d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
e1f3a66f15
freebsd-dev/sys/netinet/sctp_input.c

6061 lines
183 KiB
C
Raw Normal View History

Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*-
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
* Copyright (c) 2001-2008, by Cisco Systems, Inc. All rights reserved.
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
* Copyright (c) 2008-2011, by Randall Stewart. All rights reserved.
* Copyright (c) 2008-2011, by Michael Tuexen. All rights reserved.
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* a) Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
*
* b) Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the distribution.
*
* c) Neither the name of Cisco Systems, Inc. nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
* THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
* THE POSSIBILITY OF SUCH DAMAGE.
*/
/* $KAME: sctp_input.c,v 1.27 2005/03/06 16:04:17 itojun Exp $ */
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <netinet/sctp_os.h>
#include <netinet/sctp_var.h>
- Sysctl's move to seperate file - moved away from ifn/ifa access to sctp_ifa/sctp_ifn built and managed by the add-ip code. - cleaned up add-ip code to use the iterator - made iterator be a thread, which enables auto-asconf now. - rewrote and cleaned up source address selection (also made it use new structures). - Fixed a couple of memory leaks. - DACK now settable as to how many packets to delay as well as time. - connectx() to latest socket API, new associd arg. - Fixed issue with revoking and loosing potential to send when we inflate the flight size. We now inflate the cwnd too and deflate it later when the revoked chunk is sent or acked. - Got rid of some temp debug code - src addr selection moved to a common file (sctp_output.c) - Support for simple VRF's (we have support for multi-vfr via compile switch that is scrubbed from BSD but we won't need multi-vrf until we first get VRF :-D) - Rest of mib work for address information now done - Limit number of addresses in INIT/INIT-ACK to a #def (30). Reviewed by: gnn
2007-03-15 11:27:14 +00:00
#include <netinet/sctp_sysctl.h>
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
#include <netinet/sctp_pcb.h>
#include <netinet/sctp_header.h>
#include <netinet/sctputil.h>
#include <netinet/sctp_output.h>
#include <netinet/sctp_input.h>
#include <netinet/sctp_auth.h>
#include <netinet/sctp_indata.h>
#include <netinet/sctp_asconf.h>
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
#include <netinet/sctp_bsd_addr.h>
- send call has a reference to uio->uio_resid in the recent send code, but uio may be NULL on sendfile calls. Change to use sndlen variable. - EMSGSIZE is not being returned in non-blocking mode and needs a small tweak to look if the msg would ever fit when returning EWOULDBLOCK. - FWD-TSN has a bug in stream processing which could cause a panic. This is a follow on to the codenomicon fix. - PDAPI level 1 and 2 do not work unless the reader gets his returned buffer full. Fix so we can break out when at level 1 or 2. - Fix fast-handoff features to copy across properly on accepted sockets - Fix sctp_peeloff() system call when no true system call exists to screen arguments for errors. In cases where a real system call exists the system call itself does this. - Fix raddr leak in recent add-ip code change for bundled asconfs (even when non-bundled asconfs are received) - Make sure ipi_addr lock is held when walking global addr list. Need to change this lock type to a rwlock(). - Add don't wake flag on both input and output when the socket is closing. - When deleting an address verify the interface is correct before allowing the delete to process. This protects panda and unnumbered. - Clean up old sysctl stuff and get rid of the old Open/Net BSD structures. - Add a function to watch the ranges in the sysctl sets. - When appending in the reassembly queue, validate that the assoc has not gone to about to be freed. If so (in the middle) abort out. Note this especially effects MAC I think due to the lock/unlock they do (or with LOCK testing in place). - Netstat patch to get rid of warnings. - Make sure that no data gets queued to inactive/unconfirmed destinations. This especially effect CMT but also makes a impact on regular SCTP as well. - During init collision when we detect seq number out of sync we need to treat it like Case C and discard the cookie (no invarient needed here). - Atomic access to the random store. - When we declare a vtag good, we need to shove it into the time wait hash to prevent further use. When the tag is put into the assoc hash, we need to remove it from the twait hash (where it will surely be). This prevents duplicate tag assignments. - Move decr-ref count to better protect sysctl out of data. - ltrace error corrections in sctp6_usrreq.c - Add hook for interface up/down to be sent to us. - Make sysctl() exported structures independent of processor architecture. - Fix route and src addr cache clearing for delete address case. - Make sure address marked SCTP_DEL_IP_ADDRESS is never selected as src addr. - in icmp handling fixed so we actually look at the icmp codes to figure out what to do. - Modified mobility code. Reception of DELETE IP ADDRESS for a primary destination and SET PRIMARY for a new primary destination is used for retransmission trigger to the new primary destination. Also, in this case, destination of chunks in send_queue are changed to the new primary destination. - Fix so that we disallow sending by mbuf to ever have EEOR mode set upon it. Approved by: re@freebsd.org (B Mah)
2007-09-08 17:48:46 +00:00
#include <netinet/sctp_timer.h>
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
#include <netinet/sctp_crc32.h>
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
#include <netinet/udp.h>
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
#include <sys/smp.h>
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
static void
sctp_stop_all_cookie_timers(struct sctp_tcb *stcb)
{
struct sctp_nets *net;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
/*
* This now not only stops all cookie timers it also stops any INIT
* timers as well. This will make sure that the timers are stopped
* in all collision cases.
*/
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_LOCK_ASSERT(stcb);
TAILQ_FOREACH(net, &stcb->asoc.nets, sctp_next) {
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
if (net->rxt_timer.type == SCTP_TIMER_TYPE_COOKIE) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_COOKIE,
stcb->sctp_ep,
stcb,
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_1);
} else if (net->rxt_timer.type == SCTP_TIMER_TYPE_INIT) {
sctp_timer_stop(SCTP_TIMER_TYPE_INIT,
stcb->sctp_ep,
stcb,
net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_2);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
}
/* INIT handler */
static void
sctp_handle_init(struct mbuf *m, int iphlen, int offset, struct sctphdr *sh,
struct sctp_init_chunk *cp, struct sctp_inpcb *inp, struct sctp_tcb *stcb,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
struct sctp_nets *net, int *abort_no_unlock, uint32_t vrf_id, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_init *init;
struct mbuf *op_err;
uint32_t init_limit;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "sctp_handle_init: handling INIT tcb:%p\n",
stcb);
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
if (stcb == NULL) {
SCTP_INP_RLOCK(inp);
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
goto outnow;
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
op_err = NULL;
init = &cp->init;
/* First are we accepting? */
if ((inp->sctp_socket->so_qlimit == 0) && (stcb == NULL)) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_init: Abort, so_qlimit:%d\n",
inp->sctp_socket->so_qlimit);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* FIX ME ?? What about TCP model and we have a
- Define changes in sctp.h - Bug in CA that does not get us incrementing the PBA properly which made us more conservative. - comment updated in sctp_input.c - memsets added before we log - added arg to hmac id's MFC after: 2 weeks
2008-05-20 09:51:36 +00:00
* match/restart case? Actually no fix is needed. the lookup
* will always find the existing assoc so stcb would not be
* NULL. It may be questionable to do this since we COULD
* just send back the INIT-ACK and hope that the app did
* accept()'s by the time the COOKIE was sent. But there is
* a price to pay for COOKIE generation and I don't want to
* pay it on the chance that the app will actually do some
* accepts(). The App just looses and should NOT be in this
* state :-)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (ntohs(cp->ch.chunk_length) < sizeof(struct sctp_init_chunk)) {
/* Invalid length */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* validate parameters */
if (init->initiate_tag == 0) {
/* protocol error... send abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (ntohl(init->a_rwnd) < SCTP_MIN_RWND) {
/* invalid parameter... send abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (init->num_inbound_streams == 0) {
/* protocol error... send abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (init->num_outbound_streams == 0) {
/* protocol error... send abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, op_err,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
init_limit = offset + ntohs(cp->ch.chunk_length);
if (sctp_validate_init_auth_params(m, offset + sizeof(*cp),
init_limit)) {
/* auth parameter(s) error... send abort */
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sctp_abort_association(inp, stcb, m, iphlen, sh, NULL, vrf_id, port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (stcb)
*abort_no_unlock = 1;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
goto outnow;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* send an INIT-ACK w/cookie */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "sctp_handle_init: sending INIT-ACK\n");
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sctp_send_initiate_ack(inp, stcb, m, iphlen, offset, sh, cp, vrf_id, port,
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
((stcb == NULL) ? SCTP_HOLDS_LOCK : SCTP_NOT_LOCKED));
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
outnow:
if (stcb == NULL) {
SCTP_INP_RUNLOCK(inp);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* process peer "INIT/INIT-ACK" chunk returns value < 0 on error
*/
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
int
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_is_there_unsent_data(struct sctp_tcb *stcb, int so_locked
#if !defined(__APPLE__) && !defined(SCTP_SO_LOCK_TESTING)
SCTP_UNUSED
#endif
)
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
{
int unsent_data = 0;
Add stream scheduling support. This work is based on a patch received from Robin Seggelmann. MFC after: 3 months.
2011-01-23 19:36:28 +00:00
unsigned int i;
struct sctp_stream_queue_pending *sp;
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
struct sctp_association *asoc;
/*
* This function returns the number of streams that have true unsent
* data on them. Note that as it looks through it will clean up any
* places that have old data that has been sent but left at top of
* stream queue.
*/
asoc = &stcb->asoc;
SCTP_TCB_SEND_LOCK(stcb);
Add stream scheduling support. This work is based on a patch received from Robin Seggelmann. MFC after: 3 months.
2011-01-23 19:36:28 +00:00
if (!stcb->asoc.ss_functions.sctp_ss_is_empty(stcb, asoc)) {
/* Check to see if some data queued */
for (i = 0; i < stcb->asoc.streamoutcnt; i++) {
/* sa_ignore FREED_MEMORY */
sp = TAILQ_FIRST(&stcb->asoc.strmout[i].outqueue);
if (sp == NULL) {
continue;
}
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
if ((sp->msg_is_complete) &&
(sp->length == 0) &&
(sp->sender_all_done)) {
/*
* We are doing differed cleanup. Last time
* through when we took all the data the
* sender_all_done was not set.
*/
if (sp->put_last_out == 0) {
SCTP_PRINTF("Gak, put out entire msg with NO end!-1\n");
SCTP_PRINTF("sender_done:%d len:%d msg_comp:%d put_last_out:%d\n",
sp->sender_all_done,
sp->length,
sp->msg_is_complete,
sp->put_last_out);
}
Add stream scheduling support. This work is based on a patch received from Robin Seggelmann. MFC after: 3 months.
2011-01-23 19:36:28 +00:00
atomic_subtract_int(&stcb->asoc.stream_queue_cnt, 1);
TAILQ_REMOVE(&stcb->asoc.strmout[i].outqueue, sp, next);
Delay the assignment of a path for DATA chunk until they hit the sent_queue. Honor a given path when the SCTP_ADDR_OVER flag is set. MFC after: 2 weeks.
2010-09-15 23:10:45 +00:00
if (sp->net) {
sctp_free_remote_addr(sp->net);
sp->net = NULL;
}
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
if (sp->data) {
sctp_m_freem(sp->data);
sp->data = NULL;
}
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_strmoq(stcb, sp, so_locked);
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
} else {
unsent_data++;
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
break;
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
}
}
}
SCTP_TCB_SEND_UNLOCK(stcb);
return (unsent_data);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
static int
sctp_process_init(struct sctp_init_chunk *cp, struct sctp_tcb *stcb,
struct sctp_nets *net)
{
struct sctp_init *init;
struct sctp_association *asoc;
struct sctp_nets *lnet;
unsigned int i;
init = &cp->init;
asoc = &stcb->asoc;
/* save off parameters */
asoc->peer_vtag = ntohl(init->initiate_tag);
asoc->peers_rwnd = ntohl(init->a_rwnd);
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
/* init tsn's */
asoc->highest_tsn_inside_map = asoc->asconf_seq_in = ntohl(init->initial_tsn) - 1;
Delay the assignment of a path for DATA chunk until they hit the sent_queue. Honor a given path when the SCTP_ADDR_OVER flag is set. MFC after: 2 weeks.
2010-09-15 23:10:45 +00:00
if (!TAILQ_EMPTY(&asoc->nets)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* update any ssthresh's that may have a default */
TAILQ_FOREACH(lnet, &asoc->nets, sctp_next) {
lnet->ssthresh = asoc->peers_rwnd;
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & (SCTP_CWND_MONITOR_ENABLE | SCTP_CWND_LOGGING_ENABLE)) {
- Fix so ifn's are properly deleted when the ref count goes to 0. - Fix so VRF's will clean themselves up when no references are around. - Allow sctp_ifa to be passed into inpcb_bind, addr_mgmt_ep_sa to bypass normal validation checks. - turn auto-asconf off for subset bound sockets - Moves all logging to use KTR. This gets rid of most of the logging #ifdef's with a few exceptions reducing the number of config options for SCTP.
2007-06-14 22:59:04 +00:00
sctp_log_cwnd(stcb, lnet, 0, SCTP_CWND_INITIALIZATION);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
SCTP_TCB_SEND_LOCK(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->pre_open_streams > ntohs(init->num_inbound_streams)) {
unsigned int newcnt;
struct sctp_stream_out *outs;
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
struct sctp_stream_queue_pending *sp, *nsp;
struct sctp_tmit_chunk *chk, *nchk;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
* Fix a bug where PR-SCTP settings are ignore when using implicit association setup. * Fix a bug where message with illegal stream ids are not deleted. * Fix a crash when reporting back unsent messages from the send_queue. * Fix a bug related to INIT retransmission when the socket is already closed. * Fix a bug where associations were stalled when partial delivery API was enabled. * Fix a bug where the receive buffer size was smaller than the partial_delivery_point. Approved by: re, rrs (mentor) MFC after: One day.
2009-08-15 21:10:52 +00:00
/* abandon the upper streams */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
newcnt = ntohs(init->num_inbound_streams);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH_SAFE(chk, &asoc->send_queue, sctp_next, nchk) {
if (chk->rec.data.stream_number >= newcnt) {
TAILQ_REMOVE(&asoc->send_queue, chk, sctp_next);
asoc->send_queue_cnt--;
if (chk->data != NULL) {
sctp_free_bufspace(stcb, asoc, chk, 1);
sctp_ulp_notify(SCTP_NOTIFY_DG_FAIL, stcb,
SCTP_NOTIFY_DATAGRAM_UNSENT, chk, SCTP_SO_NOT_LOCKED);
if (chk->data) {
sctp_m_freem(chk->data);
chk->data = NULL;
* Fix a bug where PR-SCTP settings are ignore when using implicit association setup. * Fix a bug where message with illegal stream ids are not deleted. * Fix a crash when reporting back unsent messages from the send_queue. * Fix a bug related to INIT retransmission when the socket is already closed. * Fix a bug where associations were stalled when partial delivery API was enabled. * Fix a bug where the receive buffer size was smaller than the partial_delivery_point. Approved by: re, rrs (mentor) MFC after: One day.
2009-08-15 21:10:52 +00:00
}
}
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_chunk(stcb, chk, SCTP_SO_NOT_LOCKED);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
/* sa_ignore FREED_MEMORY */
* Fix a bug where PR-SCTP settings are ignore when using implicit association setup. * Fix a bug where message with illegal stream ids are not deleted. * Fix a crash when reporting back unsent messages from the send_queue. * Fix a bug related to INIT retransmission when the socket is already closed. * Fix a bug where associations were stalled when partial delivery API was enabled. * Fix a bug where the receive buffer size was smaller than the partial_delivery_point. Approved by: re, rrs (mentor) MFC after: One day.
2009-08-15 21:10:52 +00:00
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->strmout) {
for (i = newcnt; i < asoc->pre_open_streams; i++) {
outs = &asoc->strmout[i];
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH_SAFE(sp, &outs->outqueue, next, nsp) {
* Fix a bug where PR-SCTP settings are ignore when using implicit association setup. * Fix a bug where message with illegal stream ids are not deleted. * Fix a crash when reporting back unsent messages from the send_queue. * Fix a bug related to INIT retransmission when the socket is already closed. * Fix a bug where associations were stalled when partial delivery API was enabled. * Fix a bug where the receive buffer size was smaller than the partial_delivery_point. Approved by: re, rrs (mentor) MFC after: One day.
2009-08-15 21:10:52 +00:00
TAILQ_REMOVE(&outs->outqueue, sp, next);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->stream_queue_cnt--;
sctp_ulp_notify(SCTP_NOTIFY_SPECIAL_SP_FAIL,
stcb, SCTP_NOTIFY_DATAGRAM_UNSENT,
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sp, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (sp->data) {
sctp_m_freem(sp->data);
sp->data = NULL;
}
Delay the assignment of a path for DATA chunk until they hit the sent_queue. Honor a given path when the SCTP_ADDR_OVER flag is set. MFC after: 2 weeks.
2010-09-15 23:10:45 +00:00
if (sp->net) {
sctp_free_remote_addr(sp->net);
sp->net = NULL;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Free the chunk */
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_strmoq(stcb, sp, SCTP_SO_NOT_LOCKED);
- Fixed 1-2-1 model to not worry about associd in sockopts - Fixed RTOinfo for bounding. - Fixed connect() to return ECONNREFUSED when an ABORT is received. - Added comments to direct Static Analysis not to look at some things it does not understand (comments are /* sa_ignore XXXXX */) - Bind when colliding was broken, missing not_found = 1 before checking to see if the port was in use caused endless bind loop. - Cookie life needs to be in milliseconds to conform to socket api. - Cookie life is not supposed to change if its 0, On the assoc level set we changed it to 0 opps. - Two more static analysis issues identified by the cisco tool. Null checks needed. - An issue for sendfile(). Need to validate the correct input argument. - When sending failed due to a no route to host, we leaked the mbuf chain failing to call m_freem(). - Fix #ifdef issue for getting hash block len when HAVE_SHA2 is NOT defined Reviewed by: gnn
2007-05-17 12:16:24 +00:00
/* sa_ignore FREED_MEMORY */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
}
* Fix a bug where PR-SCTP settings are ignore when using implicit association setup. * Fix a bug where message with illegal stream ids are not deleted. * Fix a crash when reporting back unsent messages from the send_queue. * Fix a bug related to INIT retransmission when the socket is already closed. * Fix a bug where associations were stalled when partial delivery API was enabled. * Fix a bug where the receive buffer size was smaller than the partial_delivery_point. Approved by: re, rrs (mentor) MFC after: One day.
2009-08-15 21:10:52 +00:00
/* cut back the count */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->pre_open_streams = newcnt;
}
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
SCTP_TCB_SEND_UNLOCK(stcb);
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
asoc->strm_realoutsize = asoc->streamoutcnt = asoc->pre_open_streams;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/* EY - nr_sack: initialize highest tsn in nr_mapping_array */
asoc->highest_tsn_inside_nr_map = asoc->highest_tsn_inside_map;
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MAP_LOGGING_ENABLE) {
- Fix so ifn's are properly deleted when the ref count goes to 0. - Fix so VRF's will clean themselves up when no references are around. - Allow sctp_ifa to be passed into inpcb_bind, addr_mgmt_ep_sa to bypass normal validation checks. - turn auto-asconf off for subset bound sockets - Moves all logging to use KTR. This gets rid of most of the logging #ifdef's with a few exceptions reducing the number of config options for SCTP.
2007-06-14 22:59:04 +00:00
sctp_log_map(0, 5, asoc->highest_tsn_inside_map, SCTP_MAP_SLIDE_RESULT);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* This is the next one we expect */
asoc->str_reset_seq_in = asoc->asconf_seq_in + 1;
asoc->mapping_array_base_tsn = ntohl(init->initial_tsn);
- Out with some printfs. - Fix a initialization of last_tsn_used - Fix handling of mapped IPv4 addresses Obtained from: Michael Tuexen and I :-) MFC after: 1 week
2008-07-29 09:06:35 +00:00
asoc->tsn_last_delivered = asoc->cumulative_tsn = asoc->asconf_seq_in;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->advanced_peer_ack_point = asoc->last_acked_seq;
/* open the requested streams */
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->strmin != NULL) {
/* Free the old ones */
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
struct sctp_queued_to_read *ctl, *nctl;
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
for (i = 0; i < asoc->streamincnt; i++) {
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH_SAFE(ctl, &asoc->strmin[i].inqueue, next, nctl) {
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
TAILQ_REMOVE(&asoc->strmin[i].inqueue, ctl, next);
sctp_free_remote_addr(ctl->whoFrom);
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
ctl->whoFrom = NULL;
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
sctp_m_freem(ctl->data);
ctl->data = NULL;
sctp_free_a_readq(stcb, ctl);
}
}
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTP_FREE(asoc->strmin, SCTP_M_STRMI);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Turns out we would reset the TSN seq counter during a colliding INIT. This if fine except when we have data outstanding... we basically reset it to the previous value it was.. so then we end up assigning the same TSN to two different data chunks. This patch: 1) Finds a missing lock for when we change the stream numbers during COOKIE and INIT-ACK processing.. we were NOT locking the send_buffer.. which COULD cause problems (found by inspection looking for <2>) 2) Fixes a case during a colliding INIT where we incorrectly reset the sending Sequence thus in some cases duplicately assigning a TSN. 3) Additional enhancments to logging so we can see strm/tsn in the receiver AND new tracking to watch what the sender is doing with TSN and STRM seq's. Approved by: gnn
2006-11-11 15:59:01 +00:00
asoc->streamincnt = ntohs(init->num_outbound_streams);
if (asoc->streamincnt > MAX_SCTP_STREAMS) {
asoc->streamincnt = MAX_SCTP_STREAMS;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_MALLOC(asoc->strmin, struct sctp_stream_in *, asoc->streamincnt *
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
sizeof(struct sctp_stream_in), SCTP_M_STRMI);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->strmin == NULL) {
/* we didn't get memory for the streams! */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "process_init: couldn't get memory for the streams!\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
for (i = 0; i < asoc->streamincnt; i++) {
asoc->strmin[i].stream_no = i;
asoc->strmin[i].last_sequence_delivered = 0xffff;
/*
* U-stream ranges will be set when the cookie is unpacked.
* Or for the INIT sender they are un set (if pr-sctp not
* supported) when the INIT-ACK arrives.
*/
TAILQ_INIT(&asoc->strmin[i].inqueue);
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave.
2007-04-22 11:06:27 +00:00
asoc->strmin[i].delivery_started = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* load_address_from_init will put the addresses into the
* association when the COOKIE is processed or the INIT-ACK is
* processed. Both types of COOKIE's existing and new call this
* routine. It will remove addresses that are no longer in the
* association (for the restarting case where addresses are
* removed). Up front when the INIT arrives we will discard it if it
* is a restart and new addresses have been added.
*/
- Fixed 1-2-1 model to not worry about associd in sockopts - Fixed RTOinfo for bounding. - Fixed connect() to return ECONNREFUSED when an ABORT is received. - Added comments to direct Static Analysis not to look at some things it does not understand (comments are /* sa_ignore XXXXX */) - Bind when colliding was broken, missing not_found = 1 before checking to see if the port was in use caused endless bind loop. - Cookie life needs to be in milliseconds to conform to socket api. - Cookie life is not supposed to change if its 0, On the assoc level set we changed it to 0 opps. - Two more static analysis issues identified by the cisco tool. Null checks needed. - An issue for sendfile(). Need to validate the correct input argument. - When sending failed due to a no route to host, we leaked the mbuf chain failing to call m_freem(). - Fix #ifdef issue for getting hash block len when HAVE_SHA2 is NOT defined Reviewed by: gnn
2007-05-17 12:16:24 +00:00
/* sa_ignore MEMLEAK */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (0);
}
/*
* INIT-ACK message processing/consumption returns value < 0 on error
*/
static int
sctp_process_init_ack(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_init_ack_chunk *cp, struct sctp_tcb *stcb,
- Take out the broken table-id concept. Panda Routers have a M-VRF concept that is NOT well thought out for a multi-homed transport protocol. So the useless table-id entries passed around need to be removed. - Add a event timer for the zero copy api. - Fix a bug in sctp_timer.c when searching for an alternate with the largest ssthresh (the compare was wrong).
2007-06-01 11:19:54 +00:00
struct sctp_nets *net, int *abort_no_unlock, uint32_t vrf_id)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_association *asoc;
struct mbuf *op_err;
int retval, abort_flag;
uint32_t initack_limit;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
int nat_friendly = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* First verify that we have no illegal param's */
abort_flag = 0;
op_err = NULL;
op_err = sctp_arethere_unrecognized_parameters(m,
(offset + sizeof(struct sctp_init_chunk)),
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
&abort_flag, (struct sctp_chunkhdr *)cp, &nat_friendly);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_flag) {
/* Send an abort and notify peer */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_abort_an_association(stcb->sctp_ep, stcb, SCTP_CAUSE_PROTOCOL_VIOLATION, op_err, SCTP_SO_NOT_LOCKED);
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
asoc = &stcb->asoc;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
asoc->peer_supports_nat = (uint8_t) nat_friendly;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* process the peer's parameters in the INIT-ACK */
retval = sctp_process_init((struct sctp_init_chunk *)cp, stcb, net);
if (retval < 0) {
return (retval);
}
initack_limit = offset + ntohs(cp->ch.chunk_length);
/* load all addresses */
if ((retval = sctp_load_addresses_from_init(stcb, m, iphlen,
(offset + sizeof(struct sctp_init_chunk)), initack_limit, sh,
NULL))) {
/* Huh, we should abort */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"Load addresses from INIT causes an abort %d\n",
retval);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
NULL, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
- added pre-checks to the bindx call. - use proper tick gathering macro instead of ticks directly. - Placed reasonable boundaries on sets that a user can do that are converted to ticks from ms. - Fix CMT_PF to always check to be sure CMT is on. - Fix ticks use of CMT_PF. - put back code to allow asconfs to be queued while INITs are in flight and before the assoc is established. - During window probes, an ack'd packet might be left with the window probe mark on it causing it to be retransmitted. Change so that the flight decrease macro clears the window_probe mark. - Additional logging flight size/reading and ASOC LOG. This is only enabled if you manually insert things into opt_sctp.h since its a set of debug code only. - Found an interesting SMP race in the way data was appended which could cause a reader to lose a part of a message, had to reorder when we marked the message was complete to after the data was appended. - bug in ADD-IP for the subset bound socket case when the peer has only one address - fix ASCONF implicit success/error handling case - proper support of jails in Freebsd 6> - copy out the timeval for the 64 bit sparc world on cookie-echo alignment error crashes without this). Approved by: re(Ken Smith)
2007-07-17 20:58:26 +00:00
/* if the peer doesn't support asconf, flush the asconf queue */
if (asoc->peer_supports_asconf == 0) {
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
struct sctp_asconf_addr *param, *nparam;
- added pre-checks to the bindx call. - use proper tick gathering macro instead of ticks directly. - Placed reasonable boundaries on sets that a user can do that are converted to ticks from ms. - Fix CMT_PF to always check to be sure CMT is on. - Fix ticks use of CMT_PF. - put back code to allow asconfs to be queued while INITs are in flight and before the assoc is established. - During window probes, an ack'd packet might be left with the window probe mark on it causing it to be retransmitted. Change so that the flight decrease macro clears the window_probe mark. - Additional logging flight size/reading and ASOC LOG. This is only enabled if you manually insert things into opt_sctp.h since its a set of debug code only. - Found an interesting SMP race in the way data was appended which could cause a reader to lose a part of a message, had to reorder when we marked the message was complete to after the data was appended. - bug in ADD-IP for the subset bound socket case when the peer has only one address - fix ASCONF implicit success/error handling case - proper support of jails in Freebsd 6> - copy out the timeval for the 64 bit sparc world on cookie-echo alignment error crashes without this). Approved by: re(Ken Smith)
2007-07-17 20:58:26 +00:00
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH_SAFE(param, &asoc->asconf_queue, next, nparam) {
TAILQ_REMOVE(&asoc->asconf_queue, param, next);
SCTP_FREE(param, SCTP_M_ASC_ADDR);
- added pre-checks to the bindx call. - use proper tick gathering macro instead of ticks directly. - Placed reasonable boundaries on sets that a user can do that are converted to ticks from ms. - Fix CMT_PF to always check to be sure CMT is on. - Fix ticks use of CMT_PF. - put back code to allow asconfs to be queued while INITs are in flight and before the assoc is established. - During window probes, an ack'd packet might be left with the window probe mark on it causing it to be retransmitted. Change so that the flight decrease macro clears the window_probe mark. - Additional logging flight size/reading and ASOC LOG. This is only enabled if you manually insert things into opt_sctp.h since its a set of debug code only. - Found an interesting SMP race in the way data was appended which could cause a reader to lose a part of a message, had to reorder when we marked the message was complete to after the data was appended. - bug in ADD-IP for the subset bound socket case when the peer has only one address - fix ASCONF implicit success/error handling case - proper support of jails in Freebsd 6> - copy out the timeval for the 64 bit sparc world on cookie-echo alignment error crashes without this). Approved by: re(Ken Smith)
2007-07-17 20:58:26 +00:00
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.peer_hmac_id = sctp_negotiate_hmacid(stcb->asoc.peer_hmacs,
stcb->asoc.local_hmacs);
if (op_err) {
sctp_queue_op_err(stcb, op_err);
/* queuing will steal away the mbuf chain to the out queue */
op_err = NULL;
}
/* extract the cookie and queue it to "echo" it back... */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
net->error_count = 0;
/*
* Cancel the INIT timer, We do this first before queueing the
* cookie. We always cancel at the primary to assue that we are
* canceling the timer started by the INIT which always goes to the
* primary.
*/
sctp_timer_stop(SCTP_TIMER_TYPE_INIT, stcb->sctp_ep, stcb,
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
asoc->primary_destination, SCTP_FROM_SCTP_INPUT + SCTP_LOC_4);
/* calculate the RTO */
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
net->RTO = sctp_calculate_rto(stcb, asoc, net, &asoc->time_entered, sctp_align_safe_nocopy,
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
SCTP_RTT_FROM_NON_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
retval = sctp_send_cookie_echo(m, offset, stcb, net);
if (retval < 0) {
/*
* No cookie, we probably should send a op error. But in any
* case if there is no cookie in the INIT-ACK, we can
* abandon the peer, its broke.
*/
if (retval == -3) {
/* We abort with an error of missing mandatory param */
op_err =
sctp_generate_invmanparam(SCTP_CAUSE_MISSING_PARAM);
if (op_err) {
/*
* Expand beyond to include the mandatory
* param cookie
*/
struct sctp_inv_mandatory_param *mp;
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_LEN(op_err) =
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sizeof(struct sctp_inv_mandatory_param);
mp = mtod(op_err,
struct sctp_inv_mandatory_param *);
/* Subtract the reserved param */
mp->length =
htons(sizeof(struct sctp_inv_mandatory_param) - 2);
mp->num_param = htonl(1);
mp->param = htons(SCTP_STATE_COOKIE);
mp->resv = 0;
}
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sh, op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
return (retval);
}
return (0);
}
static void
sctp_handle_heartbeat_ack(struct sctp_heartbeat_chunk *cp,
struct sctp_tcb *stcb, struct sctp_nets *net)
{
struct sockaddr_storage store;
Fix some CLANG warnings. One clang warning is left due to the fact that its bogus.. nam->sa_family will not change from AF_INET6 to AF_INET (but clang thinks it does ;-D)
2010-09-05 13:41:45 +00:00
struct sctp_nets *r_net, *f_net;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
struct timeval tv;
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
int req_prim = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#ifdef INET
struct sockaddr_in *sin;
#endif
#ifdef INET6
struct sockaddr_in6 *sin6;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ntohs(cp->ch.chunk_length) != sizeof(struct sctp_heartbeat_chunk)) {
/* Invalid length */
return;
}
memset(&store, 0, sizeof(store));
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
switch (cp->heartbeat.hb_info.addr_family) {
#ifdef INET
case AF_INET:
if (cp->heartbeat.hb_info.addr_len == sizeof(struct sockaddr_in)) {
sin = (struct sockaddr_in *)&store;
sin->sin_family = cp->heartbeat.hb_info.addr_family;
sin->sin_len = cp->heartbeat.hb_info.addr_len;
sin->sin_port = stcb->rport;
memcpy(&sin->sin_addr, cp->heartbeat.hb_info.address,
sizeof(sin->sin_addr));
} else {
return;
}
break;
#endif
#ifdef INET6
case AF_INET6:
if (cp->heartbeat.hb_info.addr_len == sizeof(struct sockaddr_in6)) {
sin6 = (struct sockaddr_in6 *)&store;
sin6->sin6_family = cp->heartbeat.hb_info.addr_family;
sin6->sin6_len = cp->heartbeat.hb_info.addr_len;
sin6->sin6_port = stcb->rport;
memcpy(&sin6->sin6_addr, cp->heartbeat.hb_info.address,
sizeof(sin6->sin6_addr));
} else {
return;
}
break;
#endif
default:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return;
}
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
r_net = sctp_findnet(stcb, (struct sockaddr *)&store);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (r_net == NULL) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Huh? I can't find the address I sent it to, discard\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return;
}
if ((r_net && (r_net->dest_state & SCTP_ADDR_UNCONFIRMED)) &&
(r_net->heartbeat_random1 == cp->heartbeat.hb_info.random_value1) &&
(r_net->heartbeat_random2 == cp->heartbeat.hb_info.random_value2)) {
/*
* If the its a HB and it's random value is correct when can
* confirm the destination.
*/
r_net->dest_state &= ~SCTP_ADDR_UNCONFIRMED;
- Sysctl's move to seperate file - moved away from ifn/ifa access to sctp_ifa/sctp_ifn built and managed by the add-ip code. - cleaned up add-ip code to use the iterator - made iterator be a thread, which enables auto-asconf now. - rewrote and cleaned up source address selection (also made it use new structures). - Fixed a couple of memory leaks. - DACK now settable as to how many packets to delay as well as time. - connectx() to latest socket API, new associd arg. - Fixed issue with revoking and loosing potential to send when we inflate the flight size. We now inflate the cwnd too and deflate it later when the revoked chunk is sent or acked. - Got rid of some temp debug code - src addr selection moved to a common file (sctp_output.c) - Support for simple VRF's (we have support for multi-vfr via compile switch that is scrubbed from BSD but we won't need multi-vrf until we first get VRF :-D) - Rest of mib work for address information now done - Limit number of addresses in INIT/INIT-ACK to a #def (30). Reviewed by: gnn
2007-03-15 11:27:14 +00:00
if (r_net->dest_state & SCTP_ADDR_REQ_PRIMARY) {
stcb->asoc.primary_destination = r_net;
r_net->dest_state &= ~SCTP_ADDR_WAS_PRIMARY;
r_net->dest_state &= ~SCTP_ADDR_REQ_PRIMARY;
Fix some CLANG warnings. One clang warning is left due to the fact that its bogus.. nam->sa_family will not change from AF_INET6 to AF_INET (but clang thinks it does ;-D)
2010-09-05 13:41:45 +00:00
f_net = TAILQ_FIRST(&stcb->asoc.nets);
if (f_net != r_net) {
- Sysctl's move to seperate file - moved away from ifn/ifa access to sctp_ifa/sctp_ifn built and managed by the add-ip code. - cleaned up add-ip code to use the iterator - made iterator be a thread, which enables auto-asconf now. - rewrote and cleaned up source address selection (also made it use new structures). - Fixed a couple of memory leaks. - DACK now settable as to how many packets to delay as well as time. - connectx() to latest socket API, new associd arg. - Fixed issue with revoking and loosing potential to send when we inflate the flight size. We now inflate the cwnd too and deflate it later when the revoked chunk is sent or acked. - Got rid of some temp debug code - src addr selection moved to a common file (sctp_output.c) - Support for simple VRF's (we have support for multi-vfr via compile switch that is scrubbed from BSD but we won't need multi-vrf until we first get VRF :-D) - Rest of mib work for address information now done - Limit number of addresses in INIT/INIT-ACK to a #def (30). Reviewed by: gnn
2007-03-15 11:27:14 +00:00
/*
* first one on the list is NOT the primary
* sctp_cmpaddr() is much more efficent if
* the primary is the first on the list,
* make it so.
*/
Fix some CLANG warnings. One clang warning is left due to the fact that its bogus.. nam->sa_family will not change from AF_INET6 to AF_INET (but clang thinks it does ;-D)
2010-09-05 13:41:45 +00:00
TAILQ_REMOVE(&stcb->asoc.nets, r_net, sctp_next);
TAILQ_INSERT_HEAD(&stcb->asoc.nets, r_net, sctp_next);
- Sysctl's move to seperate file - moved away from ifn/ifa access to sctp_ifa/sctp_ifn built and managed by the add-ip code. - cleaned up add-ip code to use the iterator - made iterator be a thread, which enables auto-asconf now. - rewrote and cleaned up source address selection (also made it use new structures). - Fixed a couple of memory leaks. - DACK now settable as to how many packets to delay as well as time. - connectx() to latest socket API, new associd arg. - Fixed issue with revoking and loosing potential to send when we inflate the flight size. We now inflate the cwnd too and deflate it later when the revoked chunk is sent or acked. - Got rid of some temp debug code - src addr selection moved to a common file (sctp_output.c) - Support for simple VRF's (we have support for multi-vfr via compile switch that is scrubbed from BSD but we won't need multi-vrf until we first get VRF :-D) - Rest of mib work for address information now done - Limit number of addresses in INIT/INIT-ACK to a #def (30). Reviewed by: gnn
2007-03-15 11:27:14 +00:00
}
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
req_prim = 1;
- Sysctl's move to seperate file - moved away from ifn/ifa access to sctp_ifa/sctp_ifn built and managed by the add-ip code. - cleaned up add-ip code to use the iterator - made iterator be a thread, which enables auto-asconf now. - rewrote and cleaned up source address selection (also made it use new structures). - Fixed a couple of memory leaks. - DACK now settable as to how many packets to delay as well as time. - connectx() to latest socket API, new associd arg. - Fixed issue with revoking and loosing potential to send when we inflate the flight size. We now inflate the cwnd too and deflate it later when the revoked chunk is sent or acked. - Got rid of some temp debug code - src addr selection moved to a common file (sctp_output.c) - Support for simple VRF's (we have support for multi-vfr via compile switch that is scrubbed from BSD but we won't need multi-vrf until we first get VRF :-D) - Rest of mib work for address information now done - Limit number of addresses in INIT/INIT-ACK to a #def (30). Reviewed by: gnn
2007-03-15 11:27:14 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_CONFIRMED,
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
stcb, 0, (void *)r_net, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
r_net->error_count = 0;
r_net->hb_responded = 1;
tv.tv_sec = cp->heartbeat.hb_info.time_value_1;
tv.tv_usec = cp->heartbeat.hb_info.time_value_2;
if (r_net->dest_state & SCTP_ADDR_NOT_REACHABLE) {
r_net->dest_state &= ~SCTP_ADDR_NOT_REACHABLE;
r_net->dest_state |= SCTP_ADDR_REACHABLE;
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_UP, stcb,
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
SCTP_HEARTBEAT_SUCCESS, (void *)r_net, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* now was it the primary? if so restore */
if (r_net->dest_state & SCTP_ADDR_WAS_PRIMARY) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
(void)sctp_set_primary_addr(stcb, (struct sockaddr *)NULL, r_net);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
/*
* JRS 5/14/07 - If CMT PF is on and the destination is in PF state,
* set the destination to active state and set the cwnd to one or
* two MTU's based on whether PF1 or PF2 is being used. If a T3
* timer is running, for the destination, stop the timer because a
* PF-heartbeat was received.
*/
Improve plausibility check in sctp_handle_sack(). Allow cmt_on_off to support values 0 (no CMT), 1 (CMT), and 2 (CMT/RP). MFC after: 3 months.
2010-12-22 17:59:38 +00:00
if ((stcb->asoc.sctp_cmt_on_off > 0) &&
Fix the switching on/off of CMT using sysctl and socket option. Fix the switching on/off of PF and NR-SACKs using sysctl. Add minor improvement in handling malloc failures. Improve the address checks when sending. MFC after: 4 weeks
2010-08-28 17:59:51 +00:00
(stcb->asoc.sctp_cmt_pf > 0) &&
((net->dest_state & SCTP_ADDR_PF) == SCTP_ADDR_PF)) {
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
if (SCTP_OS_TIMER_PENDING(&net->rxt_timer.timer)) {
sctp_timer_stop(SCTP_TIMER_TYPE_SEND, stcb->sctp_ep,
stcb, net,
SCTP_FROM_SCTP_INPUT + SCTP_LOC_5);
}
net->dest_state &= ~SCTP_ADDR_PF;
Fix the switching on/off of CMT using sysctl and socket option. Fix the switching on/off of PF and NR-SACKs using sysctl. Add minor improvement in handling malloc failures. Improve the address checks when sending. MFC after: 4 weeks
2010-08-28 17:59:51 +00:00
net->cwnd = net->mtu * stcb->asoc.sctp_cmt_pf;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Destination %p moved from PF to reachable with cwnd %d.\n",
net, net->cwnd);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Now lets do a RTO with this */
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
r_net->RTO = sctp_calculate_rto(stcb, &stcb->asoc, r_net, &tv, sctp_align_safe_nocopy,
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
SCTP_RTT_FROM_NON_DATA);
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
/* Mobility adaptation */
if (req_prim) {
if ((sctp_is_mobility_feature_on(stcb->sctp_ep,
SCTP_MOBILITY_BASE) ||
sctp_is_mobility_feature_on(stcb->sctp_ep,
SCTP_MOBILITY_FASTHANDOFF)) &&
sctp_is_mobility_feature_on(stcb->sctp_ep,
SCTP_MOBILITY_PRIM_DELETED)) {
sctp_timer_stop(SCTP_TIMER_TYPE_PRIM_DELETED, stcb->sctp_ep, stcb, NULL, SCTP_FROM_SCTP_TIMER + SCTP_LOC_7);
if (sctp_is_mobility_feature_on(stcb->sctp_ep,
SCTP_MOBILITY_FASTHANDOFF)) {
sctp_assoc_immediate_retrans(stcb,
stcb->asoc.primary_destination);
}
if (sctp_is_mobility_feature_on(stcb->sctp_ep,
SCTP_MOBILITY_BASE)) {
Delay the assignment of a path for DATA chunk until they hit the sent_queue. Honor a given path when the SCTP_ADDR_OVER flag is set. MFC after: 2 weeks.
2010-09-15 23:10:45 +00:00
sctp_move_chunks_from_net(stcb,
stcb->asoc.deleted_primary);
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
}
sctp_delete_prim_timer(stcb->sctp_ep, stcb,
stcb->asoc.deleted_primary);
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
static int
sctp_handle_nat_colliding_state(struct sctp_tcb *stcb)
{
/*
* return 0 means we want you to proceed with the abort non-zero
* means no abort processing
*/
struct sctpasochead *head;
if (SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_COOKIE_WAIT) {
/* generate a new vtag and send init */
LIST_REMOVE(stcb, sctp_asocs);
stcb->asoc.my_vtag = sctp_select_a_tag(stcb->sctp_ep, stcb->sctp_ep->sctp_lport, stcb->rport, 1);
head = &SCTP_BASE_INFO(sctp_asochash)[SCTP_PCBHASH_ASOC(stcb->asoc.my_vtag, SCTP_BASE_INFO(hashasocmark))];
/*
* put it in the bucket in the vtag hash of assoc's for the
* system
*/
LIST_INSERT_HEAD(head, stcb, sctp_asocs);
sctp_send_initiate(stcb->sctp_ep, stcb, SCTP_SO_NOT_LOCKED);
return (1);
}
if (SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_COOKIE_ECHOED) {
/*
* treat like a case where the cookie expired i.e.: - dump
* current cookie. - generate a new vtag. - resend init.
*/
/* generate a new vtag and send init */
LIST_REMOVE(stcb, sctp_asocs);
stcb->asoc.state &= ~SCTP_STATE_COOKIE_ECHOED;
stcb->asoc.state |= SCTP_STATE_COOKIE_WAIT;
sctp_stop_all_cookie_timers(stcb);
sctp_toss_old_cookies(stcb, &stcb->asoc);
stcb->asoc.my_vtag = sctp_select_a_tag(stcb->sctp_ep, stcb->sctp_ep->sctp_lport, stcb->rport, 1);
head = &SCTP_BASE_INFO(sctp_asochash)[SCTP_PCBHASH_ASOC(stcb->asoc.my_vtag, SCTP_BASE_INFO(hashasocmark))];
/*
* put it in the bucket in the vtag hash of assoc's for the
* system
*/
LIST_INSERT_HEAD(head, stcb, sctp_asocs);
sctp_send_initiate(stcb->sctp_ep, stcb, SCTP_SO_NOT_LOCKED);
return (1);
}
return (0);
}
static int
sctp_handle_nat_missing_state(struct sctp_tcb *stcb,
struct sctp_nets *net)
{
/*
* return 0 means we want you to proceed with the abort non-zero
* means no abort processing
*/
if (stcb->asoc.peer_supports_auth == 0) {
SCTPDBG(SCTP_DEBUG_INPUT2, "sctp_handle_nat_missing_state: Peer does not support AUTH, cannot send an asconf\n");
return (0);
}
sctp_asconf_send_nat_state_update(stcb, net);
return (1);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
static void
sctp_handle_abort(struct sctp_abort_chunk *cp,
struct sctp_tcb *stcb, struct sctp_nets *net)
{
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
uint16_t len;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "sctp_handle_abort: handling ABORT\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL)
return;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
len = ntohs(cp->ch.chunk_length);
if (len > sizeof(struct sctp_chunkhdr)) {
/*
* Need to check the cause codes for our two magic nat
* aborts which don't kill the assoc necessarily.
*/
struct sctp_abort_chunk *cpnext;
struct sctp_missing_nat_state *natc;
uint16_t cause;
cpnext = cp;
cpnext++;
natc = (struct sctp_missing_nat_state *)cpnext;
cause = ntohs(natc->cause);
if (cause == SCTP_CAUSE_NAT_COLLIDING_STATE) {
SCTPDBG(SCTP_DEBUG_INPUT2, "Received Colliding state abort flags:%x\n",
cp->ch.chunk_flags);
if (sctp_handle_nat_colliding_state(stcb)) {
return;
}
} else if (cause == SCTP_CAUSE_NAT_MISSING_STATE) {
SCTPDBG(SCTP_DEBUG_INPUT2, "Received missing state abort flags:%x\n",
cp->ch.chunk_flags);
if (sctp_handle_nat_missing_state(stcb, net)) {
return;
}
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* stop any receive timers */
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_RECV, stcb->sctp_ep, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_6);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* notify user of the abort and clean up... */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_abort_notification(stcb, 0, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* free the tcb */
- Bug fix managing congestion parameter on immediate retransmittion by handover event (fast mobility code) - Fixed problem of mobility code which is caused by remaining parameters in the deleted primary destination. - Add a missing lock. When a peer sends an INIT, and while we are processing it to send an INIT-ACK the socket is closed, we did not hold a lock to keep the socket from going away. Add protection for this case. - Fix so that arwnd is alway uses the minimal rwnd if the user has set the socket buffer smaller. Found this when the test org decided to see what happens when you set in a rwnd of 10 bytes (which is not allowed per RFC .. 4k is minimum). - Fixes so a cookie-echo ootb will NOT cause an abort to be sent. This was happening in a MPI collision case. - Examined all panics and unless there was no recovery, moved any that were not already to INVARANTS. Approved by: re@freebsd.org (gnn)
2007-10-01 03:22:29 +00:00
#if defined(SCTP_PANIC_ON_ABORT)
printf("stcb:%p state:%d rport:%d net:%p\n",
stcb, stcb->asoc.state, stcb->rport, net);
if (!(stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET)) {
panic("Received an ABORT");
} else {
printf("No panic its in state %x closed\n", stcb->asoc.state);
}
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR_COUNTER32(sctps_aborted);
if ((SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_OPEN) ||
(SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_SHUTDOWN_RECEIVED)) {
SCTP_STAT_DECR_GAUGE32(sctps_currestab);
}
- More work on making send lock contention. - Removed free-oqueue cache. - Fix counter for sq entries - Increased the amount of information retained on ASOC_TSN logging on the association. - Made it so with the ASOC_TSN logging on sending or recieving an abort we dump the log. - Went through and added invariant's around some panic's that needed them. - decrements went to atomic_subtact_int instead of add -1 - Removed residual count increment that threw off a strm oq count. - Tracks and complaints if we don't have a LAST fragment and clean up the sp structure. - Track a new stat that counts number of abandoned msgs that happen if you close without reading. - Fix lookup of frag point to be aware of a 0 assoc-id. Reviewed by: gnn
2007-04-19 11:28:43 +00:00
#ifdef SCTP_ASOCLOG_OF_TSNS
sctp_print_out_track_log(stcb);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#endif
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
- More work on making send lock contention. - Removed free-oqueue cache. - Fix counter for sq entries - Increased the amount of information retained on ASOC_TSN logging on the association. - Made it so with the ASOC_TSN logging on sending or recieving an abort we dump the log. - Went through and added invariant's around some panic's that needed them. - decrements went to atomic_subtact_int instead of add -1 - Removed residual count increment that threw off a strm oq count. - Tracks and complaints if we don't have a LAST fragment and clean up the sp structure. - Track a new stat that counts number of abandoned msgs that happen if you close without reading. - Fix lookup of frag point to be aware of a 0 assoc-id. Reviewed by: gnn
2007-04-19 11:28:43 +00:00
#endif
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
stcb->asoc.state |= SCTP_STATE_WAS_ABORTED;
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(stcb->sctp_ep, stcb, SCTP_NORMAL_PROC,
SCTP_FROM_SCTP_INPUT + SCTP_LOC_6);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "sctp_handle_abort: finished\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
static void
sctp_handle_shutdown(struct sctp_shutdown_chunk *cp,
struct sctp_tcb *stcb, struct sctp_nets *net, int *abort_flag)
{
struct sctp_association *asoc;
int some_on_streamwheel;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_shutdown: handling SHUTDOWN\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL)
return;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
asoc = &stcb->asoc;
if ((SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_WAIT) ||
(SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return;
}
if (ntohs(cp->ch.chunk_length) != sizeof(struct sctp_shutdown_chunk)) {
/* Shutdown NOT the expected size */
return;
} else {
sctp_update_acked(stcb, cp, net, abort_flag);
Fix a bug where the system panics when a SHUTDOWN is received with an illegal TSN. Approved by: rrs (mentor) MFC after: ASAP
2009-11-18 12:17:06 +00:00
if (*abort_flag) {
return;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
if (asoc->control_pdapi) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* With a normal shutdown we assume the end of last record.
*/
SCTP_INP_READ_LOCK(stcb->sctp_ep);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
asoc->control_pdapi->end_added = 1;
asoc->control_pdapi->pdapi_aborted = 1;
asoc->control_pdapi = NULL;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_INP_READ_UNLOCK(stcb->sctp_ep);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
/* assoc was freed while we were unlocked */
SCTP_SOCKET_UNLOCK(so, 1);
return;
}
#endif
Tons of fixes to get all the 64bit issues removed. This also moves two 16 bit int's to become 32 bit values so we do not have to use atomic_add_16. Most of the changes are %p, casts and other various nasty's that were in the orignal code base. With this commit my machine will now do a build universe.. however I as yet have not tested on a 64bit machine .. it may not work :-(
2006-11-05 13:25:18 +00:00
sctp_sorwakeup(stcb->sctp_ep, stcb->sctp_socket);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* goto SHUTDOWN_RECEIVED state to block new requests */
if (stcb->sctp_socket) {
if ((SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_RECEIVED) &&
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
(SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_ACK_SENT) &&
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_SENT)) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_SHUTDOWN_RECEIVED);
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
SCTP_CLEAR_SUBSTATE(asoc, SCTP_STATE_SHUTDOWN_PENDING);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* notify upper layer that peer has initiated a
* shutdown
*/
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_PEER_SHUTDOWN, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* reset time */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)SCTP_GETTIME_TIMEVAL(&asoc->time_entered);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
if (SCTP_GET_STATE(asoc) == SCTP_STATE_SHUTDOWN_SENT) {
/*
* stop the shutdown timer, since we WILL move to
* SHUTDOWN-ACK-SENT.
*/
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_SHUTDOWN, stcb->sctp_ep, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_8);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- Consolidate the code that free's chunks to actually also call the sctp_free_remote_address() function. - Assure that when we allocate a chunk the whoTo is NULL, also when we free it and place it into the cache we NULL it (that way the consolidation code will always work). - Fix a small race, when a empty data holder is left on the stream out queue, and both sides do a shutdown, the empty data holder would prevent us from sending a SHUTDOWN-ACK and at the same time we never would cleanup the empty holder (since nothing was ever in queue). We now add a utility function that a) cleans up empty holders and b) properly determines if there are still pending data chunks on the stream out wheel. Approved by: re@freebsd.org (Ken Smith)
2007-07-02 19:22:22 +00:00
/* Now is there unsent data on a stream somewhere? */
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
some_on_streamwheel = sctp_is_there_unsent_data(stcb, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (!TAILQ_EMPTY(&asoc->send_queue) ||
!TAILQ_EMPTY(&asoc->sent_queue) ||
some_on_streamwheel) {
/* By returning we will push more data out */
return;
} else {
/* no outstanding data to send, so move on... */
/* send SHUTDOWN-ACK */
sctp_send_shutdown_ack(stcb, stcb->asoc.primary_destination);
/* move to SHUTDOWN-ACK-SENT state */
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
if ((SCTP_GET_STATE(asoc) == SCTP_STATE_OPEN) ||
(SCTP_GET_STATE(asoc) == SCTP_STATE_SHUTDOWN_RECEIVED)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_DECR_GAUGE32(sctps_currestab);
}
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_SHUTDOWN_ACK_SENT);
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
SCTP_CLEAR_SUBSTATE(asoc, SCTP_STATE_SHUTDOWN_PENDING);
Not only stop all timers when entering the SHUTDOWN_SENT state, but also when entering the SHUTDOWN_ACK_SEND state. MFC after: 3 days.
2010-11-07 14:39:40 +00:00
sctp_stop_timers_for_shutdown(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNACK, stcb->sctp_ep,
stcb, net);
}
}
static void
sctp_handle_shutdown_ack(struct sctp_shutdown_ack_chunk *cp,
Fix handling of SHUTDOWN-ACK chunk in COOKIE_WAIT and COOKIE_ECHOED. MFC after: 1 week
2010-02-20 20:30:40 +00:00
struct sctp_tcb *stcb,
struct sctp_nets *net)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_association *asoc;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
so = SCTP_INP_SO(stcb->sctp_ep);
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_shutdown_ack: handling SHUTDOWN ACK\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL)
return;
asoc = &stcb->asoc;
/* process according to association state */
Fix handling of SHUTDOWN-ACK chunk in COOKIE_WAIT and COOKIE_ECHOED. MFC after: 1 week
2010-02-20 20:30:40 +00:00
if ((SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_WAIT) ||
(SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED)) {
/* unexpected SHUTDOWN-ACK... do OOTB handling... */
sctp_send_shutdown_complete(stcb, net, 1);
SCTP_TCB_UNLOCK(stcb);
return;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_SENT) &&
(SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_ACK_SENT)) {
/* unexpected SHUTDOWN-ACK... so ignore... */
SCTP_TCB_UNLOCK(stcb);
return;
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
if (asoc->control_pdapi) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* With a normal shutdown we assume the end of last record.
*/
SCTP_INP_READ_LOCK(stcb->sctp_ep);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
asoc->control_pdapi->end_added = 1;
asoc->control_pdapi->pdapi_aborted = 1;
asoc->control_pdapi = NULL;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_INP_READ_UNLOCK(stcb->sctp_ep);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
/* assoc was freed while we were unlocked */
SCTP_SOCKET_UNLOCK(so, 1);
return;
}
#endif
Tons of fixes to get all the 64bit issues removed. This also moves two 16 bit int's to become 32 bit values so we do not have to use atomic_add_16. Most of the changes are %p, casts and other various nasty's that were in the orignal code base. With this commit my machine will now do a build universe.. however I as yet have not tested on a 64bit machine .. it may not work :-(
2006-11-05 13:25:18 +00:00
sctp_sorwakeup(stcb->sctp_ep, stcb->sctp_socket);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* are the queues empty? */
if (!TAILQ_EMPTY(&asoc->send_queue) ||
!TAILQ_EMPTY(&asoc->sent_queue) ||
Add stream scheduling support. This work is based on a patch received from Robin Seggelmann. MFC after: 3 months.
2011-01-23 19:36:28 +00:00
!stcb->asoc.ss_functions.sctp_ss_is_empty(stcb, asoc)) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_report_all_outbound(stcb, 0, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* stop the timer */
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_SHUTDOWN, stcb->sctp_ep, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_9);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* send SHUTDOWN-COMPLETE */
Fix handling of SHUTDOWN-ACK chunk in COOKIE_WAIT and COOKIE_ECHOED. MFC after: 1 week
2010-02-20 20:30:40 +00:00
sctp_send_shutdown_complete(stcb, net, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* notify upper layer protocol */
if (stcb->sctp_socket) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_ASSOC_DOWN, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) ||
(stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_IN_TCPPOOL)) {
/* Set the connected flag to disconnected */
stcb->sctp_ep->sctp_socket->so_snd.sb_cc = 0;
}
}
SCTP_STAT_INCR_COUNTER32(sctps_shutdown);
/* free the TCB but first save off the ep */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(stcb->sctp_ep, stcb, SCTP_NORMAL_PROC,
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
SCTP_FROM_SCTP_INPUT + SCTP_LOC_10);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* Skip past the param header and then we will find the chunk that caused the
* problem. There are two possiblities ASCONF or FWD-TSN other than that and
* our peer must be broken.
*/
static void
sctp_process_unrecog_chunk(struct sctp_tcb *stcb, struct sctp_paramhdr *phdr,
struct sctp_nets *net)
{
struct sctp_chunkhdr *chk;
chk = (struct sctp_chunkhdr *)((caddr_t)phdr + sizeof(*phdr));
switch (chk->chunk_type) {
case SCTP_ASCONF_ACK:
case SCTP_ASCONF:
sctp_asconf_cleanup(stcb, net);
break;
case SCTP_FORWARD_CUM_TSN:
stcb->asoc.peer_supports_prsctp = 0;
break;
default:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"Peer does not support chunk type %d(%x)??\n",
chk->chunk_type, (uint32_t) chk->chunk_type);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
}
}
/*
* Skip past the param header and then we will find the param that caused the
* problem. There are a number of param's in a ASCONF OR the prsctp param
* these will turn of specific features.
*/
static void
sctp_process_unrecog_param(struct sctp_tcb *stcb, struct sctp_paramhdr *phdr)
{
struct sctp_paramhdr *pbad;
pbad = phdr + 1;
switch (ntohs(pbad->param_type)) {
/* pr-sctp draft */
case SCTP_PRSCTP_SUPPORTED:
stcb->asoc.peer_supports_prsctp = 0;
break;
case SCTP_SUPPORTED_CHUNK_EXT:
break;
/* draft-ietf-tsvwg-addip-sctp */
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
case SCTP_HAS_NAT_SUPPORT:
stcb->asoc.peer_supports_nat = 0;
break;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_ADD_IP_ADDRESS:
case SCTP_DEL_IP_ADDRESS:
case SCTP_SET_PRIM_ADDR:
stcb->asoc.peer_supports_asconf = 0;
break;
case SCTP_SUCCESS_REPORT:
case SCTP_ERROR_CAUSE_IND:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "Huh, the peer does not support success? or error cause?\n");
SCTPDBG(SCTP_DEBUG_INPUT2,
"Turning off ASCONF to this strange peer\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.peer_supports_asconf = 0;
break;
default:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"Peer does not support param type %d(%x)??\n",
pbad->param_type, (uint32_t) pbad->param_type);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
}
}
static int
sctp_handle_error(struct sctp_chunkhdr *ch,
struct sctp_tcb *stcb, struct sctp_nets *net)
{
int chklen;
struct sctp_paramhdr *phdr;
uint16_t error_type;
uint16_t error_len;
struct sctp_association *asoc;
int adjust;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* parse through all of the errors and process */
asoc = &stcb->asoc;
phdr = (struct sctp_paramhdr *)((caddr_t)ch +
sizeof(struct sctp_chunkhdr));
chklen = ntohs(ch->chunk_length) - sizeof(struct sctp_chunkhdr);
while ((size_t)chklen >= sizeof(struct sctp_paramhdr)) {
/* Process an Error Cause */
error_type = ntohs(phdr->param_type);
error_len = ntohs(phdr->param_length);
if ((error_len > chklen) || (error_len == 0)) {
/* invalid param length for this param */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Bogus length in error param- chunk left:%d errorlen:%d\n",
chklen, error_len);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (0);
}
switch (error_type) {
case SCTP_CAUSE_INVALID_STREAM:
case SCTP_CAUSE_MISSING_PARAM:
case SCTP_CAUSE_INVALID_PARAM:
case SCTP_CAUSE_NO_USER_DATA:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Software error we got a %d back? We have a bug :/ (or do they?)\n",
error_type);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
case SCTP_CAUSE_NAT_COLLIDING_STATE:
SCTPDBG(SCTP_DEBUG_INPUT2, "Received Colliding state abort flags:%x\n",
ch->chunk_flags);
if (sctp_handle_nat_colliding_state(stcb)) {
return (0);
}
break;
case SCTP_CAUSE_NAT_MISSING_STATE:
SCTPDBG(SCTP_DEBUG_INPUT2, "Received missing state abort flags:%x\n",
ch->chunk_flags);
if (sctp_handle_nat_missing_state(stcb, net)) {
return (0);
}
break;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_CAUSE_STALE_COOKIE:
/*
* We only act if we have echoed a cookie and are
* waiting.
*/
if (SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED) {
int *p;
p = (int *)((caddr_t)phdr + sizeof(*phdr));
/* Save the time doubled */
asoc->cookie_preserve_req = ntohl(*p) << 1;
asoc->stale_cookie_count++;
if (asoc->stale_cookie_count >
asoc->max_init_times) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_abort_notification(stcb, 0, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* now free the asoc */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(stcb->sctp_ep, stcb, SCTP_NORMAL_PROC,
SCTP_FROM_SCTP_INPUT + SCTP_LOC_11);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
/* blast back to INIT state */
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
sctp_toss_old_cookies(stcb, &stcb->asoc);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->state &= ~SCTP_STATE_COOKIE_ECHOED;
asoc->state |= SCTP_STATE_COOKIE_WAIT;
sctp_stop_all_cookie_timers(stcb);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_send_initiate(stcb->sctp_ep, stcb, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
break;
case SCTP_CAUSE_UNRESOLVABLE_ADDR:
/*
* Nothing we can do here, we don't do hostname
* addresses so if the peer does not like my IPv6
* (or IPv4 for that matter) it does not matter. If
* they don't support that type of address, they can
* NOT possibly get that packet type... i.e. with no
* IPv6 you can't recieve a IPv6 packet. so we can
* safely ignore this one. If we ever added support
* for HOSTNAME Addresses, then we would need to do
* something here.
*/
break;
case SCTP_CAUSE_UNRECOG_CHUNK:
sctp_process_unrecog_chunk(stcb, phdr, net);
break;
case SCTP_CAUSE_UNRECOG_PARAM:
sctp_process_unrecog_param(stcb, phdr);
break;
case SCTP_CAUSE_COOKIE_IN_SHUTDOWN:
/*
* We ignore this since the timer will drive out a
* new cookie anyway and there timer will drive us
* to send a SHUTDOWN_COMPLETE. We can't send one
* here since we don't have their tag.
*/
break;
case SCTP_CAUSE_DELETING_LAST_ADDR:
case SCTP_CAUSE_RESOURCE_SHORTAGE:
case SCTP_CAUSE_DELETING_SRC_ADDR:
/*
* We should NOT get these here, but in a
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
* ASCONF-ACK.
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "Peer sends ASCONF errors in a Operational Error?<%d>?\n",
error_type);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_CAUSE_OUT_OF_RESC:
/*
* And what, pray tell do we do with the fact that
* the peer is out of resources? Not really sure we
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
* could do anything but abort. I suspect this
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
* should have came WITH an abort instead of in a
* OP-ERROR.
*/
break;
default:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "sctp_handle_error: unknown error type = 0x%xh\n",
error_type);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
}
adjust = SCTP_SIZE32(error_len);
chklen -= adjust;
phdr = (struct sctp_paramhdr *)((caddr_t)phdr + adjust);
}
return (0);
}
static int
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_handle_init_ack(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_init_ack_chunk *cp, struct sctp_tcb *stcb,
- Take out the broken table-id concept. Panda Routers have a M-VRF concept that is NOT well thought out for a multi-homed transport protocol. So the useless table-id entries passed around need to be removed. - Add a event timer for the zero copy api. - Fix a bug in sctp_timer.c when searching for an alternate with the largest ssthresh (the compare was wrong).
2007-06-01 11:19:54 +00:00
struct sctp_nets *net, int *abort_no_unlock, uint32_t vrf_id)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_init_ack *init_ack;
struct mbuf *op_err;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_init_ack: handling INIT-ACK\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_init_ack: TCB is null\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
if (ntohs(cp->ch.chunk_length) < sizeof(struct sctp_init_ack_chunk)) {
/* Invalid length */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
init_ack = &cp->init;
/* validate parameters */
if (init_ack->initiate_tag == 0) {
/* protocol error... send an abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
if (ntohl(init_ack->a_rwnd) < SCTP_MIN_RWND) {
/* protocol error... send an abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
if (init_ack->num_inbound_streams == 0) {
/* protocol error... send an abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
if (init_ack->num_outbound_streams == 0) {
/* protocol error... send an abort */
op_err = sctp_generate_invmanparam(SCTP_CAUSE_INVALID_PARAM);
sctp_abort_association(stcb->sctp_ep, stcb, m, iphlen, sh,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
op_err, 0, net->port);
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
*abort_no_unlock = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (-1);
}
/* process according to association state... */
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
switch (stcb->asoc.state & SCTP_STATE_MASK) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_STATE_COOKIE_WAIT:
/* this is the expected state for this chunk */
/* process the INIT-ACK parameters */
if (stcb->asoc.primary_destination->dest_state &
SCTP_ADDR_UNCONFIRMED) {
/*
* The primary is where we sent the INIT, we can
* always consider it confirmed when the INIT-ACK is
* returned. Do this before we load addresses
* though.
*/
stcb->asoc.primary_destination->dest_state &=
~SCTP_ADDR_UNCONFIRMED;
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_CONFIRMED,
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
stcb, 0, (void *)stcb->asoc.primary_destination, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
if (sctp_process_init_ack(m, iphlen, offset, sh, cp, stcb,
- Take out the broken table-id concept. Panda Routers have a M-VRF concept that is NOT well thought out for a multi-homed transport protocol. So the useless table-id entries passed around need to be removed. - Add a event timer for the zero copy api. - Fix a bug in sctp_timer.c when searching for an alternate with the largest ssthresh (the compare was wrong).
2007-06-01 11:19:54 +00:00
net, abort_no_unlock, vrf_id) < 0) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* error in parsing parameters */
return (-1);
}
/* update our state */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "moving to COOKIE-ECHOED state\n");
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(&stcb->asoc, SCTP_STATE_COOKIE_ECHOED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* reset the RTO calc */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)SCTP_GETTIME_TIMEVAL(&stcb->asoc.time_entered);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* collapse the init timer back in case of a exponential
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
* backoff
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
sctp_timer_start(SCTP_TIMER_TYPE_COOKIE, stcb->sctp_ep,
stcb, net);
/*
* the send at the end of the inbound data processing will
* cause the cookie to be sent
*/
break;
case SCTP_STATE_SHUTDOWN_SENT:
/* incorrect state... discard */
break;
case SCTP_STATE_COOKIE_ECHOED:
/* incorrect state... discard */
break;
case SCTP_STATE_OPEN:
/* incorrect state... discard */
break;
case SCTP_STATE_EMPTY:
case SCTP_STATE_INUSE:
default:
/* incorrect state... discard */
return (-1);
break;
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Leaving handle-init-ack end\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (0);
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
static struct sctp_tcb *
sctp_process_cookie_new(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_state_cookie *cookie, int cookie_len,
struct sctp_inpcb *inp, struct sctp_nets **netp,
struct sockaddr *init_src, int *notification,
int auth_skipped, uint32_t auth_offset, uint32_t auth_len,
uint32_t vrf_id, uint16_t port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* handle a state cookie for an existing association m: input packet mbuf
* chain-- assumes a pullup on IP/SCTP/COOKIE-ECHO chunk note: this is a
* "split" mbuf and the cookie signature does not exist offset: offset into
* mbuf to the cookie-echo chunk
*/
static struct sctp_tcb *
sctp_process_cookie_existing(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_state_cookie *cookie, int cookie_len,
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
struct sctp_inpcb *inp, struct sctp_tcb *stcb, struct sctp_nets **netp,
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
struct sockaddr *init_src, int *notification, sctp_assoc_t * sac_assoc_id,
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
uint32_t vrf_id, int auth_skipped, uint32_t auth_offset, uint32_t auth_len, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_association *asoc;
struct sctp_init_chunk *init_cp, init_buf;
struct sctp_init_ack_chunk *initack_cp, initack_buf;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
struct sctp_nets *net;
struct mbuf *op_err;
struct sctp_paramhdr *ph;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int chk_length;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
int init_offset, initack_offset, i;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int retval;
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
int spec_flag = 0;
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
uint32_t how_indx;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
net = *netp;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* I know that the TCB is non-NULL from the caller */
asoc = &stcb->asoc;
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
for (how_indx = 0; how_indx < sizeof(asoc->cookie_how); how_indx++) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (asoc->cookie_how[how_indx] == 0)
break;
}
if (how_indx < sizeof(asoc->cookie_how)) {
asoc->cookie_how[how_indx] = 1;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (SCTP_GET_STATE(asoc) == SCTP_STATE_SHUTDOWN_ACK_SENT) {
/* SHUTDOWN came in after sending INIT-ACK */
sctp_send_shutdown_ack(stcb, stcb->asoc.primary_destination);
op_err = sctp_get_mbuf_for_msg(sizeof(struct sctp_paramhdr),
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
0, M_DONTWAIT, 1, MT_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (op_err == NULL) {
/* FOOBAR */
return (NULL);
}
/* Set the len */
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_LEN(op_err) = sizeof(struct sctp_paramhdr);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
ph = mtod(op_err, struct sctp_paramhdr *);
ph->param_type = htons(SCTP_CAUSE_COOKIE_IN_SHUTDOWN);
ph->param_length = htons(sizeof(struct sctp_paramhdr));
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_send_operr_to(m, iphlen, op_err, cookie->peers_vtag,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, net->port);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 2;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* find and validate the INIT chunk in the cookie (peer's info) the
* INIT should start after the cookie-echo header struct (chunk
* header, state cookie header struct)
*/
init_offset = offset += sizeof(struct sctp_cookie_echo_chunk);
init_cp = (struct sctp_init_chunk *)
sctp_m_getptr(m, init_offset, sizeof(struct sctp_init_chunk),
(uint8_t *) & init_buf);
if (init_cp == NULL) {
/* could not pull a INIT chunk in cookie */
return (NULL);
}
chk_length = ntohs(init_cp->ch.chunk_length);
if (init_cp->ch.chunk_type != SCTP_INITIATION) {
return (NULL);
}
/*
* find and validate the INIT-ACK chunk in the cookie (my info) the
* INIT-ACK follows the INIT chunk
*/
initack_offset = init_offset + SCTP_SIZE32(chk_length);
initack_cp = (struct sctp_init_ack_chunk *)
sctp_m_getptr(m, initack_offset, sizeof(struct sctp_init_ack_chunk),
(uint8_t *) & initack_buf);
if (initack_cp == NULL) {
/* could not pull INIT-ACK chunk in cookie */
return (NULL);
}
chk_length = ntohs(initack_cp->ch.chunk_length);
if (initack_cp->ch.chunk_type != SCTP_INITIATION_ACK) {
return (NULL);
}
if ((ntohl(initack_cp->init.initiate_tag) == asoc->my_vtag) &&
(ntohl(init_cp->init.initiate_tag) == asoc->peer_vtag)) {
/*
* case D in Section 5.2.4 Table 2: MMAA process accordingly
* to get into the OPEN state
*/
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (ntohl(initack_cp->init.initial_tsn) != asoc->init_seq_number) {
- send call has a reference to uio->uio_resid in the recent send code, but uio may be NULL on sendfile calls. Change to use sndlen variable. - EMSGSIZE is not being returned in non-blocking mode and needs a small tweak to look if the msg would ever fit when returning EWOULDBLOCK. - FWD-TSN has a bug in stream processing which could cause a panic. This is a follow on to the codenomicon fix. - PDAPI level 1 and 2 do not work unless the reader gets his returned buffer full. Fix so we can break out when at level 1 or 2. - Fix fast-handoff features to copy across properly on accepted sockets - Fix sctp_peeloff() system call when no true system call exists to screen arguments for errors. In cases where a real system call exists the system call itself does this. - Fix raddr leak in recent add-ip code change for bundled asconfs (even when non-bundled asconfs are received) - Make sure ipi_addr lock is held when walking global addr list. Need to change this lock type to a rwlock(). - Add don't wake flag on both input and output when the socket is closing. - When deleting an address verify the interface is correct before allowing the delete to process. This protects panda and unnumbered. - Clean up old sysctl stuff and get rid of the old Open/Net BSD structures. - Add a function to watch the ranges in the sysctl sets. - When appending in the reassembly queue, validate that the assoc has not gone to about to be freed. If so (in the middle) abort out. Note this especially effects MAC I think due to the lock/unlock they do (or with LOCK testing in place). - Netstat patch to get rid of warnings. - Make sure that no data gets queued to inactive/unconfirmed destinations. This especially effect CMT but also makes a impact on regular SCTP as well. - During init collision when we detect seq number out of sync we need to treat it like Case C and discard the cookie (no invarient needed here). - Atomic access to the random store. - When we declare a vtag good, we need to shove it into the time wait hash to prevent further use. When the tag is put into the assoc hash, we need to remove it from the twait hash (where it will surely be). This prevents duplicate tag assignments. - Move decr-ref count to better protect sysctl out of data. - ltrace error corrections in sctp6_usrreq.c - Add hook for interface up/down to be sent to us. - Make sysctl() exported structures independent of processor architecture. - Fix route and src addr cache clearing for delete address case. - Make sure address marked SCTP_DEL_IP_ADDRESS is never selected as src addr. - in icmp handling fixed so we actually look at the icmp codes to figure out what to do. - Modified mobility code. Reception of DELETE IP ADDRESS for a primary destination and SET PRIMARY for a new primary destination is used for retransmission trigger to the new primary destination. Also, in this case, destination of chunks in send_queue are changed to the new primary destination. - Fix so that we disallow sending by mbuf to ever have EEOR mode set upon it. Approved by: re@freebsd.org (B Mah)
2007-09-08 17:48:46 +00:00
/*-
* Opps, this means that we somehow generated two vtag's
* the same. I.e. we did:
* Us Peer
* <---INIT(tag=a)------
* ----INIT-ACK(tag=t)-->
* ----INIT(tag=t)------> *1
* <---INIT-ACK(tag=a)---
* <----CE(tag=t)------------- *2
*
* At point *1 we should be generating a different
* tag t'. Which means we would throw away the CE and send
* ours instead. Basically this is case C (throw away side).
*/
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 17;
return (NULL);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
switch SCTP_GET_STATE
(asoc) {
case SCTP_STATE_COOKIE_WAIT:
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
case SCTP_STATE_COOKIE_ECHOED:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
* INIT was sent but got a COOKIE_ECHO with the
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
* correct tags... just accept it...but we must
* process the init so that we can make sure we have
* the right seq no's.
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
/* First we must process the INIT !! */
retval = sctp_process_init(init_cp, stcb, net);
if (retval < 0) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 3;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* we have already processed the INIT so no problem */
sctp_timer_stop(SCTP_TIMER_TYPE_HEARTBEAT, inp, stcb,
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_12);
sctp_timer_stop(SCTP_TIMER_TYPE_INIT, inp, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_13);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* update current state */
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
if (SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED)
SCTP_STAT_INCR_COUNTER32(sctps_activeestab);
else
SCTP_STAT_INCR_COUNTER32(sctps_collisionestab);
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->state & SCTP_STATE_SHUTDOWN_PENDING) {
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNGUARD,
stcb->sctp_ep, stcb, asoc->primary_destination);
}
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
SCTP_STAT_INCR_GAUGE32(sctps_currestab);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_stop_all_cookie_timers(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) ||
(stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_IN_TCPPOOL)) &&
(inp->sctp_socket->so_qlimit == 0)
) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* Here is where collision would go if we
* did a connect() and instead got a
* init/init-ack/cookie done before the
* init-ack came back..
*/
stcb->sctp_ep->sctp_flags |=
SCTP_PCB_FLAGS_CONNECTED;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_add_int(&stcb->asoc.refcnt, -1);
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
SCTP_SOCKET_UNLOCK(so, 1);
return (NULL);
}
#endif
soisconnected(stcb->sctp_socket);
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* notify upper layer */
*notification = SCTP_NOTIFY_ASSOC_UP;
/*
* since we did not send a HB make sure we don't
* double things
*/
net->hb_responded = 1;
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
net->RTO = sctp_calculate_rto(stcb, asoc, net,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
&cookie->time_entered,
sctp_align_unsafe_makecopy,
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
SCTP_RTT_FROM_NON_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb->asoc.sctp_autoclose_ticks &&
(sctp_is_feature_on(inp, SCTP_PCB_FLAGS_AUTOCLOSE))) {
sctp_timer_start(SCTP_TIMER_TYPE_AUTOCLOSE,
inp, stcb, NULL);
}
break;
default:
/*
* we're in the OPEN state (or beyond), so peer must
* have simply lost the COOKIE-ACK
*/
break;
} /* end switch */
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_stop_all_cookie_timers(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* We ignore the return code here.. not sure if we should
* somehow abort.. but we do have an existing asoc. This
* really should not fail.
*/
if (sctp_load_addresses_from_init(stcb, m, iphlen,
init_offset + sizeof(struct sctp_init_chunk),
initack_offset, sh, init_src)) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 4;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* respond with a COOKIE-ACK */
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_toss_old_cookies(stcb, asoc);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_send_cookie_ack(stcb);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 5;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (stcb);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ntohl(initack_cp->init.initiate_tag) != asoc->my_vtag &&
ntohl(init_cp->init.initiate_tag) == asoc->peer_vtag &&
cookie->tie_tag_my_vtag == 0 &&
cookie->tie_tag_peer_vtag == 0) {
/*
* case C in Section 5.2.4 Table 2: XMOO silently discard
*/
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 6;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/*
* If nat support, and the below and stcb is established, send back
* a ABORT(colliding state) if we are established.
*/
if ((SCTP_GET_STATE(asoc) == SCTP_STATE_OPEN) &&
(asoc->peer_supports_nat) &&
((ntohl(initack_cp->init.initiate_tag) == asoc->my_vtag) &&
((ntohl(init_cp->init.initiate_tag) != asoc->peer_vtag) ||
(asoc->peer_vtag == 0)))) {
/*
* Special case - Peer's support nat. We may have two init's
* that we gave out the same tag on since one was not
* established.. i.e. we get INIT from host-1 behind the nat
* and we respond tag-a, we get a INIT from host-2 behind
* the nat and we get tag-a again. Then we bring up host-1
* (or 2's) assoc, Then comes the cookie from hsot-2 (or 1).
* Now we have colliding state. We must send an abort here
* with colliding state indication.
*/
op_err = sctp_get_mbuf_for_msg(sizeof(struct sctp_paramhdr),
0, M_DONTWAIT, 1, MT_DATA);
if (op_err == NULL) {
/* FOOBAR */
return (NULL);
}
/* pre-reserve some space */
#ifdef INET6
SCTP_BUF_RESV_UF(op_err, sizeof(struct ip6_hdr));
#else
SCTP_BUF_RESV_UF(op_err, sizeof(struct ip));
#endif
SCTP_BUF_RESV_UF(op_err, sizeof(struct sctphdr));
SCTP_BUF_RESV_UF(op_err, sizeof(struct sctp_chunkhdr));
/* Set the len */
SCTP_BUF_LEN(op_err) = sizeof(struct sctp_paramhdr);
ph = mtod(op_err, struct sctp_paramhdr *);
ph->param_type = htons(SCTP_CAUSE_NAT_COLLIDING_STATE);
ph->param_length = htons(sizeof(struct sctp_paramhdr));
sctp_send_abort(m, iphlen, sh, 0, op_err, vrf_id, port);
return (NULL);
}
if ((ntohl(initack_cp->init.initiate_tag) == asoc->my_vtag) &&
((ntohl(init_cp->init.initiate_tag) != asoc->peer_vtag) ||
(asoc->peer_vtag == 0))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* case B in Section 5.2.4 Table 2: MXAA or MOAA my info
* should be ok, re-accept peer info
*/
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (ntohl(initack_cp->init.initial_tsn) != asoc->init_seq_number) {
/*
* Extension of case C. If we hit this, then the
* random number generator returned the same vtag
* when we first sent our INIT-ACK and when we later
* sent our INIT. The side with the seq numbers that
* are different will be the one that normnally
* would have hit case C. This in effect "extends"
* our vtags in this collision case to be 64 bits.
* The same collision could occur aka you get both
* vtag and seq number the same twice in a row.. but
* is much less likely. If it did happen then we
* would proceed through and bring up the assoc.. we
* may end up with the wrong stream setup however..
* which would be bad.. but there is no way to
* tell.. until we send on a stream that does not
* exist :-)
*/
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 7;
return (NULL);
}
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 8;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_HEARTBEAT, inp, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_14);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_stop_all_cookie_timers(stcb);
/*
* since we did not send a HB make sure we don't double
* things
*/
net->hb_responded = 1;
if (stcb->asoc.sctp_autoclose_ticks &&
sctp_is_feature_on(inp, SCTP_PCB_FLAGS_AUTOCLOSE)) {
sctp_timer_start(SCTP_TIMER_TYPE_AUTOCLOSE, inp, stcb,
NULL);
}
asoc->my_rwnd = ntohl(initack_cp->init.a_rwnd);
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
asoc->pre_open_streams = ntohs(initack_cp->init.num_outbound_streams);
if (ntohl(init_cp->init.initiate_tag) != asoc->peer_vtag) {
/*
* Ok the peer probably discarded our data (if we
* echoed a cookie+data). So anything on the
* sent_queue should be marked for retransmit, we
* may not get something to kick us so it COULD
* still take a timeout to move these.. but it can't
* hurt to mark them.
*/
struct sctp_tmit_chunk *chk;
TAILQ_FOREACH(chk, &stcb->asoc.sent_queue, sctp_next) {
if (chk->sent < SCTP_DATAGRAM_RESEND) {
chk->sent = SCTP_DATAGRAM_RESEND;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_flight_size_decrease(chk);
sctp_total_flight_decrease(stcb, chk);
- Found bug in min split point bundling which caused incorrect, non-bundlable fragmentation. - Added min residual to better control split points for both how big a msg must be as well as how much needs to be left over. - With our new algo in place, we need to implicitly set "end of msg" on the sp-> structure otherwise we end up with "hung" associations. - Room reserved up front in IP header by pushing IP header to back of mbuf. - Fix so FR's peg count of retransmissions needed. - Fix so an unlucky chunk that never gets across will kill the assoc via the kill timer and send an abort too. - Fix bug in sctp_input which can result in a crash. - Do not strip off IP options anymore. - Clean up sctp_calculate_rto(). - Get rid of unused sysctl. - Fixed so we discard all M-Cast - Fixed so port check done AFTER checksum - Fixed bug in fragmentation code that prevented us from fragmenting a small complete message when we needed to. - Window probes were not marked back to unsent and flight adjusted when a sack came in with no window change or accepting of the probe data. We now fix this with having a mark on the net and the chunk so we can clear it out when the sack arrives forcing it to retran just like it was "new" this improves the handling of window probes, which were dropped by the receiver. - Tighten AUTH protocol error checks during INIT/INIT-ACK exchange
2007-03-31 11:47:30 +00:00
sctp_ucount_incr(stcb->asoc.sent_queue_retran_cnt);
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
spec_flag++;
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* process the INIT info (peer's info) */
retval = sctp_process_init(init_cp, stcb, net);
if (retval < 0) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 9;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
if (sctp_load_addresses_from_init(stcb, m, iphlen,
init_offset + sizeof(struct sctp_init_chunk),
initack_offset, sh, init_src)) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 10;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
if ((asoc->state & SCTP_STATE_COOKIE_WAIT) ||
(asoc->state & SCTP_STATE_COOKIE_ECHOED)) {
*notification = SCTP_NOTIFY_ASSOC_UP;
if (((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) ||
(stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_IN_TCPPOOL)) &&
(inp->sctp_socket->so_qlimit == 0)) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->sctp_ep->sctp_flags |=
SCTP_PCB_FLAGS_CONNECTED;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_add_int(&stcb->asoc.refcnt, -1);
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
SCTP_SOCKET_UNLOCK(so, 1);
return (NULL);
}
#endif
soisconnected(stcb->sctp_socket);
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
if (SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED)
SCTP_STAT_INCR_COUNTER32(sctps_activeestab);
else
SCTP_STAT_INCR_COUNTER32(sctps_collisionestab);
SCTP_STAT_INCR_GAUGE32(sctps_currestab);
} else if (SCTP_GET_STATE(asoc) == SCTP_STATE_OPEN) {
SCTP_STAT_INCR_COUNTER32(sctps_restartestab);
} else {
SCTP_STAT_INCR_COUNTER32(sctps_collisionestab);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->state & SCTP_STATE_SHUTDOWN_PENDING) {
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNGUARD,
stcb->sctp_ep, stcb, asoc->primary_destination);
}
sctp_stop_all_cookie_timers(stcb);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_toss_old_cookies(stcb, asoc);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_send_cookie_ack(stcb);
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
if (spec_flag) {
/*
* only if we have retrans set do we do this. What
* this call does is get only the COOKIE-ACK out and
* then when we return the normal call to
* sctp_chunk_output will get the retrans out behind
* this.
*/
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_chunk_output(inp, stcb, SCTP_OUTPUT_FROM_COOKIE_ACK, SCTP_SO_NOT_LOCKED);
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
}
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 11;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (stcb);
}
if ((ntohl(initack_cp->init.initiate_tag) != asoc->my_vtag &&
ntohl(init_cp->init.initiate_tag) != asoc->peer_vtag) &&
cookie->tie_tag_my_vtag == asoc->my_vtag_nonce &&
cookie->tie_tag_peer_vtag == asoc->peer_vtag_nonce &&
cookie->tie_tag_peer_vtag != 0) {
struct sctpasochead *head;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
if (asoc->peer_supports_nat) {
/*
* This is a gross gross hack. just call the
* cookie_new code since we are allowing a duplicate
* association. I hope this works...
*/
return (sctp_process_cookie_new(m, iphlen, offset, sh, cookie, cookie_len,
inp, netp, init_src, notification,
auth_skipped, auth_offset, auth_len,
vrf_id, port));
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* case A in Section 5.2.4 Table 2: XXMM (peer restarted)
*/
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
/* temp code */
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 12;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_INIT, inp, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_15);
sctp_timer_stop(SCTP_TIMER_TYPE_HEARTBEAT, inp, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_16);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*sac_assoc_id = sctp_get_associd(stcb);
/* notify upper layer */
*notification = SCTP_NOTIFY_ASSOC_RESTART;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
if ((SCTP_GET_STATE(asoc) != SCTP_STATE_OPEN) &&
(SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_RECEIVED) &&
(SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_SENT)) {
SCTP_STAT_INCR_GAUGE32(sctps_currestab);
}
if (SCTP_GET_STATE(asoc) == SCTP_STATE_OPEN) {
SCTP_STAT_INCR_GAUGE32(sctps_restartestab);
} else if (SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_SENT) {
SCTP_STAT_INCR_GAUGE32(sctps_collisionestab);
}
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
if (asoc->state & SCTP_STATE_SHUTDOWN_PENDING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNGUARD,
stcb->sctp_ep, stcb, asoc->primary_destination);
} else if (!(asoc->state & SCTP_STATE_SHUTDOWN_SENT)) {
/* move to OPEN state, if not in SHUTDOWN_SENT */
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
}
asoc->pre_open_streams =
ntohs(initack_cp->init.num_outbound_streams);
asoc->init_seq_number = ntohl(initack_cp->init.initial_tsn);
asoc->sending_seq = asoc->asconf_seq_out = asoc->str_reset_seq_out = asoc->init_seq_number;
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
asoc->asconf_seq_out_acked = asoc->asconf_seq_out - 1;
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
asoc->asconf_seq_in = asoc->last_acked_seq = asoc->init_seq_number - 1;
asoc->str_reset_seq_in = asoc->init_seq_number;
asoc->advanced_peer_ack_point = asoc->last_acked_seq;
- Fix a memory overwrite when the mapping array is expanded, size of expansion was not taken int consideration. - Fix so vtag hash is 1 bigger so that it modulo's out correctly, avoids a panic when restart with right modulo happens. - do not dereference stcb when control->do_not_ref_stcb is set - Fix up packet logging to not often use a lock and also to add to options. - Fix some logging option duplication in the sctputil.h
2007-05-30 17:39:45 +00:00
if (asoc->mapping_array) {
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
memset(asoc->mapping_array, 0,
asoc->mapping_array_size);
- Fix a memory overwrite when the mapping array is expanded, size of expansion was not taken int consideration. - Fix so vtag hash is 1 bigger so that it modulo's out correctly, avoids a panic when restart with right modulo happens. - do not dereference stcb when control->do_not_ref_stcb is set - Fix up packet logging to not often use a lock and also to add to options. - Fix some logging option duplication in the sctputil.h
2007-05-30 17:39:45 +00:00
}
Fix for NR-Sack code. The code was NOT working properly when enabled. Basically most of the operations were incorrect causing bad sacks when you enabled nr-sack. The fixes range across 4 files and unifiy most of the processing so that we only test nr_sack flags to decide which type of sack to generate. Optimization left for this is to combine the sack generation code and make it capable of generating either sack thus shrinking out a routine. Reviewed by: tuexen@freebsd.org
2010-03-24 19:45:36 +00:00
if (asoc->nr_mapping_array) {
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
memset(asoc->nr_mapping_array, 0,
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
asoc->mapping_array_size);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
SCTP_TCB_UNLOCK(stcb);
SCTP_INP_INFO_WLOCK();
SCTP_INP_WLOCK(stcb->sctp_ep);
SCTP_TCB_LOCK(stcb);
atomic_add_int(&stcb->asoc.refcnt, -1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* send up all the data */
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
SCTP_TCB_SEND_LOCK(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_report_all_outbound(stcb, 1, SCTP_SO_NOT_LOCKED);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
for (i = 0; i < stcb->asoc.streamoutcnt; i++) {
stcb->asoc.strmout[i].stream_no = i;
stcb->asoc.strmout[i].next_sequence_sent = 0;
stcb->asoc.strmout[i].last_msg_incomplete = 0;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* process the INIT-ACK info (my info) */
asoc->my_vtag = ntohl(initack_cp->init.initiate_tag);
asoc->my_rwnd = ntohl(initack_cp->init.a_rwnd);
/* pull from vtag hash */
LIST_REMOVE(stcb, sctp_asocs);
/* re-insert to new vtag position */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
head = &SCTP_BASE_INFO(sctp_asochash)[SCTP_PCBHASH_ASOC(stcb->asoc.my_vtag,
SCTP_BASE_INFO(hashasocmark))];
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* put it in the bucket in the vtag hash of assoc's for the
* system
*/
LIST_INSERT_HEAD(head, stcb, sctp_asocs);
/* process the INIT info (peer's info) */
In a true restart case, the send_lock was not being aquired. This meant that when we cleanup the outbound we may have one in transit to be added with the old sequence number. This is bad since then we loose a message :( Also the report_outbound needed to have the right lock when its called which it did not.. I added the lock with of course a flag since we want to have the lock before we call it in the restart case. This also fixed the FIX ME case where, in the cookie collision case, we mark for retransmit any that were bundled with the cookie that was dropped. This also means changes to the output routine so we can assure getting the COOKIE-ACK sent BEFORE we retransmit the Data. Approved by: gnn
2006-11-11 22:44:12 +00:00
SCTP_TCB_SEND_UNLOCK(stcb);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
SCTP_INP_WUNLOCK(stcb->sctp_ep);
SCTP_INP_INFO_WUNLOCK();
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
retval = sctp_process_init(init_cp, stcb, net);
if (retval < 0) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 13;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* since we did not send a HB make sure we don't double
* things
*/
net->hb_responded = 1;
if (sctp_load_addresses_from_init(stcb, m, iphlen,
init_offset + sizeof(struct sctp_init_chunk),
initack_offset, sh, init_src)) {
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 14;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* respond with a COOKIE-ACK */
sctp_stop_all_cookie_timers(stcb);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_toss_old_cookies(stcb, asoc);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_send_cookie_ack(stcb);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 15;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (stcb);
}
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
if (how_indx < sizeof(asoc->cookie_how))
asoc->cookie_how[how_indx] = 16;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* all other cases... */
return (NULL);
}
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* handle a state cookie for a new association m: input packet mbuf chain--
* assumes a pullup on IP/SCTP/COOKIE-ECHO chunk note: this is a "split" mbuf
* and the cookie signature does not exist offset: offset into mbuf to the
* cookie-echo chunk length: length of the cookie chunk to: where the init
* was from returns a new TCB
*/
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
struct sctp_tcb *
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_process_cookie_new(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_state_cookie *cookie, int cookie_len,
struct sctp_inpcb *inp, struct sctp_nets **netp,
struct sockaddr *init_src, int *notification,
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
int auth_skipped, uint32_t auth_offset, uint32_t auth_len,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
uint32_t vrf_id, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_tcb *stcb;
struct sctp_init_chunk *init_cp, init_buf;
struct sctp_init_ack_chunk *initack_cp, initack_buf;
struct sockaddr_storage sa_store;
struct sockaddr *initack_src = (struct sockaddr *)&sa_store;
struct sctp_association *asoc;
int chk_length;
int init_offset, initack_offset, initack_limit;
int retval;
int error = 0;
uint32_t old_tag;
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
uint8_t auth_chunk_buf[SCTP_PARAM_BUFFER_SIZE];
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#ifdef INET
struct sockaddr_in *sin;
#endif
#ifdef INET6
struct sockaddr_in6 *sin6;
#endif
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
so = SCTP_INP_SO(inp);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* find and validate the INIT chunk in the cookie (peer's info) the
* INIT should start after the cookie-echo header struct (chunk
* header, state cookie header struct)
*/
init_offset = offset + sizeof(struct sctp_cookie_echo_chunk);
init_cp = (struct sctp_init_chunk *)
sctp_m_getptr(m, init_offset, sizeof(struct sctp_init_chunk),
(uint8_t *) & init_buf);
if (init_cp == NULL) {
/* could not pull a INIT chunk in cookie */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"process_cookie_new: could not pull INIT chunk hdr\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
chk_length = ntohs(init_cp->ch.chunk_length);
if (init_cp->ch.chunk_type != SCTP_INITIATION) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "HUH? process_cookie_new: could not find INIT chunk!\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
initack_offset = init_offset + SCTP_SIZE32(chk_length);
/*
* find and validate the INIT-ACK chunk in the cookie (my info) the
* INIT-ACK follows the INIT chunk
*/
initack_cp = (struct sctp_init_ack_chunk *)
sctp_m_getptr(m, initack_offset, sizeof(struct sctp_init_ack_chunk),
(uint8_t *) & initack_buf);
if (initack_cp == NULL) {
/* could not pull INIT-ACK chunk in cookie */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "process_cookie_new: could not pull INIT-ACK chunk hdr\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
chk_length = ntohs(initack_cp->ch.chunk_length);
if (initack_cp->ch.chunk_type != SCTP_INITIATION_ACK) {
return (NULL);
}
/*
* NOTE: We can't use the INIT_ACK's chk_length to determine the
* "initack_limit" value. This is because the chk_length field
* includes the length of the cookie, but the cookie is omitted when
* the INIT and INIT_ACK are tacked onto the cookie...
*/
initack_limit = offset + cookie_len;
/*
* now that we know the INIT/INIT-ACK are in place, create a new TCB
* and popluate
*/
- remove duplicate code from sctp_asconf.c - remove duplicate #include <sys/priv.h> that is not under #ifdef FreeBSD version to allow compile on 6.1 - static analysis changes per the cisco SA tool including: o some SA_IGNORE comments o some checks for NULL before unlock. o type corrections int -> size_t - Fix it so sctp_alloc_asoc takes a thread/proc argument. Without this we pass a NULL in to bind on implicit assoc setup and crash :-( Approved by: re@freebsd.org(Ken Smith)
2007-07-21 21:41:32 +00:00
/*
* Here we do a trick, we set in NULL for the proc/thread argument.
* We do this since in effect we only use the p argument when the
* socket is unbound and we must do an implicit bind. Since we are
* getting a cookie, we cannot be unbound.
*/
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
stcb = sctp_aloc_assoc(inp, init_src, &error,
- remove duplicate code from sctp_asconf.c - remove duplicate #include <sys/priv.h> that is not under #ifdef FreeBSD version to allow compile on 6.1 - static analysis changes per the cisco SA tool including: o some SA_IGNORE comments o some checks for NULL before unlock. o type corrections int -> size_t - Fix it so sctp_alloc_asoc takes a thread/proc argument. Without this we pass a NULL in to bind on implicit assoc setup and crash :-( Approved by: re@freebsd.org(Ken Smith)
2007-07-21 21:41:32 +00:00
ntohl(initack_cp->init.initiate_tag), vrf_id,
(struct thread *)NULL
);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL) {
struct mbuf *op_err;
/* memory problem? */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"process_cookie_new: no room for another TCB!\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
op_err = sctp_generate_invmanparam(SCTP_CAUSE_OUT_OF_RESC);
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_abort_association(inp, (struct sctp_tcb *)NULL, m, iphlen,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sh, op_err, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* get the correct sctp_nets */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (netp)
*netp = sctp_findnet(stcb, init_src);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc = &stcb->asoc;
/* get scope variables out of cookie */
asoc->ipv4_local_scope = cookie->ipv4_scope;
asoc->site_scope = cookie->site_scope;
asoc->local_scope = cookie->local_scope;
asoc->loopback_scope = cookie->loopback_scope;
if ((asoc->ipv4_addr_legal != cookie->ipv4_addr_legal) ||
(asoc->ipv6_addr_legal != cookie->ipv6_addr_legal)) {
struct mbuf *op_err;
/*
* Houston we have a problem. The EP changed while the
* cookie was in flight. Only recourse is to abort the
* association.
*/
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
op_err = sctp_generate_invmanparam(SCTP_CAUSE_OUT_OF_RESC);
sctp_abort_association(inp, (struct sctp_tcb *)NULL, m, iphlen,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sh, op_err, vrf_id, port);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC,
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
SCTP_FROM_SCTP_INPUT + SCTP_LOC_16);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
atomic_subtract_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* process the INIT-ACK info (my info) */
old_tag = asoc->my_vtag;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
asoc->my_vtag = ntohl(initack_cp->init.initiate_tag);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->my_rwnd = ntohl(initack_cp->init.a_rwnd);
asoc->pre_open_streams = ntohs(initack_cp->init.num_outbound_streams);
asoc->init_seq_number = ntohl(initack_cp->init.initial_tsn);
asoc->sending_seq = asoc->asconf_seq_out = asoc->str_reset_seq_out = asoc->init_seq_number;
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
asoc->asconf_seq_out_acked = asoc->asconf_seq_out - 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->asconf_seq_in = asoc->last_acked_seq = asoc->init_seq_number - 1;
asoc->str_reset_seq_in = asoc->init_seq_number;
asoc->advanced_peer_ack_point = asoc->last_acked_seq;
/* process the INIT info (peer's info) */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (netp)
retval = sctp_process_init(init_cp, stcb, *netp);
else
retval = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (retval < 0) {
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_16);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
atomic_subtract_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* load all addresses */
if (sctp_load_addresses_from_init(stcb, m, iphlen,
init_offset + sizeof(struct sctp_init_chunk), initack_offset, sh,
init_src)) {
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_17);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
atomic_subtract_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* verify any preceding AUTH chunk that was skipped
*/
/* pull the local authentication parameters from the cookie/init-ack */
sctp_auth_get_cookie_params(stcb, m,
initack_offset + sizeof(struct sctp_init_ack_chunk),
initack_limit - (initack_offset + sizeof(struct sctp_init_ack_chunk)));
if (auth_skipped) {
struct sctp_auth_chunk *auth;
auth = (struct sctp_auth_chunk *)
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
sctp_m_getptr(m, auth_offset, auth_len, auth_chunk_buf);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((auth == NULL) || sctp_handle_auth(stcb, auth, m, auth_offset)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* auth HMAC failed, dump the assoc and packet */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_AUTH1,
"COOKIE-ECHO: AUTH failed\n");
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_18);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
atomic_subtract_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
} else {
/* remaining chunks checked... good to go */
stcb->asoc.authenticated = 1;
}
}
/* update current state */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "moving to OPEN state\n");
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->state & SCTP_STATE_SHUTDOWN_PENDING) {
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNGUARD,
stcb->sctp_ep, stcb, asoc->primary_destination);
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
sctp_stop_all_cookie_timers(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR_COUNTER32(sctps_passiveestab);
SCTP_STAT_INCR_GAUGE32(sctps_currestab);
/*
* if we're doing ASCONFs, check to see if we have any new local
* addresses that need to get added to the peer (eg. addresses
* changed while cookie echo in flight). This needs to be done
* after we go to the OPEN state to do the correct asconf
* processing. else, make sure we have the correct addresses in our
* lists
*/
/* warning, we re-use sin, sin6, sa_store here! */
/* pull in local_address (our "from" address) */
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
switch (cookie->laddr_type) {
#ifdef INET
case SCTP_IPV4_ADDRESS:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* source addr is IPv4 */
sin = (struct sockaddr_in *)initack_src;
memset(sin, 0, sizeof(*sin));
sin->sin_family = AF_INET;
sin->sin_len = sizeof(struct sockaddr_in);
sin->sin_addr.s_addr = cookie->laddress[0];
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
break;
#endif
#ifdef INET6
case SCTP_IPV6_ADDRESS:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* source addr is IPv6 */
sin6 = (struct sockaddr_in6 *)initack_src;
memset(sin6, 0, sizeof(*sin6));
sin6->sin6_family = AF_INET6;
sin6->sin6_len = sizeof(struct sockaddr_in6);
sin6->sin6_scope_id = cookie->scope_id;
memcpy(&sin6->sin6_addr, cookie->laddress,
sizeof(sin6->sin6_addr));
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
break;
#endif
default:
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
atomic_add_int(&stcb->asoc.refcnt, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_19);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
atomic_subtract_int(&stcb->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* set up to notify upper layer */
*notification = SCTP_NOTIFY_ASSOC_UP;
if (((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) ||
(stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_IN_TCPPOOL)) &&
(inp->sctp_socket->so_qlimit == 0)) {
/*
* This is an endpoint that called connect() how it got a
* cookie that is NEW is a bit of a mystery. It must be that
* the INIT was sent, but before it got there.. a complete
* INIT/INIT-ACK/COOKIE arrived. But of course then it
* should have went to the other code.. not here.. oh well..
* a bit of protection is worth having..
*/
stcb->sctp_ep->sctp_flags |= SCTP_PCB_FLAGS_CONNECTED;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
SCTP_SOCKET_UNLOCK(so, 1);
return (NULL);
}
#endif
soisconnected(stcb->sctp_socket);
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else if ((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) &&
(inp->sctp_socket->so_qlimit)) {
/*
* We don't want to do anything with this one. Since it is
* the listening guy. The timer will get started for
* accepted connections in the caller.
*/
;
}
/* since we did not send a HB make sure we don't double things */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((netp) && (*netp))
(*netp)->hb_responded = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb->asoc.sctp_autoclose_ticks &&
sctp_is_feature_on(inp, SCTP_PCB_FLAGS_AUTOCLOSE)) {
sctp_timer_start(SCTP_TIMER_TYPE_AUTOCLOSE, inp, stcb, NULL);
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
/* calculate the RTT */
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
(void)SCTP_GETTIME_TIMEVAL(&stcb->asoc.time_entered);
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
if ((netp) && (*netp)) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
(*netp)->RTO = sctp_calculate_rto(stcb, asoc, *netp,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
&cookie->time_entered, sctp_align_unsafe_makecopy,
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
SCTP_RTT_FROM_NON_DATA);
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
}
- Get rid of unsused constants for sysctl variables. - Fix panic from mutex unlock on freed lock when ASCONF-ACK aborts an assoc - Fix panic from addr lock recursion when ASCONFs are queued in the front states - ASCONFs "queued" in the front states should really be bundled after the COOKIE-ACK, not in front of it - Fix issue with addresses deleted in the front states from being sent with ASCONF(DELETE)-- replaced sctp_asconf_queue_add_sa() with delete specific function - Comment change in sctp.h the drafts are now RFC's Approved by: re@freebsd.org (B Mah)
2007-09-15 19:07:42 +00:00
/* respond with a COOKIE-ACK */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_send_cookie_ack(stcb);
- Get rid of unsused constants for sysctl variables. - Fix panic from mutex unlock on freed lock when ASCONF-ACK aborts an assoc - Fix panic from addr lock recursion when ASCONFs are queued in the front states - ASCONFs "queued" in the front states should really be bundled after the COOKIE-ACK, not in front of it - Fix issue with addresses deleted in the front states from being sent with ASCONF(DELETE)-- replaced sctp_asconf_queue_add_sa() with delete specific function - Comment change in sctp.h the drafts are now RFC's Approved by: re@freebsd.org (B Mah)
2007-09-15 19:07:42 +00:00
/*
* check the address lists for any ASCONFs that need to be sent
* AFTER the cookie-ack is sent
*/
sctp_check_address_list(stcb, m,
initack_offset + sizeof(struct sctp_init_ack_chunk),
initack_limit - (initack_offset + sizeof(struct sctp_init_ack_chunk)),
initack_src, cookie->local_scope, cookie->site_scope,
cookie->ipv4_scope, cookie->loopback_scope);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (stcb);
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/*
* CODE LIKE THIS NEEDS TO RUN IF the peer supports the NAT extension, i.e
* we NEED to make sure we are not already using the vtag. If so we
* need to send back an ABORT-TRY-AGAIN-WITH-NEW-TAG No middle box bit!
head = &SCTP_BASE_INFO(sctp_asochash)[SCTP_PCBHASH_ASOC(tag,
SCTP_BASE_INFO(hashasocmark))];
LIST_FOREACH(stcb, head, sctp_asocs) {
if ((stcb->asoc.my_vtag == tag) && (stcb->rport == rport) && (inp == stcb->sctp_ep)) {
-- SEND ABORT - TRY AGAIN --
}
}
*/
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* handles a COOKIE-ECHO message stcb: modified to either a new or left as
* existing (non-NULL) TCB
*/
static struct mbuf *
sctp_handle_cookie_echo(struct mbuf *m, int iphlen, int offset,
struct sctphdr *sh, struct sctp_cookie_echo_chunk *cp,
struct sctp_inpcb **inp_p, struct sctp_tcb **stcb, struct sctp_nets **netp,
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
int auth_skipped, uint32_t auth_offset, uint32_t auth_len,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
struct sctp_tcb **locked_tcb, uint32_t vrf_id, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_state_cookie *cookie;
struct sctp_tcb *l_stcb = *stcb;
struct sctp_inpcb *l_inp;
struct sockaddr *to;
sctp_assoc_t sac_restart_id;
struct sctp_pcb *ep;
struct mbuf *m_sig;
uint8_t calc_sig[SCTP_SIGNATURE_SIZE], tmp_sig[SCTP_SIGNATURE_SIZE];
uint8_t *sig;
uint8_t cookie_ok = 0;
unsigned int size_of_pkt, sig_offset, cookie_offset;
unsigned int cookie_len;
struct timeval now;
struct timeval time_expires;
struct sockaddr_storage dest_store;
struct sockaddr *localep_sa = (struct sockaddr *)&dest_store;
struct ip *iph;
int notification = 0;
struct sctp_nets *netl;
int had_a_existing_tcb = 0;
Bugfix: Make sure that the COMM_UP notificatin is delivered first also on the passive side. MFC after: 3 days.
2011-01-02 10:27:27 +00:00
int send_int_conf = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#ifdef INET
struct sockaddr_in sin;
#endif
#ifdef INET6
struct sockaddr_in6 sin6;
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_cookie: handling COOKIE-ECHO\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (inp_p == NULL) {
return (NULL);
}
/* First get the destination address setup too. */
iph = mtod(m, struct ip *);
Allow SCTP to compile without INET6. PR: 116816 Obtained from tuexen@fh-muenster.de: MFC after: 2 weeks
2008-04-16 17:24:18 +00:00
switch (iph->ip_v) {
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#ifdef INET
Allow SCTP to compile without INET6. PR: 116816 Obtained from tuexen@fh-muenster.de: MFC after: 2 weeks
2008-04-16 17:24:18 +00:00
case IPVERSION:
{
/* its IPv4 */
struct sockaddr_in *lsin;
lsin = (struct sockaddr_in *)(localep_sa);
memset(lsin, 0, sizeof(*lsin));
lsin->sin_family = AF_INET;
lsin->sin_len = sizeof(*lsin);
lsin->sin_port = sh->dest_port;
lsin->sin_addr.s_addr = iph->ip_dst.s_addr;
size_of_pkt = SCTP_GET_IPV4_LENGTH(iph);
break;
}
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#endif
Allow SCTP to compile without INET6. PR: 116816 Obtained from tuexen@fh-muenster.de: MFC after: 2 weeks
2008-04-16 17:24:18 +00:00
#ifdef INET6
case IPV6_VERSION >> 4:
{
/* its IPv6 */
struct ip6_hdr *ip6;
struct sockaddr_in6 *lsin6;
lsin6 = (struct sockaddr_in6 *)(localep_sa);
memset(lsin6, 0, sizeof(*lsin6));
lsin6->sin6_family = AF_INET6;
lsin6->sin6_len = sizeof(struct sockaddr_in6);
ip6 = mtod(m, struct ip6_hdr *);
lsin6->sin6_port = sh->dest_port;
lsin6->sin6_addr = ip6->ip6_dst;
size_of_pkt = SCTP_GET_IPV6_LENGTH(ip6) + iphlen;
break;
}
#endif
default:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
cookie = &cp->cookie;
cookie_offset = offset + sizeof(struct sctp_chunkhdr);
cookie_len = ntohs(cp->ch.chunk_length);
if ((cookie->peerport != sh->src_port) &&
(cookie->myport != sh->dest_port) &&
(cookie->my_vtag != sh->v_tag)) {
/*
* invalid ports or bad tag. Note that we always leave the
* v_tag in the header in network order and when we stored
* it in the my_vtag slot we also left it in network order.
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
* This maintains the match even though it may be in the
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
* opposite byte order of the machine :->
*/
return (NULL);
}
if (cookie_len > size_of_pkt ||
cookie_len < sizeof(struct sctp_cookie_echo_chunk) +
sizeof(struct sctp_init_chunk) +
sizeof(struct sctp_init_ack_chunk) + SCTP_SIGNATURE_SIZE) {
/* cookie too long! or too small */
return (NULL);
}
/*
* split off the signature into its own mbuf (since it should not be
* calculated in the sctp_hmac_m() call).
*/
sig_offset = offset + cookie_len - SCTP_SIGNATURE_SIZE;
if (sig_offset > size_of_pkt) {
/* packet not correct size! */
/* XXX this may already be accounted for earlier... */
return (NULL);
}
m_sig = m_split(m, sig_offset, M_DONTWAIT);
if (m_sig == NULL) {
/* out of memory or ?? */
return (NULL);
}
Use the pru_flush infrastructure to avoid a panic PR: 122710 MFC after: 1 week
2008-04-14 18:13:33 +00:00
#ifdef SCTP_MBUF_LOGGING
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MBUF_LOGGING_ENABLE) {
Use the pru_flush infrastructure to avoid a panic PR: 122710 MFC after: 1 week
2008-04-14 18:13:33 +00:00
struct mbuf *mat;
mat = m_sig;
while (mat) {
if (SCTP_BUF_IS_EXTENDED(mat)) {
sctp_log_mb(mat, SCTP_MBUF_SPLIT);
}
mat = SCTP_BUF_NEXT(mat);
}
}
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* compute the signature/digest for the cookie
*/
ep = &(*inp_p)->sctp_ep;
l_inp = *inp_p;
if (l_stcb) {
SCTP_TCB_UNLOCK(l_stcb);
}
SCTP_INP_RLOCK(l_inp);
if (l_stcb) {
SCTP_TCB_LOCK(l_stcb);
}
/* which cookie is it? */
if ((cookie->time_entered.tv_sec < (long)ep->time_of_secret_change) &&
(ep->current_secret_number != ep->last_secret_number)) {
/* it's the old cookie */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)sctp_hmac_m(SCTP_HMAC,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(uint8_t *) ep->secret_key[(int)ep->last_secret_number],
- fix send_failed notification contents - Reorder send failed to be in correct order. - Fixed calulation of init-ack to be right off mbuf lengths instead of the precalculated value. This will fix one 64 bit platform issue.
2007-06-09 13:46:57 +00:00
SCTP_SECRET_SIZE, m, cookie_offset, calc_sig, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else {
/* it's the current cookie */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)sctp_hmac_m(SCTP_HMAC,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(uint8_t *) ep->secret_key[(int)ep->current_secret_number],
- fix send_failed notification contents - Reorder send failed to be in correct order. - Fixed calulation of init-ack to be right off mbuf lengths instead of the precalculated value. This will fix one 64 bit platform issue.
2007-06-09 13:46:57 +00:00
SCTP_SECRET_SIZE, m, cookie_offset, calc_sig, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* get the signature */
SCTP_INP_RUNLOCK(l_inp);
sig = (uint8_t *) sctp_m_getptr(m_sig, 0, SCTP_SIGNATURE_SIZE, (uint8_t *) & tmp_sig);
if (sig == NULL) {
/* couldn't find signature */
-Fixes first of all the getcred on IPv6 and V4. The copy's were incorrect and so was the locking. -A bug was also found that would create a race and panic when an abort arrived on a socket being read from. -Also fix the reader to get MSG_TRUNC when a partial delivery is aborted. -Also addresses a couple of coverity caught error path memory leaks and a couple of other valid complaints Approved by: gnn
2006-11-08 00:21:13 +00:00
sctp_m_freem(m_sig);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/* compare the received digest with the computed digest */
if (memcmp(calc_sig, sig, SCTP_SIGNATURE_SIZE) != 0) {
/* try the old cookie? */
if ((cookie->time_entered.tv_sec == (long)ep->time_of_secret_change) &&
(ep->current_secret_number != ep->last_secret_number)) {
/* compute digest with old */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)sctp_hmac_m(SCTP_HMAC,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(uint8_t *) ep->secret_key[(int)ep->last_secret_number],
- fix send_failed notification contents - Reorder send failed to be in correct order. - Fixed calulation of init-ack to be right off mbuf lengths instead of the precalculated value. This will fix one 64 bit platform issue.
2007-06-09 13:46:57 +00:00
SCTP_SECRET_SIZE, m, cookie_offset, calc_sig, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* compare */
if (memcmp(calc_sig, sig, SCTP_SIGNATURE_SIZE) == 0)
cookie_ok = 1;
}
} else {
cookie_ok = 1;
}
/*
* Now before we continue we must reconstruct our mbuf so that
* normal processing of any other chunks will work.
*/
{
struct mbuf *m_at;
m_at = m;
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
while (SCTP_BUF_NEXT(m_at) != NULL) {
m_at = SCTP_BUF_NEXT(m_at);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_NEXT(m_at) = m_sig;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (cookie_ok == 0) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "handle_cookie_echo: cookie signature validation failed!\n");
SCTPDBG(SCTP_DEBUG_INPUT2,
"offset = %u, cookie_offset = %u, sig_offset = %u\n",
(uint32_t) offset, cookie_offset, sig_offset);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* check the cookie timestamps to be sure it's not stale
*/
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)SCTP_GETTIME_TIMEVAL(&now);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Expire time is in Ticks, so we convert to seconds */
- Fixed 1-2-1 model to not worry about associd in sockopts - Fixed RTOinfo for bounding. - Fixed connect() to return ECONNREFUSED when an ABORT is received. - Added comments to direct Static Analysis not to look at some things it does not understand (comments are /* sa_ignore XXXXX */) - Bind when colliding was broken, missing not_found = 1 before checking to see if the port was in use caused endless bind loop. - Cookie life needs to be in milliseconds to conform to socket api. - Cookie life is not supposed to change if its 0, On the assoc level set we changed it to 0 opps. - Two more static analysis issues identified by the cisco tool. Null checks needed. - An issue for sendfile(). Need to validate the correct input argument. - When sending failed due to a no route to host, we leaked the mbuf chain failing to call m_freem(). - Fix #ifdef issue for getting hash block len when HAVE_SHA2 is NOT defined Reviewed by: gnn
2007-05-17 12:16:24 +00:00
time_expires.tv_sec = cookie->time_entered.tv_sec + TICKS_TO_SEC(cookie->cookie_life);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
time_expires.tv_usec = cookie->time_entered.tv_usec;
1) Adds the rest of the VIMAGE change macros 2) Adds some __UserSpace__ on some of the common defines that the user space code needs 3) Fixes a bug when we send up data to a user that failed. We need to a) trim off the data chunk headers, if present, and b) make sure the frag bit is communicated properly for the msgs coming off the stream queues... i.e. we see if some of the msg has been taken. Obtained from: jeli contributed the VIMAGE changes on this pass Thanks Julain!
2008-07-09 16:45:30 +00:00
/*
* TODO sctp_constants.h needs alternative time macros when _KERNEL
* is undefined.
*/
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (timevalcmp(&now, &time_expires, >)) {
/* cookie is stale! */
struct mbuf *op_err;
struct sctp_stale_cookie_msg *scm;
uint32_t tim;
op_err = sctp_get_mbuf_for_msg(sizeof(struct sctp_stale_cookie_msg),
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
0, M_DONTWAIT, 1, MT_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (op_err == NULL) {
/* FOOBAR */
return (NULL);
}
/* Set the len */
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_LEN(op_err) = sizeof(struct sctp_stale_cookie_msg);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
scm = mtod(op_err, struct sctp_stale_cookie_msg *);
scm->ph.param_type = htons(SCTP_CAUSE_STALE_COOKIE);
scm->ph.param_length = htons((sizeof(struct sctp_paramhdr) +
(sizeof(uint32_t))));
/* seconds to usec */
tim = (now.tv_sec - time_expires.tv_sec) * 1000000;
/* add in usec */
if (tim == 0)
tim = now.tv_usec - cookie->time_entered.tv_usec;
scm->time_usec = htonl(tim);
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_send_operr_to(m, iphlen, op_err, cookie->peers_vtag,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* Now we must see with the lookup address if we have an existing
* asoc. This will only happen if we were in the COOKIE-WAIT state
* and a INIT collided with us and somewhere the peer sent the
* cookie on another address besides the single address our assoc
* had for him. In this case we will have one of the tie-tags set at
* least AND the address field in the cookie can be used to look it
* up.
*/
to = NULL;
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
switch (cookie->addr_type) {
#ifdef INET6
case SCTP_IPV6_ADDRESS:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
memset(&sin6, 0, sizeof(sin6));
sin6.sin6_family = AF_INET6;
sin6.sin6_len = sizeof(sin6);
sin6.sin6_port = sh->src_port;
sin6.sin6_scope_id = cookie->scope_id;
memcpy(&sin6.sin6_addr.s6_addr, cookie->address,
sizeof(sin6.sin6_addr.s6_addr));
to = (struct sockaddr *)&sin6;
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
break;
#endif
#ifdef INET
case SCTP_IPV4_ADDRESS:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
memset(&sin, 0, sizeof(sin));
sin.sin_family = AF_INET;
sin.sin_len = sizeof(sin);
sin.sin_port = sh->src_port;
sin.sin_addr.s_addr = cookie->address[0];
to = (struct sockaddr *)&sin;
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
break;
#endif
default:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
/* This should not happen */
return (NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if ((*stcb == NULL) && to) {
/* Yep, lets check */
*stcb = sctp_findassociation_ep_addr(inp_p, to, netp, localep_sa, NULL);
if (*stcb == NULL) {
/*
* We should have only got back the same inp. If we
* got back a different ep we have a problem. The
* original findep got back l_inp and now
*/
if (l_inp != *inp_p) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTP_PRINTF("Bad problem find_ep got a diff inp then special_locate?\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
} else {
if (*locked_tcb == NULL) {
/*
* In this case we found the assoc only
* after we locked the create lock. This
* means we are in a colliding case and we
* must make sure that we unlock the tcb if
* its one of the cases where we throw away
* the incoming packets.
*/
*locked_tcb = *stcb;
/*
* We must also increment the inp ref count
* since the ref_count flags was set when we
* did not find the TCB, now we found it
* which reduces the refcount.. we must
* raise it back out to balance it all :-)
*/
SCTP_INP_INCR_REF((*stcb)->sctp_ep);
if ((*stcb)->sctp_ep != l_inp) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTP_PRINTF("Huh? ep:%p diff then l_inp:%p?\n",
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
(*stcb)->sctp_ep, l_inp);
}
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
if (to == NULL) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
return (NULL);
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
cookie_len -= SCTP_SIGNATURE_SIZE;
if (*stcb == NULL) {
/* this is the "normal" case... get a new TCB */
*stcb = sctp_process_cookie_new(m, iphlen, offset, sh, cookie,
cookie_len, *inp_p, netp, to, &notification,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
auth_skipped, auth_offset, auth_len, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else {
/* this is abnormal... cookie-echo on existing TCB */
had_a_existing_tcb = 1;
*stcb = sctp_process_cookie_existing(m, iphlen, offset, sh,
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
cookie, cookie_len, *inp_p, *stcb, netp, to,
&notification, &sac_restart_id, vrf_id, auth_skipped, auth_offset, auth_len, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (*stcb == NULL) {
/* still no TCB... must be bad cookie-echo */
return (NULL);
}
Add support for M_FLOWID.
2011-02-05 19:13:38 +00:00
if ((*netp != NULL) && (m->m_flags & M_FLOWID)) {
(*netp)->flowid = m->m_pkthdr.flowid;
Fix bugs related to M_FLOWID: * Store the flowid when receiving an SCTP/IPv6 packet. * Store the flowid when receiving an SCTP packet with wrong CRC. * Initilize flowid correctly. * Put test code under INVARIANTS. MFC after: 3 months.
2011-02-07 15:04:23 +00:00
#ifdef INVARIANTS
1) Track when flowid does get set. MFC after: 3 months
2011-02-07 08:10:29 +00:00
(*netp)->flowidset = 1;
Fix bugs related to M_FLOWID: * Store the flowid when receiving an SCTP/IPv6 packet. * Store the flowid when receiving an SCTP packet with wrong CRC. * Initilize flowid correctly. * Put test code under INVARIANTS. MFC after: 3 months.
2011-02-07 15:04:23 +00:00
#endif
Add support for M_FLOWID.
2011-02-05 19:13:38 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* Ok, we built an association so confirm the address we sent the
* INIT-ACK to.
*/
netl = sctp_findnet(*stcb, to);
/*
* This code should in theory NOT run but
*/
if (netl == NULL) {
/* TSNH! Huh, why do I need to add this address here? */
int ret;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
ret = sctp_add_remote_addr(*stcb, to, SCTP_DONOT_SETSCOPE,
SCTP_IN_COOKIE_PROC);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
netl = sctp_findnet(*stcb, to);
}
if (netl) {
if (netl->dest_state & SCTP_ADDR_UNCONFIRMED) {
netl->dest_state &= ~SCTP_ADDR_UNCONFIRMED;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
(void)sctp_set_primary_addr((*stcb), (struct sockaddr *)NULL,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
netl);
Bugfix: Make sure that the COMM_UP notificatin is delivered first also on the passive side. MFC after: 3 days.
2011-01-02 10:27:27 +00:00
send_int_conf = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
if (*stcb) {
sctp_timer_start(SCTP_TIMER_TYPE_HEARTBEAT, *inp_p,
*stcb, NULL);
}
if ((*inp_p)->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) {
if (!had_a_existing_tcb ||
(((*inp_p)->sctp_flags & SCTP_PCB_FLAGS_CONNECTED) == 0)) {
/*
* If we have a NEW cookie or the connect never
* reached the connected state during collision we
* must do the TCP accept thing.
*/
struct socket *so, *oso;
struct sctp_inpcb *inp;
if (notification == SCTP_NOTIFY_ASSOC_RESTART) {
/*
* For a restart we will keep the same
* socket, no need to do anything. I THINK!!
*/
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(notification, *stcb, 0, (void *)&sac_restart_id, SCTP_SO_NOT_LOCKED);
Bugfix: Make sure that the COMM_UP notificatin is delivered first also on the passive side. MFC after: 3 days.
2011-01-02 10:27:27 +00:00
if (send_int_conf) {
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_CONFIRMED,
(*stcb), 0, (void *)netl, SCTP_SO_NOT_LOCKED);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (m);
}
oso = (*inp_p)->sctp_socket;
- Remove extra comment for 7.0 (no GIANT here). - Remove unneeded WLOCK/UNLOCK of inp for getting TCB lock. - Fix panic that may occur when freeing an assoc that has partial delivery in progress (may dereference null socket pointer when queuing partial delivery aborted notification) - Some spacing and comment fixes. - Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. Approved by: re@freebsd.org (Bruce Mah)
2007-08-16 01:51:22 +00:00
atomic_add_int(&(*stcb)->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK((*stcb));
Mfp4 CH=177274,177280,177284-177285,177297,177324-177325 VNET socket push back: try to minimize the number of places where we have to switch vnets and narrow down the time we stay switched. Add assertions to the socket code to catch possibly unset vnets as seen in r204147. While this reduces the number of vnet recursion in some places like NFS, POSIX local sockets and some netgraph, .. recursions are impossible to fix. The current expectations are documented at the beginning of uipc_socket.c along with the other information there. Sponsored by: The FreeBSD Foundation Sponsored by: CK Software GmbH Reviewed by: jhb Tested by: zec Tested by: Mikolaj Golub (to.my.trociny gmail.com) MFC after: 2 weeks
2011-02-16 21:29:13 +00:00
CURVNET_SET(oso->so_vnet);
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
so = sonewconn(oso, 0
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
);
Mfp4 CH=177274,177280,177284-177285,177297,177324-177325 VNET socket push back: try to minimize the number of places where we have to switch vnets and narrow down the time we stay switched. Add assertions to the socket code to catch possibly unset vnets as seen in r204147. While this reduces the number of vnet recursion in some places like NFS, POSIX local sockets and some netgraph, .. recursions are impossible to fix. The current expectations are documented at the beginning of uipc_socket.c along with the other information there. Sponsored by: The FreeBSD Foundation Sponsored by: CK Software GmbH Reviewed by: jhb Tested by: zec Tested by: Mikolaj Golub (to.my.trociny gmail.com) MFC after: 2 weeks
2011-02-16 21:29:13 +00:00
CURVNET_RESTORE();
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_LOCK((*stcb));
- Remove extra comment for 7.0 (no GIANT here). - Remove unneeded WLOCK/UNLOCK of inp for getting TCB lock. - Fix panic that may occur when freeing an assoc that has partial delivery in progress (may dereference null socket pointer when queuing partial delivery aborted notification) - Some spacing and comment fixes. - Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. Approved by: re@freebsd.org (Bruce Mah)
2007-08-16 01:51:22 +00:00
atomic_subtract_int(&(*stcb)->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (so == NULL) {
struct mbuf *op_err;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *pcb_so;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Too many sockets */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "process_cookie_new: no room for another socket!\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
op_err = sctp_generate_invmanparam(SCTP_CAUSE_OUT_OF_RESC);
sctp_abort_association(*inp_p, NULL, m, iphlen,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sh, op_err, vrf_id, port);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
pcb_so = SCTP_INP_SO(*inp_p);
atomic_add_int(&(*stcb)->asoc.refcnt, 1);
SCTP_TCB_UNLOCK((*stcb));
SCTP_SOCKET_LOCK(pcb_so, 1);
SCTP_TCB_LOCK((*stcb));
atomic_subtract_int(&(*stcb)->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(*inp_p, *stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_20);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(pcb_so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
inp = (struct sctp_inpcb *)so->so_pcb;
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
SCTP_INP_INCR_REF(inp);
/*
* We add the unbound flag here so that if we get an
* soabort() before we get the move_pcb done, we
* will properly cleanup.
*/
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
inp->sctp_flags = (SCTP_PCB_FLAGS_TCPTYPE |
SCTP_PCB_FLAGS_CONNECTED |
SCTP_PCB_FLAGS_IN_TCPPOOL |
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
SCTP_PCB_FLAGS_UNBOUND |
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(SCTP_PCB_COPY_FLAGS & (*inp_p)->sctp_flags) |
SCTP_PCB_FLAGS_DONT_WAKE);
inp->sctp_features = (*inp_p)->sctp_features;
- send call has a reference to uio->uio_resid in the recent send code, but uio may be NULL on sendfile calls. Change to use sndlen variable. - EMSGSIZE is not being returned in non-blocking mode and needs a small tweak to look if the msg would ever fit when returning EWOULDBLOCK. - FWD-TSN has a bug in stream processing which could cause a panic. This is a follow on to the codenomicon fix. - PDAPI level 1 and 2 do not work unless the reader gets his returned buffer full. Fix so we can break out when at level 1 or 2. - Fix fast-handoff features to copy across properly on accepted sockets - Fix sctp_peeloff() system call when no true system call exists to screen arguments for errors. In cases where a real system call exists the system call itself does this. - Fix raddr leak in recent add-ip code change for bundled asconfs (even when non-bundled asconfs are received) - Make sure ipi_addr lock is held when walking global addr list. Need to change this lock type to a rwlock(). - Add don't wake flag on both input and output when the socket is closing. - When deleting an address verify the interface is correct before allowing the delete to process. This protects panda and unnumbered. - Clean up old sysctl stuff and get rid of the old Open/Net BSD structures. - Add a function to watch the ranges in the sysctl sets. - When appending in the reassembly queue, validate that the assoc has not gone to about to be freed. If so (in the middle) abort out. Note this especially effects MAC I think due to the lock/unlock they do (or with LOCK testing in place). - Netstat patch to get rid of warnings. - Make sure that no data gets queued to inactive/unconfirmed destinations. This especially effect CMT but also makes a impact on regular SCTP as well. - During init collision when we detect seq number out of sync we need to treat it like Case C and discard the cookie (no invarient needed here). - Atomic access to the random store. - When we declare a vtag good, we need to shove it into the time wait hash to prevent further use. When the tag is put into the assoc hash, we need to remove it from the twait hash (where it will surely be). This prevents duplicate tag assignments. - Move decr-ref count to better protect sysctl out of data. - ltrace error corrections in sctp6_usrreq.c - Add hook for interface up/down to be sent to us. - Make sysctl() exported structures independent of processor architecture. - Fix route and src addr cache clearing for delete address case. - Make sure address marked SCTP_DEL_IP_ADDRESS is never selected as src addr. - in icmp handling fixed so we actually look at the icmp codes to figure out what to do. - Modified mobility code. Reception of DELETE IP ADDRESS for a primary destination and SET PRIMARY for a new primary destination is used for retransmission trigger to the new primary destination. Also, in this case, destination of chunks in send_queue are changed to the new primary destination. - Fix so that we disallow sending by mbuf to ever have EEOR mode set upon it. Approved by: re@freebsd.org (B Mah)
2007-09-08 17:48:46 +00:00
inp->sctp_mobility_features = (*inp_p)->sctp_mobility_features;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
inp->sctp_socket = so;
inp->sctp_frag_point = (*inp_p)->sctp_frag_point;
Fix the switching on/off of CMT using sysctl and socket option. Fix the switching on/off of PF and NR-SACKs using sysctl. Add minor improvement in handling malloc failures. Improve the address checks when sending. MFC after: 4 weeks
2010-08-28 17:59:51 +00:00
inp->sctp_cmt_on_off = (*inp_p)->sctp_cmt_on_off;
Make sure that changing the ECN sysctl does not affect exisiting associations and endpoints. MFC after: 3 months.
2011-02-03 19:59:00 +00:00
inp->sctp_ecn_enable = (*inp_p)->sctp_ecn_enable;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
inp->partial_delivery_point = (*inp_p)->partial_delivery_point;
inp->sctp_context = (*inp_p)->sctp_context;
inp->inp_starting_point_for_iterator = NULL;
/*
* copy in the authentication parameters from the
* original endpoint
*/
if (inp->sctp_ep.local_hmacs)
sctp_free_hmaclist(inp->sctp_ep.local_hmacs);
inp->sctp_ep.local_hmacs =
sctp_copy_hmaclist((*inp_p)->sctp_ep.local_hmacs);
if (inp->sctp_ep.local_auth_chunks)
sctp_free_chunklist(inp->sctp_ep.local_auth_chunks);
inp->sctp_ep.local_auth_chunks =
sctp_copy_chunklist((*inp_p)->sctp_ep.local_auth_chunks);
/*
* Now we must move it from one hash table to
* another and get the tcb in the right place.
*/
This fixes a bug with the one-2-one model socket when a user sets up a socket to a server sends data and closes the socket before the server has called accept(). It used to NOT work at all. Now we add a flag to the assoc and defer assoc cleanup so that the accept will suceed.
2010-05-11 17:02:29 +00:00
/*
* This is where the one-2-one socket is put into
* the accept state waiting for the accept!
*/
if (*stcb) {
(*stcb)->asoc.state |= SCTP_STATE_IN_ACCEPT_QUEUE;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_move_pcb_and_assoc(*inp_p, inp, *stcb);
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
atomic_add_int(&(*stcb)->asoc.refcnt, 1);
SCTP_TCB_UNLOCK((*stcb));
Correct two problems relating to sorflush(), which is called to flush read socket buffers in shutdown() and close(): - Call socantrcvmore() before sblock() to dislodge any threads that might be sleeping (potentially indefinitely) while holding sblock(), such as a thread blocked in recv(). - Flag the sblock() call as non-interruptible so that a signal delivered to the thread calling sorflush() doesn't cause sblock() to fail. The sblock() is required to ensure that all other socket consumer threads have, in fact, left, and do not enter, the socket buffer until we're done flushin it. To implement the latter, change the 'flags' argument to sblock() to accept two flags, SBL_WAIT and SBL_NOINTR, rather than one M_WAITOK flag. When SBL_NOINTR is set, it forces a non-interruptible sx acquisition, regardless of the setting of the disposition of SB_NOINTR on the socket buffer; without this change it would be possible for another thread to clear SB_NOINTR between when the socket buffer mutex is released and sblock() is invoked. Reviewed by: bz, kmacy Reported by: Jos Backus <jos at catnook dot com>
2008-01-31 08:22:24 +00:00
sctp_pull_off_control_to_new_inp((*inp_p), inp, *stcb,
0);
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
SCTP_TCB_LOCK((*stcb));
atomic_subtract_int(&(*stcb)->asoc.refcnt, 1);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
/*
* now we must check to see if we were aborted while
* the move was going on and the lock/unlock
* happened.
*/
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/*
* yep it was, we leave the assoc attached
* to the socket since the sctp_inpcb_free()
* call will send an abort for us.
*/
SCTP_INP_DECR_REF(inp);
return (NULL);
}
SCTP_INP_DECR_REF(inp);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Switch over to the new guy */
*inp_p = inp;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(notification, *stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
Bugfix: Make sure that the COMM_UP notificatin is delivered first also on the passive side. MFC after: 3 days.
2011-01-02 10:27:27 +00:00
if (send_int_conf) {
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_CONFIRMED,
(*stcb), 0, (void *)netl, SCTP_SO_NOT_LOCKED);
}
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
/*
* Pull it from the incomplete queue and wake the
* guy
*/
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
atomic_add_int(&(*stcb)->asoc.refcnt, 1);
SCTP_TCB_UNLOCK((*stcb));
SCTP_SOCKET_LOCK(so, 1);
#endif
- most all includes (#include <>) migrate to the sctp_os_bsd.h file - Finally all splxx() are removed - Count error fixed in mapping array which might cause a wrong cumack generation. - Invariants around panic for case D + printf when no invariants. - one-to-one model race condition fixed by using a pre-formed connection and then completing the work so accept won't happen on a non-formed association. - Some additional paranoia checks in sctp_output. - Locks that were missing in the accept code. Approved by: gnn
2007-01-18 09:58:43 +00:00
soisconnected(so);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
- Added some comments to tell where the htcp code comes from. - Fix a LOR on Mac OS X: Do not hold an stcb lock when calling soisconnected for a socket which has the SS_INCOMP bit set on so_state. - fix a comment to be non c++ style. Approved by: re@freebsd.org (B Mah)
2007-09-10 17:06:25 +00:00
SCTP_TCB_LOCK((*stcb));
atomic_subtract_int(&(*stcb)->asoc.refcnt, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (m);
}
}
Bugfix: Make sure that the COMM_UP notificatin is delivered first also on the passive side. MFC after: 3 days.
2011-01-02 10:27:27 +00:00
if ((*inp_p)->sctp_flags & SCTP_PCB_FLAGS_UDPTYPE) {
if (notification) {
sctp_ulp_notify(notification, *stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
}
if (send_int_conf) {
sctp_ulp_notify(SCTP_NOTIFY_INTERFACE_CONFIRMED,
(*stcb), 0, (void *)netl, SCTP_SO_NOT_LOCKED);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
return (m);
}
static void
sctp_handle_cookie_ack(struct sctp_cookie_ack_chunk *cp,
struct sctp_tcb *stcb, struct sctp_nets *net)
{
/* cp must not be used, others call this without a c-ack :-) */
struct sctp_association *asoc;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_cookie_ack: handling COOKIE-ACK\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL)
return;
asoc = &stcb->asoc;
sctp_stop_all_cookie_timers(stcb);
/* process according to association state */
if (SCTP_GET_STATE(asoc) == SCTP_STATE_COOKIE_ECHOED) {
/* state change only needed when I am in right state */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "moving to OPEN state\n");
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
SCTP_SET_STATE(asoc, SCTP_STATE_OPEN);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->state & SCTP_STATE_SHUTDOWN_PENDING) {
sctp_timer_start(SCTP_TIMER_TYPE_SHUTDOWNGUARD,
stcb->sctp_ep, stcb, asoc->primary_destination);
}
/* update RTO */
SCTP_STAT_INCR_COUNTER32(sctps_activeestab);
SCTP_STAT_INCR_GAUGE32(sctps_currestab);
if (asoc->overall_error_count == 0) {
net->RTO = sctp_calculate_rto(stcb, asoc, net,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
&asoc->time_entered, sctp_align_safe_nocopy,
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
SCTP_RTT_FROM_NON_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
(void)SCTP_GETTIME_TIMEVAL(&asoc->time_entered);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_ASSOC_UP, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) ||
(stcb->sctp_ep->sctp_flags & SCTP_PCB_FLAGS_IN_TCPPOOL)) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->sctp_ep->sctp_flags |= SCTP_PCB_FLAGS_CONNECTED;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
Fix a bug reported by Jonathan Leighton in his web-sctp testing at the Univ-of-Del. Basically when a 1-to-1 socket did a socket/bind/send(data)/close. If the timing was right we would dereference a socket that is NULL. MFC after: 1 month
2011-02-13 14:48:11 +00:00
if ((stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) == 0) {
soisconnected(stcb->sctp_socket);
}
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* since we did not send a HB make sure we don't double
* things
*/
net->hb_responded = 1;
Fix a bug reported by Jonathan Leighton in his web-sctp testing at the Univ-of-Del. Basically when a 1-to-1 socket did a socket/bind/send(data)/close. If the timing was right we would dereference a socket that is NULL. MFC after: 1 month
2011-02-13 14:48:11 +00:00
if (stcb->asoc.state & SCTP_STATE_CLOSED_SOCKET) {
/*
* We don't need to do the asconf thing, nor hb or
* autoclose if the socket is closed.
*/
goto closed_socket;
}
sctp_timer_start(SCTP_TIMER_TYPE_HEARTBEAT, stcb->sctp_ep,
stcb, net);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb->asoc.sctp_autoclose_ticks &&
sctp_is_feature_on(stcb->sctp_ep, SCTP_PCB_FLAGS_AUTOCLOSE)) {
sctp_timer_start(SCTP_TIMER_TYPE_AUTOCLOSE,
stcb->sctp_ep, stcb, NULL);
}
/*
- take out a needless panic under invariants for sctp_output.c - Fix addrs's error checking of sctp_sendx(3) when addrcnt is less than SCTP_SMALL_IOVEC_SIZE - re-add back inpcb_bind local address check bypass capability - Fix it so sctp_opt_info is independant of assoc_id postion. - Fix cookie life set to use MSEC_TO_TICKS() macro. - asconf changes o More comment changes/clarifications related to the old local address "not" list which is now an explicit restricted list. o Rename some functions for clarity: - sctp_add/del_local_addr_assoc to xxx_local_addr_restricted() - asconf related iterator functions to sctp_asconf_iterator_xxx() o Fix bug when the same address is deleted and added (and removed from the asconf queue) where the ifa is "freed" twice refcount wise, possibly freeing it completely. o Fix bug in output where the first ASCONF would not go out after the last address is changed (e.g. only goes out when retransmitted). o Fix bug where multiple ASCONFs can be bundled in the same packet with the and with the same serial numbers. o Fix asconf stcb iterator to not send ASCONF until after all work queue entries have been processed. o Change behavior so that when the last address is deleted (auto asconf on a bound all endpoint) no action is taken until an address is added; at that time, an ASCONF add+delete is sent (if the assoc is still up). o Fix local address counting so that address scoping is taken into account. o #ifdef SCTP_TIMER_BASED_ASCONF the old timer triggered sending of ASCONF (after an RTO). The default now is to send ASCONF immediately (except for the case of changing/deleting the last usable address). Approved by: re(ken smith)@freebsd.org
2007-07-24 20:06:02 +00:00
* send ASCONF if parameters are pending and ASCONFs are
* allowed (eg. addresses changed when init/cookie echo were
* in flight)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
if ((sctp_is_feature_on(stcb->sctp_ep, SCTP_PCB_FLAGS_DO_ASCONF)) &&
(stcb->asoc.peer_supports_asconf) &&
(!TAILQ_EMPTY(&stcb->asoc.asconf_queue))) {
- take out a needless panic under invariants for sctp_output.c - Fix addrs's error checking of sctp_sendx(3) when addrcnt is less than SCTP_SMALL_IOVEC_SIZE - re-add back inpcb_bind local address check bypass capability - Fix it so sctp_opt_info is independant of assoc_id postion. - Fix cookie life set to use MSEC_TO_TICKS() macro. - asconf changes o More comment changes/clarifications related to the old local address "not" list which is now an explicit restricted list. o Rename some functions for clarity: - sctp_add/del_local_addr_assoc to xxx_local_addr_restricted() - asconf related iterator functions to sctp_asconf_iterator_xxx() o Fix bug when the same address is deleted and added (and removed from the asconf queue) where the ifa is "freed" twice refcount wise, possibly freeing it completely. o Fix bug in output where the first ASCONF would not go out after the last address is changed (e.g. only goes out when retransmitted). o Fix bug where multiple ASCONFs can be bundled in the same packet with the and with the same serial numbers. o Fix asconf stcb iterator to not send ASCONF until after all work queue entries have been processed. o Change behavior so that when the last address is deleted (auto asconf on a bound all endpoint) no action is taken until an address is added; at that time, an ASCONF add+delete is sent (if the assoc is still up). o Fix local address counting so that address scoping is taken into account. o #ifdef SCTP_TIMER_BASED_ASCONF the old timer triggered sending of ASCONF (after an RTO). The default now is to send ASCONF immediately (except for the case of changing/deleting the last usable address). Approved by: re(ken smith)@freebsd.org
2007-07-24 20:06:02 +00:00
#ifdef SCTP_TIMER_BASED_ASCONF
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_ASCONF,
stcb->sctp_ep, stcb,
stcb->asoc.primary_destination);
- take out a needless panic under invariants for sctp_output.c - Fix addrs's error checking of sctp_sendx(3) when addrcnt is less than SCTP_SMALL_IOVEC_SIZE - re-add back inpcb_bind local address check bypass capability - Fix it so sctp_opt_info is independant of assoc_id postion. - Fix cookie life set to use MSEC_TO_TICKS() macro. - asconf changes o More comment changes/clarifications related to the old local address "not" list which is now an explicit restricted list. o Rename some functions for clarity: - sctp_add/del_local_addr_assoc to xxx_local_addr_restricted() - asconf related iterator functions to sctp_asconf_iterator_xxx() o Fix bug when the same address is deleted and added (and removed from the asconf queue) where the ifa is "freed" twice refcount wise, possibly freeing it completely. o Fix bug in output where the first ASCONF would not go out after the last address is changed (e.g. only goes out when retransmitted). o Fix bug where multiple ASCONFs can be bundled in the same packet with the and with the same serial numbers. o Fix asconf stcb iterator to not send ASCONF until after all work queue entries have been processed. o Change behavior so that when the last address is deleted (auto asconf on a bound all endpoint) no action is taken until an address is added; at that time, an ASCONF add+delete is sent (if the assoc is still up). o Fix local address counting so that address scoping is taken into account. o #ifdef SCTP_TIMER_BASED_ASCONF the old timer triggered sending of ASCONF (after an RTO). The default now is to send ASCONF immediately (except for the case of changing/deleting the last usable address). Approved by: re(ken smith)@freebsd.org
2007-07-24 20:06:02 +00:00
#else
- Get rid of unsused constants for sysctl variables. - Fix panic from mutex unlock on freed lock when ASCONF-ACK aborts an assoc - Fix panic from addr lock recursion when ASCONFs are queued in the front states - ASCONFs "queued" in the front states should really be bundled after the COOKIE-ACK, not in front of it - Fix issue with addresses deleted in the front states from being sent with ASCONF(DELETE)-- replaced sctp_asconf_queue_add_sa() with delete specific function - Comment change in sctp.h the drafts are now RFC's Approved by: re@freebsd.org (B Mah)
2007-09-15 19:07:42 +00:00
sctp_send_asconf(stcb, stcb->asoc.primary_destination,
SCTP_ADDR_NOT_LOCKED);
- take out a needless panic under invariants for sctp_output.c - Fix addrs's error checking of sctp_sendx(3) when addrcnt is less than SCTP_SMALL_IOVEC_SIZE - re-add back inpcb_bind local address check bypass capability - Fix it so sctp_opt_info is independant of assoc_id postion. - Fix cookie life set to use MSEC_TO_TICKS() macro. - asconf changes o More comment changes/clarifications related to the old local address "not" list which is now an explicit restricted list. o Rename some functions for clarity: - sctp_add/del_local_addr_assoc to xxx_local_addr_restricted() - asconf related iterator functions to sctp_asconf_iterator_xxx() o Fix bug when the same address is deleted and added (and removed from the asconf queue) where the ifa is "freed" twice refcount wise, possibly freeing it completely. o Fix bug in output where the first ASCONF would not go out after the last address is changed (e.g. only goes out when retransmitted). o Fix bug where multiple ASCONFs can be bundled in the same packet with the and with the same serial numbers. o Fix asconf stcb iterator to not send ASCONF until after all work queue entries have been processed. o Change behavior so that when the last address is deleted (auto asconf on a bound all endpoint) no action is taken until an address is added; at that time, an ASCONF add+delete is sent (if the assoc is still up). o Fix local address counting so that address scoping is taken into account. o #ifdef SCTP_TIMER_BASED_ASCONF the old timer triggered sending of ASCONF (after an RTO). The default now is to send ASCONF immediately (except for the case of changing/deleting the last usable address). Approved by: re(ken smith)@freebsd.org
2007-07-24 20:06:02 +00:00
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
Fix a bug reported by Jonathan Leighton in his web-sctp testing at the Univ-of-Del. Basically when a 1-to-1 socket did a socket/bind/send(data)/close. If the timing was right we would dereference a socket that is NULL. MFC after: 1 month
2011-02-13 14:48:11 +00:00
closed_socket:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* Toss the cookie if I can */
sctp_toss_old_cookies(stcb, asoc);
if (!TAILQ_EMPTY(&asoc->sent_queue)) {
/* Restart the timer if we have pending data */
struct sctp_tmit_chunk *chk;
chk = TAILQ_FIRST(&asoc->sent_queue);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_SEND, stcb->sctp_ep, stcb, chk->whoTo);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
static void
sctp_handle_ecn_echo(struct sctp_ecne_chunk *cp,
struct sctp_tcb *stcb)
{
struct sctp_nets *net;
struct sctp_tmit_chunk *lchk;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
struct sctp_ecne_chunk bkup;
uint8_t override_bit = 0;
uint32_t tsn, window_data_tsn;
int len;
1) Move per John Baldwin to mp_maxid 2) Some signed/unsigned errors found by Mac OS compiler (from Michael) 3) a couple of copyright updates on the effected files. MFC after: 3 months
2011-02-03 19:22:21 +00:00
unsigned int pkt_cnt;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
len = ntohs(cp->ch.chunk_length);
if ((len != sizeof(struct sctp_ecne_chunk)) &&
(len != sizeof(struct old_sctp_ecne_chunk))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return;
}
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if (len == sizeof(struct old_sctp_ecne_chunk)) {
/* Its the old format */
memcpy(&bkup, cp, sizeof(struct old_sctp_ecne_chunk));
bkup.num_pkts_since_cwr = htonl(1);
cp = &bkup;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR(sctps_recvecne);
tsn = ntohl(cp->tsn);
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
pkt_cnt = ntohl(cp->num_pkts_since_cwr);
lchk = TAILQ_LAST(&stcb->asoc.send_queue, sctpchunk_listhead);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (lchk == NULL) {
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
window_data_tsn = stcb->asoc.sending_seq - 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else {
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
window_data_tsn = lchk->rec.data.TSN_seq;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
/* Find where it was sent to if possible. */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
net = NULL;
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH(lchk, &stcb->asoc.sent_queue, sctp_next) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (lchk->rec.data.TSN_seq == tsn) {
net = lchk->whoTo;
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
net->ecn_prev_cwnd = lchk->rec.data.cwnd_at_send;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
}
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
if (SCTP_TSN_GT(lchk->rec.data.TSN_seq, tsn)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if (net == NULL) {
/*
* What to do. A previous send of a CWR was possibly lost.
* See how old it is, we may have it marked on the actual
* net.
*/
TAILQ_FOREACH(net, &stcb->asoc.nets, sctp_next) {
if (tsn == net->last_cwr_tsn) {
/* Found him, send it off */
goto out;
}
}
/*
* If we reach here, we need to send a special CWR that says
* hey, we did this a long time ago and you lost the
* response.
*/
net = TAILQ_FIRST(&stcb->asoc.nets);
override_bit = SCTP_CWR_REDUCE_OVERRIDE;
}
out:
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
if (SCTP_TSN_GT(tsn, net->cwr_window_tsn) &&
((override_bit & SCTP_CWR_REDUCE_OVERRIDE) == 0)) {
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
/*
* JRS - Use the congestion control given in the pluggable
* CC module
*/
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
int ocwnd;
ocwnd = net->cwnd;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
stcb->asoc.cc_functions.sctp_cwnd_update_after_ecn_echo(stcb, net, 0, pkt_cnt);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
* We reduce once every RTT. So we will only lower cwnd at
* the next sending seq i.e. the window_data_tsn
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
net->cwr_window_tsn = window_data_tsn;
net->ecn_ce_pkt_cnt += pkt_cnt;
net->lost_cnt = pkt_cnt;
net->last_cwr_tsn = tsn;
} else {
override_bit |= SCTP_CWR_IN_SAME_WINDOW;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
if (SCTP_TSN_GT(tsn, net->last_cwr_tsn) &&
((override_bit & SCTP_CWR_REDUCE_OVERRIDE) == 0)) {
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
/*
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
* Another loss in the same window update how many
* marks/packets lost we have had.
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
*/
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
int cnt = 1;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if (pkt_cnt > net->lost_cnt) {
/* Should be the case */
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
cnt = (pkt_cnt - net->lost_cnt);
net->ecn_ce_pkt_cnt += cnt;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
}
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
net->lost_cnt = pkt_cnt;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
net->last_cwr_tsn = tsn;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
/*
* Most CC functions will ignore this call, since we
* are in-window yet of the initial CE the peer saw.
*/
stcb->asoc.cc_functions.sctp_cwnd_update_after_ecn_echo(stcb, net, 1, cnt);
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* We always send a CWR this way if our previous one was lost our
* peer will get an update, or if it is not time again to reduce we
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
* still get the cwr to the peer. Note we set the override when we
* could not find the TSN on the chunk or the destination network.
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*/
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
sctp_send_cwr(stcb, net, net->last_cwr_tsn, override_bit);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
static void
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
sctp_handle_ecn_cwr(struct sctp_cwr_chunk *cp, struct sctp_tcb *stcb, struct sctp_nets *net)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
/*
* Here we get a CWR from the peer. We must look in the outqueue and
* make sure that we have a covered ECNE in teh control chunk part.
* If so remove it.
*/
struct sctp_tmit_chunk *chk;
struct sctp_ecne_chunk *ecne;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
int override;
uint32_t cwr_tsn;
cwr_tsn = ntohl(cp->tsn);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
override = cp->ch.chunk_flags & SCTP_CWR_REDUCE_OVERRIDE;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
TAILQ_FOREACH(chk, &stcb->asoc.control_send_queue, sctp_next) {
if (chk->rec.chunk_id.id != SCTP_ECN_ECHO) {
continue;
}
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if ((override == 0) && (chk->whoTo != net)) {
/* Must be from the right src unless override is set */
continue;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
ecne = mtod(chk->data, struct sctp_ecne_chunk *);
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if (SCTP_TSN_GE(cwr_tsn, ntohl(ecne->tsn))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* this covers this ECNE, we can remove it */
stcb->asoc.ecn_echo_cnt_onq--;
TAILQ_REMOVE(&stcb->asoc.control_send_queue, chk,
sctp_next);
if (chk->data) {
sctp_m_freem(chk->data);
chk->data = NULL;
}
stcb->asoc.ctrl_queue_cnt--;
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_chunk(stcb, chk, SCTP_SO_NOT_LOCKED);
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
if (override == 0) {
break;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
}
static void
sctp_handle_shutdown_complete(struct sctp_shutdown_complete_chunk *cp,
struct sctp_tcb *stcb, struct sctp_nets *net)
{
struct sctp_association *asoc;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_shutdown_complete: handling SHUTDOWN-COMPLETE\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL)
return;
asoc = &stcb->asoc;
/* process according to association state */
if (SCTP_GET_STATE(asoc) != SCTP_STATE_SHUTDOWN_ACK_SENT) {
/* unexpected SHUTDOWN-COMPLETE... so ignore... */
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_shutdown_complete: not in SCTP_STATE_SHUTDOWN_ACK_SENT --- ignore\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(stcb);
return;
}
/* notify upper layer protocol */
if (stcb->sctp_socket) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_ASSOC_DOWN, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* are the queues empty? they should be */
if (!TAILQ_EMPTY(&asoc->send_queue) ||
!TAILQ_EMPTY(&asoc->sent_queue) ||
Add stream scheduling support. This work is based on a patch received from Robin Seggelmann. MFC after: 3 months.
2011-01-23 19:36:28 +00:00
!stcb->asoc.ss_functions.sctp_ss_is_empty(stcb, asoc)) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_report_all_outbound(stcb, 0, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
/* stop the timer */
Use the pru_flush infrastructure to avoid a panic PR: 122710 MFC after: 1 week
2008-04-14 18:13:33 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_SHUTDOWNACK, stcb->sctp_ep, stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_22);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR_COUNTER32(sctps_shutdown);
/* free the TCB */
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2,
"sctp_handle_shutdown_complete: calls free-asoc\n");
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(stcb->sctp_ep);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(stcb->sctp_ep, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_23);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return;
}
static int
process_chunk_drop(struct sctp_tcb *stcb, struct sctp_chunk_desc *desc,
struct sctp_nets *net, uint8_t flg)
{
switch (desc->chunk_type) {
Fix the the SCTP_WITH_NO_CSUM option when used in combination with interface supporting CRC offload. While at it, make use of the feature that the loopback interface provides CRC offloading. MFC after: 4 weeks
2010-08-29 18:50:30 +00:00
case SCTP_DATA:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* find the tsn to resend (possibly */
{
uint32_t tsn;
struct sctp_tmit_chunk *tp1;
tsn = ntohl(desc->tsn_ifany);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH(tp1, &stcb->asoc.sent_queue, sctp_next) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (tp1->rec.data.TSN_seq == tsn) {
/* found it */
break;
}
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
if (SCTP_TSN_GT(tp1->rec.data.TSN_seq, tsn)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* not found */
tp1 = NULL;
break;
}
}
if (tp1 == NULL) {
/*
* Do it the other way , aka without paying
* attention to queue seq order.
*/
SCTP_STAT_INCR(sctps_pdrpdnfnd);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH(tp1, &stcb->asoc.sent_queue, sctp_next) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (tp1->rec.data.TSN_seq == tsn) {
/* found it */
break;
}
}
}
if (tp1 == NULL) {
SCTP_STAT_INCR(sctps_pdrptsnnf);
}
if ((tp1) && (tp1->sent < SCTP_DATAGRAM_ACKED)) {
uint8_t *ddp;
Do not resend DATA chunks without delay when dropped by the peer and the CRC was correct. MFC after: 3 days.
2010-11-06 13:43:18 +00:00
if (((flg & SCTP_BADCRC) == 0) &&
((flg & SCTP_FROM_MIDDLE_BOX) == 0)) {
return (0);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((stcb->asoc.peers_rwnd == 0) &&
((flg & SCTP_FROM_MIDDLE_BOX) == 0)) {
SCTP_STAT_INCR(sctps_pdrpdiwnp);
return (0);
}
if (stcb->asoc.peers_rwnd == 0 &&
(flg & SCTP_FROM_MIDDLE_BOX)) {
SCTP_STAT_INCR(sctps_pdrpdizrw);
return (0);
}
ddp = (uint8_t *) (mtod(tp1->data, caddr_t)+
sizeof(struct sctp_data_chunk));
{
unsigned int iii;
for (iii = 0; iii < sizeof(desc->data_bytes);
iii++) {
if (ddp[iii] != desc->data_bytes[iii]) {
SCTP_STAT_INCR(sctps_pdrpbadd);
return (-1);
}
}
}
if (tp1->do_rtt) {
/*
* this guy had a RTO calculation
* pending on it, cancel it
*/
Tunes and fixes the new DC-CC to seem to hit the right mix. Still may need some tweaks but it appears to almost not give away too much to an RFC2581 flow, but can really minimize the amount of buffers used in the net. MFC after: 3 months
2011-03-08 11:58:25 +00:00
if (tp1->whoTo->rto_needed == 0) {
tp1->whoTo->rto_needed = 1;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
tp1->do_rtt = 0;
}
SCTP_STAT_INCR(sctps_pdrpmark);
if (tp1->sent != SCTP_DATAGRAM_RESEND)
sctp_ucount_incr(stcb->asoc.sent_queue_retran_cnt);
/*
* mark it as if we were doing a FR, since
* we will be getting gap ack reports behind
* the info from the router.
*/
tp1->rec.data.doing_fast_retransmit = 1;
/*
* mark the tsn with what sequences can
* cause a new FR.
*/
if (TAILQ_EMPTY(&stcb->asoc.send_queue)) {
tp1->rec.data.fast_retran_tsn = stcb->asoc.sending_seq;
} else {
tp1->rec.data.fast_retran_tsn = (TAILQ_FIRST(&stcb->asoc.send_queue))->rec.data.TSN_seq;
}
/* restart the timer */
sctp_timer_stop(SCTP_TIMER_TYPE_SEND, stcb->sctp_ep,
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
stcb, tp1->whoTo, SCTP_FROM_SCTP_INPUT + SCTP_LOC_24);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_SEND, stcb->sctp_ep,
stcb, tp1->whoTo);
/* fix counts and things */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_FLIGHT_LOGGING_ENABLE) {
- Fix so ifn's are properly deleted when the ref count goes to 0. - Fix so VRF's will clean themselves up when no references are around. - Allow sctp_ifa to be passed into inpcb_bind, addr_mgmt_ep_sa to bypass normal validation checks. - turn auto-asconf off for subset bound sockets - Moves all logging to use KTR. This gets rid of most of the logging #ifdef's with a few exceptions reducing the number of config options for SCTP.
2007-06-14 22:59:04 +00:00
sctp_misc_ints(SCTP_FLIGHT_LOG_DOWN_PDRP,
tp1->whoTo->flight_size,
tp1->book_size,
(uintptr_t) stcb,
tp1->rec.data.TSN_seq);
}
Many bug fixes (from the IETF hack-fest): - PR-SCTP had major issues when skipping through a multi-part message. o Did not look at socket buffer. o Did not properly handle the reassmebly queue. o The MARKED segments could interfere and un-skip a chunk causing a problem with the proper FWD-TSN. o No FR of FWD-TSN's was being done. - NR-Sack code was basically disabled. It needed fixes that never got into the real code. - CMT code had issues when the two paths were NOT the same b/w. We found a few small bugs, but also the critcal one here was not dividing the rwnd amongst the paths. Obtained from: Michael Tuexen and myself at the IETF hack-fest ;-)
2009-04-04 11:43:32 +00:00
if (tp1->sent < SCTP_DATAGRAM_RESEND) {
sctp_flight_size_decrease(tp1);
sctp_total_flight_decrease(stcb, tp1);
}
Fix a flightsize bug related to the processing of PKTDRP reports. MFC after: 3 days.
2010-12-17 15:39:55 +00:00
tp1->sent = SCTP_DATAGRAM_RESEND;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} {
/* audit code */
unsigned int audit;
audit = 0;
TAILQ_FOREACH(tp1, &stcb->asoc.sent_queue, sctp_next) {
if (tp1->sent == SCTP_DATAGRAM_RESEND)
audit++;
}
TAILQ_FOREACH(tp1, &stcb->asoc.control_send_queue,
sctp_next) {
if (tp1->sent == SCTP_DATAGRAM_RESEND)
audit++;
}
if (audit != stcb->asoc.sent_queue_retran_cnt) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTP_PRINTF("**Local Audit finds cnt:%d asoc cnt:%d\n",
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
audit, stcb->asoc.sent_queue_retran_cnt);
#ifndef SCTP_AUDITING_ENABLED
stcb->asoc.sent_queue_retran_cnt = audit;
#endif
}
}
}
break;
case SCTP_ASCONF:
{
struct sctp_tmit_chunk *asconf;
TAILQ_FOREACH(asconf, &stcb->asoc.control_send_queue,
sctp_next) {
if (asconf->rec.chunk_id.id == SCTP_ASCONF) {
break;
}
}
if (asconf) {
if (asconf->sent != SCTP_DATAGRAM_RESEND)
sctp_ucount_incr(stcb->asoc.sent_queue_retran_cnt);
asconf->sent = SCTP_DATAGRAM_RESEND;
asconf->snd_count--;
}
}
break;
case SCTP_INITIATION:
/* resend the INIT */
stcb->asoc.dropped_special_cnt++;
if (stcb->asoc.dropped_special_cnt < SCTP_RETRY_DROPPED_THRESH) {
/*
* If we can get it in, in a few attempts we do
* this, otherwise we let the timer fire.
*/
sctp_timer_stop(SCTP_TIMER_TYPE_INIT, stcb->sctp_ep,
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
stcb, net, SCTP_FROM_SCTP_INPUT + SCTP_LOC_25);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_send_initiate(stcb->sctp_ep, stcb, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
break;
case SCTP_SELECTIVE_ACK:
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
case SCTP_NR_SELECTIVE_ACK:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* resend the sack */
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_send_sack(stcb, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_HEARTBEAT_REQUEST:
/* resend a demand HB */
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
if ((stcb->asoc.overall_error_count + 3) < stcb->asoc.max_send_times) {
/*
* Only retransmit if we KNOW we wont destroy the
* tcb
*/
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
(void)sctp_send_hb(stcb, 1, net, SCTP_SO_NOT_LOCKED);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_SHUTDOWN:
sctp_send_shutdown(stcb, net);
break;
case SCTP_SHUTDOWN_ACK:
sctp_send_shutdown_ack(stcb, net);
break;
case SCTP_COOKIE_ECHO:
{
struct sctp_tmit_chunk *cookie;
cookie = NULL;
TAILQ_FOREACH(cookie, &stcb->asoc.control_send_queue,
sctp_next) {
if (cookie->rec.chunk_id.id == SCTP_COOKIE_ECHO) {
break;
}
}
if (cookie) {
if (cookie->sent != SCTP_DATAGRAM_RESEND)
sctp_ucount_incr(stcb->asoc.sent_queue_retran_cnt);
cookie->sent = SCTP_DATAGRAM_RESEND;
sctp_stop_all_cookie_timers(stcb);
}
}
break;
case SCTP_COOKIE_ACK:
sctp_send_cookie_ack(stcb);
break;
case SCTP_ASCONF_ACK:
/* resend last asconf ack */
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
sctp_send_asconf_ack(stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_FORWARD_CUM_TSN:
send_forward_tsn(stcb, &stcb->asoc);
break;
/* can't do anything with these */
case SCTP_PACKET_DROPPED:
case SCTP_INITIATION_ACK: /* this should not happen */
case SCTP_HEARTBEAT_ACK:
case SCTP_ABORT_ASSOCIATION:
case SCTP_OPERATION_ERROR:
case SCTP_SHUTDOWN_COMPLETE:
case SCTP_ECN_ECHO:
case SCTP_ECN_CWR:
default:
break;
}
return (0);
}
void
sctp_reset_in_stream(struct sctp_tcb *stcb, int number_entries, uint16_t * list)
{
int i;
uint16_t temp;
/*
* We set things to 0xffff since this is the last delivered sequence
* and we will be sending in 0 after the reset.
*/
if (number_entries) {
for (i = 0; i < number_entries; i++) {
temp = ntohs(list[i]);
if (temp >= stcb->asoc.streamincnt) {
continue;
}
stcb->asoc.strmin[temp].last_sequence_delivered = 0xffff;
}
} else {
list = NULL;
for (i = 0; i < stcb->asoc.streamincnt; i++) {
stcb->asoc.strmin[i].last_sequence_delivered = 0xffff;
}
}
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_RECV, stcb, number_entries, (void *)list, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
static void
sctp_reset_out_streams(struct sctp_tcb *stcb, int number_entries, uint16_t * list)
{
int i;
if (number_entries == 0) {
for (i = 0; i < stcb->asoc.streamoutcnt; i++) {
stcb->asoc.strmout[i].next_sequence_sent = 0;
}
} else if (number_entries) {
for (i = 0; i < number_entries; i++) {
uint16_t temp;
temp = ntohs(list[i]);
if (temp >= stcb->asoc.streamoutcnt) {
/* no such stream */
continue;
}
stcb->asoc.strmout[temp].next_sequence_sent = 0;
}
}
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_SEND, stcb, number_entries, (void *)list, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
struct sctp_stream_reset_out_request *
sctp_find_stream_reset(struct sctp_tcb *stcb, uint32_t seq, struct sctp_tmit_chunk **bchk)
{
struct sctp_association *asoc;
struct sctp_stream_reset_out_req *req;
struct sctp_stream_reset_out_request *r;
struct sctp_tmit_chunk *chk;
int len, clen;
asoc = &stcb->asoc;
if (TAILQ_EMPTY(&stcb->asoc.control_send_queue)) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
asoc->stream_reset_outstanding = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
if (stcb->asoc.str_reset == NULL) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
asoc->stream_reset_outstanding = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
chk = stcb->asoc.str_reset;
if (chk->data == NULL) {
return (NULL);
}
if (bchk) {
/* he wants a copy of the chk pointer */
*bchk = chk;
}
clen = chk->send_size;
req = mtod(chk->data, struct sctp_stream_reset_out_req *);
r = &req->sr_req;
if (ntohl(r->request_seq) == seq) {
/* found it */
return (r);
}
len = SCTP_SIZE32(ntohs(r->ph.param_length));
if (clen > (len + (int)sizeof(struct sctp_chunkhdr))) {
/* move to the next one, there can only be a max of two */
r = (struct sctp_stream_reset_out_request *)((caddr_t)r + len);
if (ntohl(r->request_seq) == seq) {
return (r);
}
}
/* that seq is not here */
return (NULL);
}
static void
sctp_clean_up_stream_reset(struct sctp_tcb *stcb)
{
struct sctp_association *asoc;
struct sctp_tmit_chunk *chk = stcb->asoc.str_reset;
if (stcb->asoc.str_reset == NULL) {
return;
}
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
asoc = &stcb->asoc;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
sctp_timer_stop(SCTP_TIMER_TYPE_STRRESET, stcb->sctp_ep, stcb, chk->whoTo, SCTP_FROM_SCTP_INPUT + SCTP_LOC_26);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
TAILQ_REMOVE(&asoc->control_send_queue,
chk,
sctp_next);
if (chk->data) {
sctp_m_freem(chk->data);
chk->data = NULL;
}
asoc->ctrl_queue_cnt--;
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_chunk(stcb, chk, SCTP_SO_NOT_LOCKED);
- Incorrect error EAGAIN returned for invalid send on a locked stream (using EEOR mode). Changed to EINVAL (in sctp_output.c) - Static analysis comments added - fix in mobility code to return a value (static analysis found). - sctp6_notify function made visible instead of static (this is needed for Panda). Approved by: re@freebsd.org (B Mah)
2007-09-13 10:36:43 +00:00
/* sa_ignore NO_NULL_CHK */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.str_reset = NULL;
}
static int
sctp_handle_stream_reset_response(struct sctp_tcb *stcb,
uint32_t seq, uint32_t action,
Fixes an issue with handling of stream reset. When a reset comes in we need to calculate the length and therefore the number of listed streams (if any) based on the TLV type. Otherwise if we get a retran we could in theory panic by sending a notification to a user with a incorrect list and thus no memory listing the streams. Found in IOS by devtest :-) Approved by: gnn
2006-11-09 21:01:07 +00:00
struct sctp_stream_reset_response *respin)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
uint16_t type;
int lparm_len;
struct sctp_association *asoc = &stcb->asoc;
struct sctp_tmit_chunk *chk;
struct sctp_stream_reset_out_request *srparam;
int number_entries;
if (asoc->stream_reset_outstanding == 0) {
/* duplicate */
return (0);
}
if (seq == stcb->asoc.str_reset_seq_out) {
srparam = sctp_find_stream_reset(stcb, seq, &chk);
if (srparam) {
stcb->asoc.str_reset_seq_out++;
type = ntohs(srparam->ph.param_type);
lparm_len = ntohs(srparam->ph.param_length);
if (type == SCTP_STR_RESET_OUT_REQUEST) {
Fixes an issue with handling of stream reset. When a reset comes in we need to calculate the length and therefore the number of listed streams (if any) based on the TLV type. Otherwise if we get a retran we could in theory panic by sending a notification to a user with a incorrect list and thus no memory listing the streams. Found in IOS by devtest :-) Approved by: gnn
2006-11-09 21:01:07 +00:00
number_entries = (lparm_len - sizeof(struct sctp_stream_reset_out_request)) / sizeof(uint16_t);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
asoc->stream_reset_out_is_outstanding = 0;
if (asoc->stream_reset_outstanding)
asoc->stream_reset_outstanding--;
if (action == SCTP_STREAM_RESET_PERFORMED) {
/* do it */
sctp_reset_out_streams(stcb, number_entries, srparam->list_of_streams);
} else {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_FAILED_OUT, stcb, number_entries, srparam->list_of_streams, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
} else if (type == SCTP_STR_RESET_IN_REQUEST) {
/* Answered my request */
Fixes an issue with handling of stream reset. When a reset comes in we need to calculate the length and therefore the number of listed streams (if any) based on the TLV type. Otherwise if we get a retran we could in theory panic by sending a notification to a user with a incorrect list and thus no memory listing the streams. Found in IOS by devtest :-) Approved by: gnn
2006-11-09 21:01:07 +00:00
number_entries = (lparm_len - sizeof(struct sctp_stream_reset_in_request)) / sizeof(uint16_t);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (asoc->stream_reset_outstanding)
asoc->stream_reset_outstanding--;
if (action != SCTP_STREAM_RESET_PERFORMED) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_FAILED_IN, stcb, number_entries, srparam->list_of_streams, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
} else if (type == SCTP_STR_RESET_ADD_STREAMS) {
/* Ok we now may have more streams */
Fix the add stream feature of strm-reset to really work: - Fix the copy, we can't do a blind copy but must transfer the data from the old to the new. - Fix the ACK processing so we properly stop retransmitting the thing. - Fix it so if we get a retran we will properly reply with the saved response without doing anything. MFC after: 1 month
2009-02-27 20:54:45 +00:00
if (asoc->stream_reset_outstanding)
asoc->stream_reset_outstanding--;
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
if (action == SCTP_STREAM_RESET_PERFORMED) {
/* Put the new streams into effect */
stcb->asoc.streamoutcnt = stcb->asoc.strm_realoutsize;
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_ADD_OK, stcb,
(uint32_t) stcb->asoc.streamoutcnt, NULL, SCTP_SO_NOT_LOCKED);
} else {
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_ADD_FAIL, stcb,
(uint32_t) stcb->asoc.streamoutcnt, NULL, SCTP_SO_NOT_LOCKED);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else if (type == SCTP_STR_RESET_TSN_REQUEST) {
/**
* a) Adopt the new in tsn.
* b) reset the map
* c) Adopt the new out-tsn
*/
struct sctp_stream_reset_response_tsn *resp;
struct sctp_forward_tsn_chunk fwdtsn;
int abort_flag = 0;
if (respin == NULL) {
/* huh ? */
return (0);
}
if (action == SCTP_STREAM_RESET_PERFORMED) {
resp = (struct sctp_stream_reset_response_tsn *)respin;
asoc->stream_reset_outstanding--;
fwdtsn.ch.chunk_length = htons(sizeof(struct sctp_forward_tsn_chunk));
fwdtsn.ch.chunk_type = SCTP_FORWARD_CUM_TSN;
fwdtsn.new_cumulative_tsn = htonl(ntohl(resp->senders_next_tsn) - 1);
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
sctp_handle_forward_tsn(stcb, &fwdtsn, &abort_flag, NULL, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_flag) {
return (1);
}
stcb->asoc.highest_tsn_inside_map = (ntohl(resp->senders_next_tsn) - 1);
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MAP_LOGGING_ENABLE) {
sctp_log_map(0, 7, asoc->highest_tsn_inside_map, SCTP_MAP_SLIDE_RESULT);
}
- Out with some printfs. - Fix a initialization of last_tsn_used - Fix handling of mapped IPv4 addresses Obtained from: Michael Tuexen and I :-) MFC after: 1 week
2008-07-29 09:06:35 +00:00
stcb->asoc.tsn_last_delivered = stcb->asoc.cumulative_tsn = stcb->asoc.highest_tsn_inside_map;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.mapping_array_base_tsn = ntohl(resp->senders_next_tsn);
memset(stcb->asoc.mapping_array, 0, stcb->asoc.mapping_array_size);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
Fix for NR-Sack code. The code was NOT working properly when enabled. Basically most of the operations were incorrect causing bad sacks when you enabled nr-sack. The fixes range across 4 files and unifiy most of the processing so that we only test nr_sack flags to decide which type of sack to generate. Optimization left for this is to combine the sack generation code and make it capable of generating either sack thus shrinking out a routine. Reviewed by: tuexen@freebsd.org
2010-03-24 19:45:36 +00:00
stcb->asoc.highest_tsn_inside_nr_map = stcb->asoc.highest_tsn_inside_map;
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
memset(stcb->asoc.nr_mapping_array, 0, stcb->asoc.mapping_array_size);
Fix for NR-Sack code. The code was NOT working properly when enabled. Basically most of the operations were incorrect causing bad sacks when you enabled nr-sack. The fixes range across 4 files and unifiy most of the processing so that we only test nr_sack flags to decide which type of sack to generate. Optimization left for this is to combine the sack generation code and make it capable of generating either sack thus shrinking out a routine. Reviewed by: tuexen@freebsd.org
2010-03-24 19:45:36 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.sending_seq = ntohl(resp->receivers_next_tsn);
stcb->asoc.last_acked_seq = stcb->asoc.cumulative_tsn;
sctp_reset_out_streams(stcb, 0, (uint16_t *) NULL);
sctp_reset_in_stream(stcb, 0, (uint16_t *) NULL);
}
}
/* get rid of the request and get the request flags */
if (asoc->stream_reset_outstanding == 0) {
sctp_clean_up_stream_reset(stcb);
}
}
}
return (0);
}
static void
sctp_handle_str_reset_request_in(struct sctp_tcb *stcb,
struct sctp_tmit_chunk *chk,
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
struct sctp_stream_reset_in_request *req, int trunc)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
uint32_t seq;
int len, i;
int number_entries;
uint16_t temp;
/*
* peer wants me to send a str-reset to him for my outgoing seq's if
* seq_in is right.
*/
struct sctp_association *asoc = &stcb->asoc;
seq = ntohl(req->request_seq);
if (asoc->str_reset_seq_in == seq) {
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
if (trunc) {
/* Can't do it, since they exceeded our buffer size */
asoc->last_reset_action[1] = asoc->last_reset_action[0];
asoc->last_reset_action[0] = SCTP_STREAM_RESET_DENIED;
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[0]);
} else if (stcb->asoc.stream_reset_out_is_outstanding == 0) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
len = ntohs(req->ph.param_length);
number_entries = ((len - sizeof(struct sctp_stream_reset_in_request)) / sizeof(uint16_t));
for (i = 0; i < number_entries; i++) {
temp = ntohs(req->list_of_streams[i]);
req->list_of_streams[i] = temp;
}
/* move the reset action back one */
asoc->last_reset_action[1] = asoc->last_reset_action[0];
asoc->last_reset_action[0] = SCTP_STREAM_RESET_PERFORMED;
sctp_add_stream_reset_out(chk, number_entries, req->list_of_streams,
asoc->str_reset_seq_out,
seq, (asoc->sending_seq - 1));
asoc->stream_reset_out_is_outstanding = 1;
asoc->str_reset = chk;
sctp_timer_start(SCTP_TIMER_TYPE_STRRESET, stcb->sctp_ep, stcb, chk->whoTo);
stcb->asoc.stream_reset_outstanding++;
} else {
/* Can't do it, since we have sent one out */
asoc->last_reset_action[1] = asoc->last_reset_action[0];
asoc->last_reset_action[0] = SCTP_STREAM_RESET_TRY_LATER;
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[0]);
}
asoc->str_reset_seq_in++;
} else if (asoc->str_reset_seq_in - 1 == seq) {
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[0]);
} else if (asoc->str_reset_seq_in - 2 == seq) {
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[1]);
} else {
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_BAD_SEQNO);
}
}
static int
sctp_handle_str_reset_request_tsn(struct sctp_tcb *stcb,
struct sctp_tmit_chunk *chk,
struct sctp_stream_reset_tsn_request *req)
{
/* reset all in and out and update the tsn */
/*
* A) reset my str-seq's on in and out. B) Select a receive next,
* and set cum-ack to it. Also process this selected number as a
* fwd-tsn as well. C) set in the response my next sending seq.
*/
struct sctp_forward_tsn_chunk fwdtsn;
struct sctp_association *asoc = &stcb->asoc;
int abort_flag = 0;
uint32_t seq;
seq = ntohl(req->request_seq);
if (asoc->str_reset_seq_in == seq) {
fwdtsn.ch.chunk_length = htons(sizeof(struct sctp_forward_tsn_chunk));
fwdtsn.ch.chunk_type = SCTP_FORWARD_CUM_TSN;
fwdtsn.ch.chunk_flags = 0;
fwdtsn.new_cumulative_tsn = htonl(stcb->asoc.highest_tsn_inside_map + 1);
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
sctp_handle_forward_tsn(stcb, &fwdtsn, &abort_flag, NULL, 0);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_flag) {
return (1);
}
stcb->asoc.highest_tsn_inside_map += SCTP_STREAM_RESET_TSN_DELTA;
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MAP_LOGGING_ENABLE) {
sctp_log_map(0, 10, asoc->highest_tsn_inside_map, SCTP_MAP_SLIDE_RESULT);
}
- Out with some printfs. - Fix a initialization of last_tsn_used - Fix handling of mapped IPv4 addresses Obtained from: Michael Tuexen and I :-) MFC after: 1 week
2008-07-29 09:06:35 +00:00
stcb->asoc.tsn_last_delivered = stcb->asoc.cumulative_tsn = stcb->asoc.highest_tsn_inside_map;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.mapping_array_base_tsn = stcb->asoc.highest_tsn_inside_map + 1;
memset(stcb->asoc.mapping_array, 0, stcb->asoc.mapping_array_size);
Fix for NR-Sack code. The code was NOT working properly when enabled. Basically most of the operations were incorrect causing bad sacks when you enabled nr-sack. The fixes range across 4 files and unifiy most of the processing so that we only test nr_sack flags to decide which type of sack to generate. Optimization left for this is to combine the sack generation code and make it capable of generating either sack thus shrinking out a routine. Reviewed by: tuexen@freebsd.org
2010-03-24 19:45:36 +00:00
stcb->asoc.highest_tsn_inside_nr_map = stcb->asoc.highest_tsn_inside_map;
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
memset(stcb->asoc.nr_mapping_array, 0, stcb->asoc.mapping_array_size);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
atomic_add_int(&stcb->asoc.sending_seq, 1);
/* save off historical data for retrans */
stcb->asoc.last_sending_seq[1] = stcb->asoc.last_sending_seq[0];
stcb->asoc.last_sending_seq[0] = stcb->asoc.sending_seq;
stcb->asoc.last_base_tsnsent[1] = stcb->asoc.last_base_tsnsent[0];
stcb->asoc.last_base_tsnsent[0] = stcb->asoc.mapping_array_base_tsn;
sctp_add_stream_reset_result_tsn(chk,
ntohl(req->request_seq),
SCTP_STREAM_RESET_PERFORMED,
stcb->asoc.sending_seq,
stcb->asoc.mapping_array_base_tsn);
sctp_reset_out_streams(stcb, 0, (uint16_t *) NULL);
sctp_reset_in_stream(stcb, 0, (uint16_t *) NULL);
stcb->asoc.last_reset_action[1] = stcb->asoc.last_reset_action[0];
stcb->asoc.last_reset_action[0] = SCTP_STREAM_RESET_PERFORMED;
asoc->str_reset_seq_in++;
} else if (asoc->str_reset_seq_in - 1 == seq) {
sctp_add_stream_reset_result_tsn(chk, seq, asoc->last_reset_action[0],
stcb->asoc.last_sending_seq[0],
stcb->asoc.last_base_tsnsent[0]
);
} else if (asoc->str_reset_seq_in - 2 == seq) {
sctp_add_stream_reset_result_tsn(chk, seq, asoc->last_reset_action[1],
stcb->asoc.last_sending_seq[1],
stcb->asoc.last_base_tsnsent[1]
);
} else {
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_BAD_SEQNO);
}
return (0);
}
static void
sctp_handle_str_reset_request_out(struct sctp_tcb *stcb,
struct sctp_tmit_chunk *chk,
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
struct sctp_stream_reset_out_request *req, int trunc)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
uint32_t seq, tsn;
int number_entries, len;
struct sctp_association *asoc = &stcb->asoc;
seq = ntohl(req->request_seq);
/* now if its not a duplicate we process it */
if (asoc->str_reset_seq_in == seq) {
len = ntohs(req->ph.param_length);
number_entries = ((len - sizeof(struct sctp_stream_reset_out_request)) / sizeof(uint16_t));
/*
* the sender is resetting, handle the list issue.. we must
* a) verify if we can do the reset, if so no problem b) If
* we can't do the reset we must copy the request. c) queue
* it, and setup the data in processor to trigger it off
* when needed and dequeue all the queued data.
*/
tsn = ntohl(req->send_reset_at_tsn);
/* move the reset action back one */
asoc->last_reset_action[1] = asoc->last_reset_action[0];
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
if (trunc) {
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_DENIED);
asoc->last_reset_action[0] = SCTP_STREAM_RESET_DENIED;
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
} else if (SCTP_TSN_GE(asoc->cumulative_tsn, tsn)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* we can do it now */
sctp_reset_in_stream(stcb, number_entries, req->list_of_streams);
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_PERFORMED);
asoc->last_reset_action[0] = SCTP_STREAM_RESET_PERFORMED;
} else {
/*
* we must queue it up and thus wait for the TSN's
* to arrive that are at or before tsn
*/
struct sctp_stream_reset_list *liste;
int siz;
siz = sizeof(struct sctp_stream_reset_list) + (number_entries * sizeof(uint16_t));
SCTP_MALLOC(liste, struct sctp_stream_reset_list *,
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
siz, SCTP_M_STRESET);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (liste == NULL) {
/* gak out of memory */
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_DENIED);
asoc->last_reset_action[0] = SCTP_STREAM_RESET_DENIED;
return;
}
liste->tsn = tsn;
liste->number_entries = number_entries;
memcpy(&liste->req, req,
(sizeof(struct sctp_stream_reset_out_request) + (number_entries * sizeof(uint16_t))));
TAILQ_INSERT_TAIL(&asoc->resetHead, liste, next_resp);
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_PERFORMED);
asoc->last_reset_action[0] = SCTP_STREAM_RESET_PERFORMED;
}
asoc->str_reset_seq_in++;
} else if ((asoc->str_reset_seq_in - 1) == seq) {
/*
* one seq back, just echo back last action since my
* response was lost.
*/
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[0]);
} else if ((asoc->str_reset_seq_in - 2) == seq) {
/*
* two seq back, just echo back last action since my
* response was lost.
*/
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[1]);
} else {
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_BAD_SEQNO);
}
}
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
static void
sctp_handle_str_reset_add_strm(struct sctp_tcb *stcb, struct sctp_tmit_chunk *chk,
struct sctp_stream_reset_add_strm *str_add)
{
/*
* Peer is requesting to add more streams. If its within our
* max-streams we will allow it.
*/
uint16_t num_stream, i;
uint32_t seq;
Fix the add stream feature of strm-reset to really work: - Fix the copy, we can't do a blind copy but must transfer the data from the old to the new. - Fix the ACK processing so we properly stop retransmitting the thing. - Fix it so if we get a retran we will properly reply with the saved response without doing anything. MFC after: 1 month
2009-02-27 20:54:45 +00:00
struct sctp_association *asoc = &stcb->asoc;
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
struct sctp_queued_to_read *ctl, *nctl;
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
/* Get the number. */
seq = ntohl(str_add->request_seq);
num_stream = ntohs(str_add->number_of_streams);
/* Now what would be the new total? */
Fix the add stream feature of strm-reset to really work: - Fix the copy, we can't do a blind copy but must transfer the data from the old to the new. - Fix the ACK processing so we properly stop retransmitting the thing. - Fix it so if we get a retran we will properly reply with the saved response without doing anything. MFC after: 1 month
2009-02-27 20:54:45 +00:00
if (asoc->str_reset_seq_in == seq) {
num_stream += stcb->asoc.streamincnt;
if (num_stream > stcb->asoc.max_inbound_streams) {
/* We must reject it they ask for to many */
denied:
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_DENIED);
stcb->asoc.last_reset_action[1] = stcb->asoc.last_reset_action[0];
stcb->asoc.last_reset_action[0] = SCTP_STREAM_RESET_DENIED;
} else {
/* Ok, we can do that :-) */
struct sctp_stream_in *oldstrm;
/* save off the old */
oldstrm = stcb->asoc.strmin;
SCTP_MALLOC(stcb->asoc.strmin, struct sctp_stream_in *,
(num_stream * sizeof(struct sctp_stream_in)),
SCTP_M_STRMI);
if (stcb->asoc.strmin == NULL) {
stcb->asoc.strmin = oldstrm;
goto denied;
}
/* copy off the old data */
for (i = 0; i < stcb->asoc.streamincnt; i++) {
TAILQ_INIT(&stcb->asoc.strmin[i].inqueue);
stcb->asoc.strmin[i].stream_no = i;
stcb->asoc.strmin[i].last_sequence_delivered = oldstrm[i].last_sequence_delivered;
stcb->asoc.strmin[i].delivery_started = oldstrm[i].delivery_started;
/* now anything on those queues? */
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
TAILQ_FOREACH_SAFE(ctl, &oldstrm[i].inqueue, next, nctl) {
Fix the add stream feature of strm-reset to really work: - Fix the copy, we can't do a blind copy but must transfer the data from the old to the new. - Fix the ACK processing so we properly stop retransmitting the thing. - Fix it so if we get a retran we will properly reply with the saved response without doing anything. MFC after: 1 month
2009-02-27 20:54:45 +00:00
TAILQ_REMOVE(&oldstrm[i].inqueue, ctl, next);
TAILQ_INSERT_TAIL(&stcb->asoc.strmin[i].inqueue, ctl, next);
}
}
/* Init the new streams */
for (i = stcb->asoc.streamincnt; i < num_stream; i++) {
TAILQ_INIT(&stcb->asoc.strmin[i].inqueue);
stcb->asoc.strmin[i].stream_no = i;
stcb->asoc.strmin[i].last_sequence_delivered = 0xffff;
stcb->asoc.strmin[i].delivery_started = 0;
}
SCTP_FREE(oldstrm, SCTP_M_STRMI);
/* update the size */
stcb->asoc.streamincnt = num_stream;
/* Send the ack */
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_PERFORMED);
stcb->asoc.last_reset_action[1] = stcb->asoc.last_reset_action[0];
stcb->asoc.last_reset_action[0] = SCTP_STREAM_RESET_PERFORMED;
sctp_ulp_notify(SCTP_NOTIFY_STR_RESET_INSTREAM_ADD_OK, stcb,
(uint32_t) stcb->asoc.streamincnt, NULL, SCTP_SO_NOT_LOCKED);
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
}
Fix the add stream feature of strm-reset to really work: - Fix the copy, we can't do a blind copy but must transfer the data from the old to the new. - Fix the ACK processing so we properly stop retransmitting the thing. - Fix it so if we get a retran we will properly reply with the saved response without doing anything. MFC after: 1 month
2009-02-27 20:54:45 +00:00
} else if ((asoc->str_reset_seq_in - 1) == seq) {
/*
* one seq back, just echo back last action since my
* response was lost.
*/
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[0]);
} else if ((asoc->str_reset_seq_in - 2) == seq) {
/*
* two seq back, just echo back last action since my
* response was lost.
*/
sctp_add_stream_reset_result(chk, seq, asoc->last_reset_action[1]);
} else {
sctp_add_stream_reset_result(chk, seq, SCTP_STREAM_RESET_BAD_SEQNO);
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
}
}
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
#ifdef __GNUC__
__attribute__((noinline))
#endif
static int
sctp_handle_stream_reset(struct sctp_tcb *stcb, struct mbuf *m, int offset,
struct sctp_stream_reset_out_req *sr_req)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
int chk_length, param_len, ptype;
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
struct sctp_paramhdr pstore;
uint8_t cstore[SCTP_CHUNK_BUFFER_SIZE];
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
uint32_t seq;
int num_req = 0;
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
int trunc = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
struct sctp_tmit_chunk *chk;
struct sctp_chunkhdr *ch;
struct sctp_paramhdr *ph;
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
int ret_code = 0;
int num_param = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* now it may be a reset or a reset-response */
chk_length = ntohs(sr_req->ch.chunk_length);
/* setup for adding the response */
sctp_alloc_a_chunk(stcb, chk);
if (chk == NULL) {
return (ret_code);
}
chk->rec.chunk_id.id = SCTP_STREAM_RESET;
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
chk->rec.chunk_id.can_take_data = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
chk->asoc = &stcb->asoc;
chk->no_fr_allowed = 0;
chk->book_size = chk->send_size = sizeof(struct sctp_chunkhdr);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
chk->book_size_scale = 0;
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
chk->data = sctp_get_mbuf_for_msg(MCLBYTES, 0, M_DONTWAIT, 1, MT_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (chk->data == NULL) {
strres_nochunk:
if (chk->data) {
sctp_m_freem(chk->data);
chk->data = NULL;
}
Fix a locking issue showing up on Mac OS X when subscribing to authentication events. DTLS/SCTP renegotiations trigger the bug. MFC after: 2 weeks.
2011-05-08 09:11:59 +00:00
sctp_free_a_chunk(stcb, chk, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (ret_code);
}
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_RESV_UF(chk->data, SCTP_MIN_OVERHEAD);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* setup chunk parameters */
chk->sent = SCTP_DATAGRAM_UNSENT;
chk->snd_count = 0;
chk->whoTo = stcb->asoc.primary_destination;
atomic_add_int(&chk->whoTo->ref_count, 1);
ch = mtod(chk->data, struct sctp_chunkhdr *);
ch->chunk_type = SCTP_STREAM_RESET;
ch->chunk_flags = 0;
ch->chunk_length = htons(chk->send_size);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_LEN(chk->data) = SCTP_SIZE32(chk->send_size);
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
offset += sizeof(struct sctp_chunkhdr);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
while ((size_t)chk_length >= sizeof(struct sctp_stream_reset_tsn_request)) {
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
ph = (struct sctp_paramhdr *)sctp_m_getptr(m, offset, sizeof(pstore), (uint8_t *) & pstore);
if (ph == NULL)
break;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
param_len = ntohs(ph->param_length);
if (param_len < (int)sizeof(struct sctp_stream_reset_tsn_request)) {
/* bad param */
break;
}
- Fix type casts in calling sctp_m_getptr, it expects a int not an unsigned (returned by sizeof) also add cast to comparison check for size bounds. Approved by: re(bmah@freebsd.org)
2007-06-22 14:40:09 +00:00
ph = (struct sctp_paramhdr *)sctp_m_getptr(m, offset, min(param_len, (int)sizeof(cstore)),
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
(uint8_t *) & cstore);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
ptype = ntohs(ph->param_type);
num_param++;
- Fix type casts in calling sctp_m_getptr, it expects a int not an unsigned (returned by sizeof) also add cast to comparison check for size bounds. Approved by: re(bmah@freebsd.org)
2007-06-22 14:40:09 +00:00
if (param_len > (int)sizeof(cstore)) {
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
trunc = 1;
} else {
trunc = 0;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (num_param > SCTP_MAX_RESET_PARAMS) {
/* hit the max of parameters already sorry.. */
break;
}
if (ptype == SCTP_STR_RESET_OUT_REQUEST) {
struct sctp_stream_reset_out_request *req_out;
req_out = (struct sctp_stream_reset_out_request *)ph;
num_req++;
if (stcb->asoc.stream_reset_outstanding) {
seq = ntohl(req_out->response_seq);
if (seq == stcb->asoc.str_reset_seq_out) {
/* implicit ack */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
(void)sctp_handle_stream_reset_response(stcb, seq, SCTP_STREAM_RESET_PERFORMED, NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
sctp_handle_str_reset_request_out(stcb, chk, req_out, trunc);
Add the add-stream capability. Still needs more testing.. MFC after: 1 month
2009-02-20 15:03:54 +00:00
} else if (ptype == SCTP_STR_RESET_ADD_STREAMS) {
struct sctp_stream_reset_add_strm *str_add;
str_add = (struct sctp_stream_reset_add_strm *)ph;
num_req++;
sctp_handle_str_reset_add_strm(stcb, chk, str_add);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else if (ptype == SCTP_STR_RESET_IN_REQUEST) {
struct sctp_stream_reset_in_request *req_in;
num_req++;
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
req_in = (struct sctp_stream_reset_in_request *)ph;
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
sctp_handle_str_reset_request_in(stcb, chk, req_in, trunc);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else if (ptype == SCTP_STR_RESET_TSN_REQUEST) {
struct sctp_stream_reset_tsn_request *req_tsn;
num_req++;
req_tsn = (struct sctp_stream_reset_tsn_request *)ph;
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (sctp_handle_str_reset_request_tsn(stcb, chk, req_tsn)) {
ret_code = 1;
goto strres_nochunk;
}
/* no more */
break;
} else if (ptype == SCTP_STR_RESET_RESPONSE) {
struct sctp_stream_reset_response *resp;
uint32_t result;
resp = (struct sctp_stream_reset_response *)ph;
seq = ntohl(resp->response_seq);
result = ntohl(resp->result);
if (sctp_handle_stream_reset_response(stcb, seq, result, resp)) {
ret_code = 1;
goto strres_nochunk;
}
} else {
break;
}
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
offset += SCTP_SIZE32(param_len);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
chk_length -= SCTP_SIZE32(param_len);
}
if (num_req == 0) {
/* we have no response free the stuff */
goto strres_nochunk;
}
/* ok we have a chunk to link in */
TAILQ_INSERT_TAIL(&stcb->asoc.control_send_queue,
chk,
sctp_next);
stcb->asoc.ctrl_queue_cnt++;
return (ret_code);
}
/*
* Handle a router or endpoints report of a packet loss, there are two ways
* to handle this, either we get the whole packet and must disect it
* ourselves (possibly with truncation and or corruption) or it is a summary
* from a middle box that did the disectting for us.
*/
static void
sctp_handle_packet_dropped(struct sctp_pktdrop_chunk *cp,
- Issue one, new stack reduction left packet_drop handling still thinking it had the whole chunk. This could cause a crash if a large packet drop came in. Fixed by adjusting the trunc length down to the limit. - Large sacks with lots of segments could also have same issue. Changed duplicate and segment handling to use proper get_m_ptr function to pull each block from mbuf chains.
2007-06-15 17:59:57 +00:00
struct sctp_tcb *stcb, struct sctp_nets *net, uint32_t limit)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
uint32_t bottle_bw, on_queue;
uint16_t trunc_len;
unsigned int chlen;
unsigned int at;
struct sctp_chunk_desc desc;
struct sctp_chunkhdr *ch;
chlen = ntohs(cp->ch.chunk_length);
chlen -= sizeof(struct sctp_pktdrop_chunk);
/* XXX possible chlen underflow */
if (chlen == 0) {
ch = NULL;
if (cp->ch.chunk_flags & SCTP_FROM_MIDDLE_BOX)
SCTP_STAT_INCR(sctps_pdrpbwrpt);
} else {
ch = (struct sctp_chunkhdr *)(cp->data + sizeof(struct sctphdr));
chlen -= sizeof(struct sctphdr);
/* XXX possible chlen underflow */
memset(&desc, 0, sizeof(desc));
}
trunc_len = (uint16_t) ntohs(cp->trunc_len);
- Issue one, new stack reduction left packet_drop handling still thinking it had the whole chunk. This could cause a crash if a large packet drop came in. Fixed by adjusting the trunc length down to the limit. - Large sacks with lots of segments could also have same issue. Changed duplicate and segment handling to use proper get_m_ptr function to pull each block from mbuf chains.
2007-06-15 17:59:57 +00:00
if (trunc_len > limit) {
trunc_len = limit;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* now the chunks themselves */
while ((ch != NULL) && (chlen >= sizeof(struct sctp_chunkhdr))) {
desc.chunk_type = ch->chunk_type;
/* get amount we need to move */
at = ntohs(ch->chunk_length);
if (at < sizeof(struct sctp_chunkhdr)) {
/* corrupt chunk, maybe at the end? */
SCTP_STAT_INCR(sctps_pdrpcrupt);
break;
}
if (trunc_len == 0) {
/* we are supposed to have all of it */
if (at > chlen) {
/* corrupt skip it */
SCTP_STAT_INCR(sctps_pdrpcrupt);
break;
}
} else {
/* is there enough of it left ? */
if (desc.chunk_type == SCTP_DATA) {
if (chlen < (sizeof(struct sctp_data_chunk) +
sizeof(desc.data_bytes))) {
break;
}
} else {
if (chlen < sizeof(struct sctp_chunkhdr)) {
break;
}
}
}
if (desc.chunk_type == SCTP_DATA) {
/* can we get out the tsn? */
if ((cp->ch.chunk_flags & SCTP_FROM_MIDDLE_BOX))
SCTP_STAT_INCR(sctps_pdrpmbda);
if (chlen >= (sizeof(struct sctp_data_chunk) + sizeof(uint32_t))) {
/* yep */
struct sctp_data_chunk *dcp;
uint8_t *ddp;
unsigned int iii;
dcp = (struct sctp_data_chunk *)ch;
ddp = (uint8_t *) (dcp + 1);
for (iii = 0; iii < sizeof(desc.data_bytes); iii++) {
desc.data_bytes[iii] = ddp[iii];
}
desc.tsn_ifany = dcp->dp.tsn;
} else {
/* nope we are done. */
SCTP_STAT_INCR(sctps_pdrpnedat);
break;
}
} else {
if ((cp->ch.chunk_flags & SCTP_FROM_MIDDLE_BOX))
SCTP_STAT_INCR(sctps_pdrpmbct);
}
if (process_chunk_drop(stcb, &desc, net, cp->ch.chunk_flags)) {
SCTP_STAT_INCR(sctps_pdrppdbrk);
break;
}
if (SCTP_SIZE32(at) > chlen) {
break;
}
chlen -= SCTP_SIZE32(at);
if (chlen < sizeof(struct sctp_chunkhdr)) {
/* done, none left */
break;
}
ch = (struct sctp_chunkhdr *)((caddr_t)ch + SCTP_SIZE32(at));
}
/* Now update any rwnd --- possibly */
if ((cp->ch.chunk_flags & SCTP_FROM_MIDDLE_BOX) == 0) {
/* From a peer, we get a rwnd report */
uint32_t a_rwnd;
SCTP_STAT_INCR(sctps_pdrpfehos);
bottle_bw = ntohl(cp->bottle_bw);
on_queue = ntohl(cp->current_onq);
if (bottle_bw && on_queue) {
/* a rwnd report is in here */
if (bottle_bw > on_queue)
a_rwnd = bottle_bw - on_queue;
else
a_rwnd = 0;
if (a_rwnd == 0)
stcb->asoc.peers_rwnd = 0;
else {
if (a_rwnd > stcb->asoc.total_flight) {
stcb->asoc.peers_rwnd =
a_rwnd - stcb->asoc.total_flight;
} else {
stcb->asoc.peers_rwnd = 0;
}
if (stcb->asoc.peers_rwnd <
stcb->sctp_ep->sctp_ep.sctp_sws_sender) {
/* SWS sender side engages */
stcb->asoc.peers_rwnd = 0;
}
}
}
} else {
SCTP_STAT_INCR(sctps_pdrpfmbox);
}
/* now middle boxes in sat networks get a cwnd bump */
if ((cp->ch.chunk_flags & SCTP_FROM_MIDDLE_BOX) &&
(stcb->asoc.sat_t3_loss_recovery == 0) &&
(stcb->asoc.sat_network)) {
/*
* This is debateable but for sat networks it makes sense
* Note if a T3 timer has went off, we will prohibit any
* changes to cwnd until we exit the t3 loss recovery.
*/
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
stcb->asoc.cc_functions.sctp_cwnd_update_after_packet_dropped(stcb,
net, cp, &bottle_bw, &on_queue);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
/*
* handles all control chunks in a packet inputs: - m: mbuf chain, assumed to
* still contain IP/SCTP header - stcb: is the tcb found for this packet -
* offset: offset into the mbuf chain to first chunkhdr - length: is the
* length of the complete packet outputs: - length: modified to remaining
* length after control processing - netp: modified to new sctp_nets after
* cookie-echo processing - return NULL to discard the packet (ie. no asoc,
* bad packet,...) otherwise return the tcb for this packet
*/
Adds gcc attribute to prevent inlining of a function. If it goes inline we may well blow the stack if witness and such are enabled.
2007-05-29 14:17:47 +00:00
#ifdef __GNUC__
__attribute__((noinline))
#endif
static struct sctp_tcb *
sctp_process_control(struct mbuf *m, int iphlen, int *offset, int length,
struct sctphdr *sh, struct sctp_chunkhdr *ch, struct sctp_inpcb *inp,
struct sctp_tcb *stcb, struct sctp_nets **netp, int *fwd_tsn_seen,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
uint32_t vrf_id, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
struct sctp_association *asoc;
uint32_t vtag_in;
int num_chunks = 0; /* number of control chunks processed */
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
uint32_t chk_length;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int ret;
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
int abort_no_unlock = 0;
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
int ecne_seen = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* How big should this be, and should it be alloc'd? Lets try the
* d-mtu-ceiling for now (2k) and that should hopefully work ...
* until we get into jumbo grams and such..
*/
- Copyright updates (aka 2007) - ZONE get now also take a type cast so it does the cast like mtod does. - New macro SCTP_LIST_EMPTY, which in bsd is just LIST_EMPTY - Removal of const in some of the static hmac functions (not needed) - Store length changes to allow for new fields in auth - Auth code updated to current draft (this should be the RFC version we think). - use uint8_t instead of u_char in LOOPBACK address comparison - Some u_int32_t converted to uint32_t (in crc code) - A bug was found in the mib counts for ordered/unordered count, this was fixed (was referencing a freed mbuf). - SCTP_ASOCLOG_OF_TSNS added (code will probably disappear after my testing completes. It allows us to keep a small log on each assoc of the last 40 TSN's in/out and stream assignment. It is NOT in options and so is only good for private builds. - Some CMT changes in prep for Jana fixing his problem with reneging when CMT is enabled (Concurrent Multipath Transfer = CMT). - Some missing mib stats added. - Correction to number of open assoc's count in mib - Correction to os_bsd.h to get right sha2 macros - Add of special AUTH_04 flags so you can compile the code with the old format (in case the peer does not yet support the latest auth code). - Nonce sum was incorrectly being set in when ecn_nonce was NOT on. - LOR in listen with implicit bind found and fixed. - Moved away from using mbuf's for socket options to using just data pointers. The mbufs were used to harmonize NetBSD code since both Net and Open used this method. We have decided to move away from that and more conform to FreeBSD style (which makes more sense). - Very very nasty bug found in some of my "debug" code. The cookie_how collision case tracking had an endless loop in it if you got a second retransmission of a cookie collision case. This would lock up a CPU .. ugly.. - auth function goes to using size_t instead of int which conforms to socketapi better - Found the nasty bug that happens after 9 days of testing.. you get the data chunk, deliver it and due to the reference to a ch-> that every now and then has been deleted (depending on the postion in the mbuf) you have an invalid ch->ch.flags.. and thus you don't advance the stream sequence number.. so you block the stream permanently. The fix is to make local variables of these guys and set them up before you have any chance of trimming the mbuf. - style fix in sctp_util.h, not sure how this got bad maybe in the last patch? (aka it may not be in the real source). - Found interesting bug when using the extended snd/rcv info where we would get an error on receiving with this. Thats because it was NOT padded to the same size as the snd_rcv info. We increase (add the pad) so the two structs are the same size in sctp_uio.h - In sctp_usrreq.c one of the most common things we did for socket options was to cast the pointer and validate the size. This as been macro-ized to help make the code more readable. - in sctputil.c two things, the socketapi class found a missing flag type (the next msg is a notification) and a missing scope recovery was also fixed. Reviewed by: gnn
2007-02-12 23:24:31 +00:00
uint8_t chunk_buf[SCTP_CHUNK_BUFFER_SIZE];
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
struct sctp_tcb *locked_tcb = stcb;
int got_auth = 0;
uint32_t auth_offset = 0, auth_len = 0;
int auth_skipped = 0;
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
int asconf_cnt = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
struct socket *so;
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "sctp_process_control: iphlen=%u, offset=%u, length=%u stcb:%p\n",
iphlen, *offset, length, stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* validate chunk header length... */
if (ntohs(ch->chunk_length) < sizeof(*ch)) {
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Invalid header length %d\n",
ntohs(ch->chunk_length));
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
/*
* validate the verification tag
*/
vtag_in = ntohl(sh->v_tag);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
if (locked_tcb) {
SCTP_TCB_LOCK_ASSERT(locked_tcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ch->chunk_type == SCTP_INITIATION) {
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Its an INIT of len:%d vtag:%x\n",
ntohs(ch->chunk_length), vtag_in);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (vtag_in != 0) {
/* protocol error- silently discard... */
SCTP_STAT_INCR(sctps_badvtag);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
} else if (ch->chunk_type != SCTP_COOKIE_ECHO) {
/*
* If there is no stcb, skip the AUTH chunk and process
* later after a stcb is found (to validate the lookup was
* valid.
*/
if ((ch->chunk_type == SCTP_AUTHENTICATION) &&
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
(stcb == NULL) &&
!SCTP_BASE_SYSCTL(sctp_auth_disable)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* save this chunk for later processing */
auth_skipped = 1;
auth_offset = *offset;
auth_len = ntohs(ch->chunk_length);
/* (temporarily) move past this chunk */
*offset += SCTP_SIZE32(auth_len);
if (*offset >= length) {
/* no more data left in the mbuf chain */
*offset = length;
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, *offset,
sizeof(struct sctp_chunkhdr), chunk_buf);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (ch == NULL) {
/* Help */
*offset = length;
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ch->chunk_type == SCTP_COOKIE_ECHO) {
goto process_control_chunks;
}
/*
* first check if it's an ASCONF with an unknown src addr we
* need to look inside to find the association
*/
if (ch->chunk_type == SCTP_ASCONF && stcb == NULL) {
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
struct sctp_chunkhdr *asconf_ch = ch;
uint32_t asconf_offset = 0, asconf_len = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* inp's refcount may be reduced */
SCTP_INP_INCR_REF(inp);
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
asconf_offset = *offset;
do {
asconf_len = ntohs(asconf_ch->chunk_length);
if (asconf_len < sizeof(struct sctp_asconf_paramhdr))
break;
stcb = sctp_findassociation_ep_asconf(m, iphlen,
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
*offset, sh, &inp, netp, vrf_id);
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if (stcb != NULL)
break;
asconf_offset += SCTP_SIZE32(asconf_len);
asconf_ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, asconf_offset,
sizeof(struct sctp_chunkhdr), chunk_buf);
} while (asconf_ch != NULL && asconf_ch->chunk_type == SCTP_ASCONF);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb == NULL) {
/*
* reduce inp's refcount if not reduced in
* sctp_findassociation_ep_asconf().
*/
SCTP_INP_DECR_REF(inp);
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
} else {
locked_tcb = stcb;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* now go back and verify any auth chunk to be sure */
if (auth_skipped && (stcb != NULL)) {
struct sctp_auth_chunk *auth;
auth = (struct sctp_auth_chunk *)
sctp_m_getptr(m, auth_offset,
auth_len, chunk_buf);
got_auth = 1;
auth_skipped = 0;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((auth == NULL) || sctp_handle_auth(stcb, auth, m,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
auth_offset)) {
/* auth HMAC failed so dump it */
*offset = length;
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
} else {
/* remaining chunks are HMAC checked */
stcb->asoc.authenticated = 1;
}
}
}
if (stcb == NULL) {
/* no association, so it's out of the blue... */
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_handle_ootb(m, iphlen, *offset, sh, inp, NULL,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
asoc = &stcb->asoc;
/* ABORT and SHUTDOWN can use either v_tag... */
if ((ch->chunk_type == SCTP_ABORT_ASSOCIATION) ||
(ch->chunk_type == SCTP_SHUTDOWN_COMPLETE) ||
(ch->chunk_type == SCTP_PACKET_DROPPED)) {
if ((vtag_in == asoc->my_vtag) ||
((ch->chunk_flags & SCTP_HAD_NO_TCB) &&
(vtag_in == asoc->peer_vtag))) {
/* this is valid */
} else {
/* drop this packet... */
SCTP_STAT_INCR(sctps_badvtag);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
} else if (ch->chunk_type == SCTP_SHUTDOWN_ACK) {
if (vtag_in != asoc->my_vtag) {
/*
* this could be a stale SHUTDOWN-ACK or the
* peer never got the SHUTDOWN-COMPLETE and
* is still hung; we have started a new asoc
* but it won't complete until the shutdown
* is completed
*/
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_handle_ootb(m, iphlen, *offset, sh, inp,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
NULL, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
} else {
/* for all other chunks, vtag must match */
if (vtag_in != asoc->my_vtag) {
/* invalid vtag... */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3,
"invalid vtag: %xh, expect %xh\n",
vtag_in, asoc->my_vtag);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR(sctps_badvtag);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
}
} /* end if !SCTP_COOKIE_ECHO */
/*
* process all control chunks...
*/
if (((ch->chunk_type == SCTP_SELECTIVE_ACK) ||
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/* EY */
(ch->chunk_type == SCTP_NR_SELECTIVE_ACK) ||
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(ch->chunk_type == SCTP_HEARTBEAT_REQUEST)) &&
(SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_COOKIE_ECHOED)) {
/* implied cookie-ack.. we must have lost the ack */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_cookie_ack((struct sctp_cookie_ack_chunk *)ch, stcb,
*netp);
}
process_control_chunks:
while (IS_SCTP_CONTROL(ch)) {
/* validate chunk length */
chk_length = ntohs(ch->chunk_length);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT2, "sctp_process_control: processing a chunk type=%u, len=%u\n",
ch->chunk_type, chk_length);
- Fix so ifn's are properly deleted when the ref count goes to 0. - Fix so VRF's will clean themselves up when no references are around. - Allow sctp_ifa to be passed into inpcb_bind, addr_mgmt_ep_sa to bypass normal validation checks. - turn auto-asconf off for subset bound sockets - Moves all logging to use KTR. This gets rid of most of the logging #ifdef's with a few exceptions reducing the number of config options for SCTP.
2007-06-14 22:59:04 +00:00
SCTP_LTRACE_CHK(inp, stcb, ch->chunk_type, chk_length);
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
if (chk_length < sizeof(*ch) ||
(*offset + (int)chk_length) > length) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
SCTP_STAT_INCR_COUNTER64(sctps_incontrolchunks);
/*
* INIT-ACK only gets the init ack "header" portion only
* because we don't have to process the peer's COOKIE. All
* others get a complete chunk.
*/
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if ((ch->chunk_type == SCTP_INITIATION_ACK) ||
(ch->chunk_type == SCTP_INITIATION)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* get an init-ack chunk */
ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, *offset,
sizeof(struct sctp_init_ack_chunk), chunk_buf);
if (ch == NULL) {
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
} else {
- Update the comment lines in sctp_input.c - We need to init the INP_LOCK since otherwise for non-SMP kernels you crash when you set the TOS.
2007-06-15 19:28:58 +00:00
/* For cookies and all other chunks. */
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if (chk_length > sizeof(chunk_buf)) {
/*
* use just the size of the chunk buffer so
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
* the front part of our chunks fit in
* contiguous space up to the chunk buffer
* size (508 bytes). For chunks that need to
- Update the comment lines in sctp_input.c - We need to init the INP_LOCK since otherwise for non-SMP kernels you crash when you set the TOS.
2007-06-15 19:28:58 +00:00
* get more than that they must use the
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
* sctp_m_getptr() function or other means
- Update the comment lines in sctp_input.c - We need to init the INP_LOCK since otherwise for non-SMP kernels you crash when you set the TOS.
2007-06-15 19:28:58 +00:00
* (e.g. know how to parse mbuf chains).
* Cookies do this already.
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*/
ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, *offset,
(sizeof(chunk_buf) - 4),
chunk_buf);
if (ch == NULL) {
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
return (NULL);
}
} else {
/* We can fit it all */
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, *offset,
chk_length, chunk_buf);
if (ch == NULL) {
SCTP_PRINTF("sctp_process_control: Can't get the all data....\n");
*offset = length;
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
}
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
return (NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
}
num_chunks++;
/* Save off the last place we got a control from */
if (stcb != NULL) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (((netp != NULL) && (*netp != NULL)) || (ch->chunk_type == SCTP_ASCONF)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* allow last_control to be NULL if
* ASCONF... ASCONF processing will find the
* right net later
*/
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((netp != NULL) && (*netp != NULL))
stcb->asoc.last_control_chunk_from = *netp;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
#ifdef SCTP_AUDITING_ENABLED
sctp_audit_log(0xB0, ch->chunk_type);
#endif
/* check to see if this chunk required auth, but isn't */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if ((stcb != NULL) &&
!SCTP_BASE_SYSCTL(sctp_auth_disable) &&
sctp_auth_is_required_chunk(ch->chunk_type, stcb->asoc.local_auth_chunks) &&
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
!stcb->asoc.authenticated) {
/* "silently" ignore */
SCTP_STAT_INCR(sctps_recvauthmissing);
goto next_chunk;
}
switch (ch->chunk_type) {
case SCTP_INITIATION:
/* must be first and only chunk */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_INIT\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/* We are not interested anymore? */
if ((stcb) && (stcb->asoc.total_output_queue_size)) {
/*
* collision case where we are
* sending to them too
*/
;
} else {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if ((chk_length > SCTP_LARGEST_INIT_ACCEPTED) ||
(num_chunks > 1) ||
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
(SCTP_BASE_SYSCTL(sctp_strict_init) && (length - *offset > (int)SCTP_SIZE32(chk_length)))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
if ((stcb != NULL) &&
(SCTP_GET_STATE(&stcb->asoc) ==
SCTP_STATE_SHUTDOWN_ACK_SENT)) {
sctp_send_shutdown_ack(stcb,
stcb->asoc.primary_destination);
*offset = length;
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_chunk_output(inp, stcb, SCTP_OUTPUT_FROM_CONTROL_PROC, SCTP_SO_NOT_LOCKED);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (netp) {
sctp_handle_init(m, iphlen, *offset, sh,
(struct sctp_init_chunk *)ch, inp,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
stcb, *netp, &abort_no_unlock, vrf_id, port);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (abort_no_unlock)
return (NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
break;
- Fixed cookie handling to calc an RTO when its an INIT collision case. - Fixed RTO calc to maintain a seperate variable to track if a RTO calc as been done, this allows the RTO var to be doubled during initial timeouts. - Reduces the amount of stack used by process control. - Use a constant for the peer chunk overhead. - Name change to spell candidate correctly.
2007-06-13 01:31:53 +00:00
case SCTP_PAD_CHUNK:
break;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_INITIATION_ACK:
/* must be first and only chunk */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_INIT-ACK\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/* We are not interested anymore */
if ((stcb) && (stcb->asoc.total_output_queue_size)) {
;
} else {
1) Further enhance the INVARIANT lock validation (no locks) are held by checking the create and inp locks as well. 2) Fix a bug in that when a socket is closed an INIT-ACK is returned, we do NOT unlock the locked_tcb unless its different (an unlikely scenario). If we blindly unlock as we were doing before we can end up unlocking the actual stcb thats about to be sent down to the free function which requires the lock be held. MFC after: 1 week
2010-06-06 16:11:16 +00:00
if (locked_tcb != stcb) {
/* Very unlikely */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
if (stcb) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(inp);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_27);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
return (NULL);
}
}
if ((num_chunks > 1) ||
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
(SCTP_BASE_SYSCTL(sctp_strict_init) && (length - *offset > (int)SCTP_SIZE32(chk_length)))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((netp) && (*netp)) {
ret = sctp_handle_init_ack(m, iphlen, *offset, sh,
- Take out the broken table-id concept. Panda Routers have a M-VRF concept that is NOT well thought out for a multi-homed transport protocol. So the useless table-id entries passed around need to be removed. - Add a event timer for the zero copy api. - Fix a bug in sctp_timer.c when searching for an alternate with the largest ssthresh (the compare was wrong).
2007-06-01 11:19:54 +00:00
(struct sctp_init_ack_chunk *)ch, stcb, *netp, &abort_no_unlock, vrf_id);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
} else {
ret = -1;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* Special case, I must call the output routine to
* get the cookie echoed
*/
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
if (abort_no_unlock)
return (NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((stcb) && ret == 0)
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_chunk_output(stcb->sctp_ep, stcb, SCTP_OUTPUT_FROM_CONTROL_PROC, SCTP_SO_NOT_LOCKED);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
break;
case SCTP_SELECTIVE_ACK:
{
struct sctp_sack_chunk *sack;
int abort_now = 0;
uint32_t a_rwnd, cum_ack;
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
uint16_t num_seg, num_dup;
uint8_t flags;
int offset_seg, offset_dup;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Fix the switching on/off of CMT using sysctl and socket option. Fix the switching on/off of PF and NR-SACKs using sysctl. Add minor improvement in handling malloc failures. Improve the address checks when sending. MFC after: 4 weeks
2010-08-28 17:59:51 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_SACK\n");
SCTP_STAT_INCR(sctps_recvsacks);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (stcb == NULL) {
SCTPDBG(SCTP_DEBUG_INDATA1, "No stcb when processing SACK chunk\n");
break;
}
if (chk_length < sizeof(struct sctp_sack_chunk)) {
SCTPDBG(SCTP_DEBUG_INDATA1, "Bad size on SACK chunk, too small\n");
break;
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if (SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_SHUTDOWN_ACK_SENT) {
/*-
* If we have sent a shutdown-ack, we will pay no
* attention to a sack sent in to us since
* we don't care anymore.
*/
-Improvement: Add '\n' on debug output in sctp_lower_sosend(). -Improvement: panic() on INVARIANTS kernels if memory allocation fails for a tagblock in sctp_add_vtag_to_timewait(). -Bugfix: Protect code in sctp_is_in_timewait() by SCTP_INP_INFO_WLOCK/SCTP_INP_INFO_WUNLOCK. -Cleanup: Get rid of unused variable now in sctp_init_asoc(). -Bugfix: Reuse the correct vtag in sctp_add_vtag_to_timewait(). -Cleanup: Get rid of unused constant SCTP_TIME_WAIT_SHORT in sctp_constants.h. -Improvement: Use all hash buckets of the vtag hash table. -Cleanup: Get rid of then unused constant SCTP_STACK_VTAG_HASH_SIZE_A. -Bugfix: Handle SHUTDOWN;SACK packet correctly. -Bugfix: Last TSN in a gap ack block was not being "ack'd" in the internal scoreboard. Obtained from: (with help from Michael Tuexen)
2008-11-12 14:16:39 +00:00
break;
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sack = (struct sctp_sack_chunk *)ch;
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
flags = ch->chunk_flags;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
cum_ack = ntohl(sack->sack.cum_tsn_ack);
num_seg = ntohs(sack->sack.num_gap_ack_blks);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
num_dup = ntohs(sack->sack.num_dup_tsns);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
a_rwnd = (uint32_t) ntohl(sack->sack.a_rwnd);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (sizeof(struct sctp_sack_chunk) +
num_seg * sizeof(struct sctp_gap_ack_block) +
num_dup * sizeof(uint32_t) != chk_length) {
SCTPDBG(SCTP_DEBUG_INDATA1, "Bad size of SACK chunk\n");
break;
}
offset_seg = *offset + sizeof(struct sctp_sack_chunk);
offset_dup = offset_seg + num_seg * sizeof(struct sctp_gap_ack_block);
- Restructure so bindx functions are not done inline to socket option but are a seperate call that can be re-used if needed. - 64 bit issues o re-arrange cookie so it is better 64 bit aligned o For wire level things we need the packed attribute.
2007-06-12 11:21:00 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_SACK process cum_ack:%x num_seg:%d a_rwnd:%d\n",
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
cum_ack, num_seg, a_rwnd);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.seen_a_sack_this_pkt = 1;
if ((stcb->asoc.pr_sctp_cnt == 0) &&
(num_seg == 0) &&
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
SCTP_TSN_GE(cum_ack, stcb->asoc.last_acked_seq) &&
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(stcb->asoc.saw_sack_with_frags == 0) &&
Fix a bug where also the number of non-renegable gap reports was considered to be potentially renegable. MFC after: 1 day.
2010-12-04 19:29:49 +00:00
(stcb->asoc.saw_sack_with_nr_frags == 0) &&
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
(!TAILQ_EMPTY(&stcb->asoc.sent_queue))
) {
/*
* We have a SIMPLE sack having no
* prior segments and data on sent
* queue to be acked.. Use the
* faster path sack processing. We
* also allow window update sacks
* with no missing segments to go
* this way too.
*/
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
sctp_express_handle_sack(stcb, cum_ack, a_rwnd, &abort_now, ecne_seen);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
} else {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (netp && *netp)
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
sctp_handle_sack(m, offset_seg, offset_dup,
stcb, *netp,
num_seg, 0, num_dup, &abort_now, flags,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
cum_ack, a_rwnd, ecne_seen);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_now) {
/* ABORT signal from sack processing */
*offset = length;
return (NULL);
}
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (TAILQ_EMPTY(&stcb->asoc.send_queue) &&
TAILQ_EMPTY(&stcb->asoc.sent_queue) &&
(stcb->asoc.stream_queue_cnt == 0)) {
sctp_ulp_notify(SCTP_NOTIFY_SENDER_DRY, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
break;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/*
* EY - nr_sack: If the received chunk is an
* nr_sack chunk
*/
case SCTP_NR_SELECTIVE_ACK:
{
struct sctp_nr_sack_chunk *nr_sack;
int abort_now = 0;
uint32_t a_rwnd, cum_ack;
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
uint16_t num_seg, num_nr_seg, num_dup;
uint8_t flags;
int offset_seg, offset_dup;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
Fix the switching on/off of CMT using sysctl and socket option. Fix the switching on/off of PF and NR-SACKs using sysctl. Add minor improvement in handling malloc failures. Improve the address checks when sending. MFC after: 4 weeks
2010-08-28 17:59:51 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_NR_SACK\n");
SCTP_STAT_INCR(sctps_recvsacks);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (stcb == NULL) {
SCTPDBG(SCTP_DEBUG_INDATA1, "No stcb when processing NR-SACK chunk\n");
break;
}
Fix some CLANG warnings. One clang warning is left due to the fact that its bogus.. nam->sa_family will not change from AF_INET6 to AF_INET (but clang thinks it does ;-D)
2010-09-05 13:41:45 +00:00
if ((stcb->asoc.sctp_nr_sack_on_off == 0) ||
(stcb->asoc.peer_supports_nr_sack == 0)) {
goto unknown_chunk;
}
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (chk_length < sizeof(struct sctp_nr_sack_chunk)) {
SCTPDBG(SCTP_DEBUG_INDATA1, "Bad size on NR-SACK chunk, too small\n");
break;
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
if (SCTP_GET_STATE(&stcb->asoc) == SCTP_STATE_SHUTDOWN_ACK_SENT) {
/*-
* If we have sent a shutdown-ack, we will pay no
* attention to a sack sent in to us since
* we don't care anymore.
*/
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
break;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
}
nr_sack = (struct sctp_nr_sack_chunk *)ch;
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
flags = ch->chunk_flags;
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
cum_ack = ntohl(nr_sack->nr_sack.cum_tsn_ack);
num_seg = ntohs(nr_sack->nr_sack.num_gap_ack_blks);
num_nr_seg = ntohs(nr_sack->nr_sack.num_nr_gap_ack_blks);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
num_dup = ntohs(nr_sack->nr_sack.num_dup_tsns);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
a_rwnd = (uint32_t) ntohl(nr_sack->nr_sack.a_rwnd);
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (sizeof(struct sctp_nr_sack_chunk) +
(num_seg + num_nr_seg) * sizeof(struct sctp_gap_ack_block) +
num_dup * sizeof(uint32_t) != chk_length) {
SCTPDBG(SCTP_DEBUG_INDATA1, "Bad size of NR_SACK chunk\n");
break;
}
offset_seg = *offset + sizeof(struct sctp_nr_sack_chunk);
offset_dup = offset_seg + num_seg * sizeof(struct sctp_gap_ack_block);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_NR_SACK process cum_ack:%x num_seg:%d a_rwnd:%d\n",
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
cum_ack, num_seg, a_rwnd);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
stcb->asoc.seen_a_sack_this_pkt = 1;
if ((stcb->asoc.pr_sctp_cnt == 0) &&
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
(num_seg == 0) && (num_nr_seg == 0) &&
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
SCTP_TSN_GE(cum_ack, stcb->asoc.last_acked_seq) &&
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
(stcb->asoc.saw_sack_with_frags == 0) &&
Fix a bug where also the number of non-renegable gap reports was considered to be potentially renegable. MFC after: 1 day.
2010-12-04 19:29:49 +00:00
(stcb->asoc.saw_sack_with_nr_frags == 0) &&
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
(!TAILQ_EMPTY(&stcb->asoc.sent_queue))) {
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
/*
* We have a SIMPLE sack having no
* prior segments and data on sent
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
* queue to be acked. Use the faster
* path sack processing. We also
* allow window update sacks with no
* missing segments to go this way
* too.
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
*/
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
sctp_express_handle_sack(stcb, cum_ack, a_rwnd,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
&abort_now, ecne_seen);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
} else {
if (netp && *netp)
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
sctp_handle_sack(m, offset_seg, offset_dup,
stcb, *netp,
num_seg, num_nr_seg, num_dup, &abort_now, flags,
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
cum_ack, a_rwnd, ecne_seen);
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
}
if (abort_now) {
/* ABORT signal from sack processing */
*offset = length;
return (NULL);
}
Get rid of a lot of duplicated code for NR-SACK handle. Generalize the SACK to code handle also NR-SACKs.
2010-01-17 21:00:28 +00:00
if (TAILQ_EMPTY(&stcb->asoc.send_queue) &&
TAILQ_EMPTY(&stcb->asoc.sent_queue) &&
(stcb->asoc.stream_queue_cnt == 0)) {
sctp_ulp_notify(SCTP_NOTIFY_SENDER_DRY, stcb, 0, NULL, SCTP_SO_NOT_LOCKED);
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
}
break;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_HEARTBEAT_REQUEST:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_HEARTBEAT\n");
if ((stcb) && netp && *netp) {
SCTP_STAT_INCR(sctps_recvheartbeat);
sctp_send_heartbeat_ack(stcb, m, *offset,
chk_length, *netp);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
/* He's alive so give him credit */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
stcb->asoc.overall_error_count = 0;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_HEARTBEAT_ACK:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_HEARTBEAT-ACK\n");
if ((stcb == NULL) || (chk_length != sizeof(struct sctp_heartbeat_chunk))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
/* Its not ours */
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* He's alive so give him credit */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
SCTP_STAT_INCR(sctps_recvheartbeatack);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (netp && *netp)
sctp_handle_heartbeat_ack((struct sctp_heartbeat_chunk *)ch,
stcb, *netp);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_ABORT_ASSOCIATION:
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_ABORT, stcb %p\n",
stcb);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb) && netp && *netp)
sctp_handle_abort((struct sctp_abort_chunk *)ch,
stcb, *netp);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
break;
case SCTP_SHUTDOWN:
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_SHUTDOWN, stcb %p\n",
stcb);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb == NULL) || (chk_length != sizeof(struct sctp_shutdown_chunk))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
return (NULL);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
if (netp && *netp) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int abort_flag = 0;
sctp_handle_shutdown((struct sctp_shutdown_chunk *)ch,
stcb, *netp, &abort_flag);
if (abort_flag) {
*offset = length;
return (NULL);
}
}
break;
case SCTP_SHUTDOWN_ACK:
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_SHUTDOWN-ACK, stcb %p\n", stcb);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb) && (netp) && (*netp))
sctp_handle_shutdown_ack((struct sctp_shutdown_ack_chunk *)ch, stcb, *netp);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
break;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
case SCTP_OPERATION_ERROR:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_OP-ERR\n");
if ((stcb) && netp && *netp && sctp_handle_error(ch, stcb, *netp) < 0) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
break;
case SCTP_COOKIE_ECHO:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3,
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
"SCTP_COOKIE-ECHO, stcb %p\n", stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((stcb) && (stcb->asoc.total_output_queue_size)) {
;
} else {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* We are not interested anymore */
Ok, yet another bug in killing off all the hundreds of apitesters.. Basically we end up with attempting to destroy a lock thats contended on. A cookie echo arrives at the same time that the close is happening. The close gets the lock but the cookie echo has already passed the check for the gone flag and is then locked waiting on the create lock.. when we go to destroy it bam. For now we do the timer destroy for all calls to close.. We can probably optimize this later so that we check whats being contended on and if there is contention then do the timer thing. but this is probably safest since the inp has been removed from all lists and references and only the timer can find it.. once the locks are released all other places will instantly see the GONE flag and bail (thats what the change in sctp_input is one place that was lacking the bail code). MFC after: 1 week
2010-06-06 19:24:32 +00:00
abend:
This does two changes: 1) Makes it so that the INVARIANT function validate nolocks is available anywhere. 2) Fixes a BUG where a close has been done on a collision socket and the cookie processing would return leaving a lock held. MFC after: 1 week
2010-06-05 21:20:28 +00:00
if (stcb) {
SCTP_TCB_UNLOCK(stcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
}
/*
* First are we accepting? We do this again here
This fixes a bug with the one-2-one model socket when a user sets up a socket to a server sends data and closes the socket before the server has called accept(). It used to NOT work at all. Now we add a flag to the assoc and defer assoc cleanup so that the accept will suceed.
2010-05-11 17:02:29 +00:00
* since it is possible that a previous endpoint WAS
* listening responded to a INIT-ACK and then
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
* closed. We opened and bound.. and are now no
* longer listening.
*/
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
if ((stcb == NULL) && (inp->sctp_socket->so_qlen >= inp->sctp_socket->so_qlimit)) {
if ((inp->sctp_flags & SCTP_PCB_FLAGS_TCPTYPE) &&
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
(SCTP_BASE_SYSCTL(sctp_abort_if_one_2_one_hits_limit))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
struct mbuf *oper;
struct sctp_paramhdr *phdr;
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
oper = sctp_get_mbuf_for_msg(sizeof(struct sctp_paramhdr),
0, M_DONTWAIT, 1, MT_DATA);
if (oper) {
SCTP_BUF_LEN(oper) =
sizeof(struct sctp_paramhdr);
phdr = mtod(oper,
struct sctp_paramhdr *);
phdr->param_type =
htons(SCTP_CAUSE_OUT_OF_RESC);
phdr->param_length =
htons(sizeof(struct sctp_paramhdr));
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
sctp_abort_association(inp, stcb, m,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
iphlen, sh, oper, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- fix sctp_ifn initial refcount issue (prevents deletion) - fix a bug during cookie collision that prevented an association from coming up in a specific restart case. - Fix it so the shutdown-pending flag gets removed (this is more for correctness then needed) when we enter shutdown-sent or shutdown-ack-sent states. - Fix a bug that caused the receiver to sometimes NOT send a SACK when a duplicate TSN arrived. Without this fix it was possible for the association to fall down if the - Deleted primary destination is also stored when SCTP_MOBILITY_BASE. (Previously, it is stored when only SCTP_MOBILITY_FASTHANDOFF) - Fix a locking issue where we might call send_initiate_ack() and incorrectly state the lock held/not held. Also fix it so that when we release the lock the inp cannot be deleted on us. - Add the debug option that can cause the stack to panic instead of aborting an assoc. This does not and should never show up in options but is useful for debugging unexpected aborts. - Add cumack_log sent to track sending cumack information for the debug case where we are running a special log per assoc. - Added extra () aroudn sctp_sbspace macro to avoid compile warnings. MFC after: 1 week
2007-10-16 14:05:51 +00:00
*offset = length;
return (NULL);
} else {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
struct mbuf *ret_buf;
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
struct sctp_inpcb *linp;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (stcb) {
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
linp = NULL;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
} else {
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
linp = inp;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (linp) {
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
SCTP_ASOC_CREATE_LOCK(linp);
Ok, yet another bug in killing off all the hundreds of apitesters.. Basically we end up with attempting to destroy a lock thats contended on. A cookie echo arrives at the same time that the close is happening. The close gets the lock but the cookie echo has already passed the check for the gone flag and is then locked waiting on the create lock.. when we go to destroy it bam. For now we do the timer destroy for all calls to close.. We can probably optimize this later so that we check whats being contended on and if there is contention then do the timer thing. but this is probably safest since the inp has been removed from all lists and references and only the timer can find it.. once the locks are released all other places will instantly see the GONE flag and bail (thats what the change in sctp_input is one place that was lacking the bail code). MFC after: 1 week
2010-06-06 19:24:32 +00:00
if ((inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) ||
(inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_ALLGONE)) {
SCTP_ASOC_CREATE_UNLOCK(linp);
goto abend;
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
if (netp) {
ret_buf =
sctp_handle_cookie_echo(m, iphlen,
*offset, sh,
(struct sctp_cookie_echo_chunk *)ch,
&inp, &stcb, netp,
auth_skipped,
auth_offset,
auth_len,
&locked_tcb,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id,
port);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
} else {
ret_buf = NULL;
}
if (linp) {
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
SCTP_ASOC_CREATE_UNLOCK(linp);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ret_buf == NULL) {
if (locked_tcb) {
SCTP_TCB_UNLOCK(locked_tcb);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3,
"GAK, null buffer\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
auth_skipped = 0;
*offset = length;
return (NULL);
}
/* if AUTH skipped, see if it verified... */
if (auth_skipped) {
got_auth = 1;
auth_skipped = 0;
}
if (!TAILQ_EMPTY(&stcb->asoc.sent_queue)) {
/*
* Restart the timer if we have
* pending data
*/
struct sctp_tmit_chunk *chk;
chk = TAILQ_FIRST(&stcb->asoc.sent_queue);
Code cleanup: Use LIST_FOREACH, LIST_FOREACH_SAFE, TAILQ_FOREACH, TAILQ_FOREACH_SAFE where appropriate. No functional change. MFC after: 3 months.
2010-12-30 16:56:20 +00:00
sctp_timer_start(SCTP_TIMER_TYPE_SEND, stcb->sctp_ep, stcb, chk->whoTo);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
break;
case SCTP_COOKIE_ACK:
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_COOKIE-ACK, stcb %p\n", stcb);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb == NULL) || chk_length != sizeof(struct sctp_cookie_ack_chunk)) {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/* We are not interested anymore */
if ((stcb) && (stcb->asoc.total_output_queue_size)) {
;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
} else if (stcb) {
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(inp);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_27);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
}
/* He's alive so give him credit */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb) && netp && *netp) {
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_cookie_ack((struct sctp_cookie_ack_chunk *)ch, stcb, *netp);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_ECN_ECHO:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_ECN-ECHO\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* He's alive so give him credit */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb == NULL) || (chk_length != sizeof(struct sctp_ecne_chunk))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (stcb) {
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_ecn_echo((struct sctp_ecne_chunk *)ch,
stcb);
1) Allow a chunk to track the cwnd it was at when sent. 2) Add separate max-bursts for retransmit and hb. These are set to sysctlable values but not settable via the socket api. This makes sure we don't blast out HB's or fast-retransmits. 3) Determine on the first data transmission on a net if its local-lan (by being under or over a RTT). This can later be used to think about different algorithms based on locallan vs big-i (experimental) 4) The cwnd should NOT be allowed to grow when an ECNEcho is seen (TCP has this same bug). We fix this in SCTP so an ECNe being seen prevents an advance of cwnd. 5) CWR's should not be sent multiple times to the same network, instead just updating the TSN being transmitted if needed. MFC after: 1 Month
2011-02-02 11:13:23 +00:00
ecne_seen = 1;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_ECN_CWR:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_ECN-CWR\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* He's alive so give him credit */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb == NULL) || (chk_length != sizeof(struct sctp_cwr_chunk))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (stcb) {
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
stcb->asoc.overall_error_count = 0;
Fixes to ECN in SCTP. 1) ECN was on an association basis, this is incorrect and will not work with CMT or for that matter if the user is sending to multiple addresses. This commit makes ECN on a per path basis. 2) Adopt the new format for the ECN internet draft. This also maintains compatability with old format chunks as well. 3) Keep track of the real time of a RTT down to micro seconds. For some future conditional features (for like a data center this is good information to have). MFC after: 1 month
2011-01-29 19:55:29 +00:00
sctp_handle_ecn_cwr((struct sctp_cwr_chunk *)ch, stcb, *netp);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_SHUTDOWN_COMPLETE:
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_SHUTDOWN-COMPLETE, stcb %p\n", stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* must be first and only chunk */
if ((num_chunks > 1) ||
- Fixed (Apple) compiler warnings in sctp_input.c, sctputil.c, sctp_output.c - Fixed a LOR in handling a cookie. Turns out create lock is applied. And if we abort processing, this causes LOR. Changed to force the timer to clean up, that way create lock is released.
2007-05-30 22:34:21 +00:00
(length - *offset > (int)SCTP_SIZE32(chk_length))) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb) && netp && *netp) {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
sctp_handle_shutdown_complete((struct sctp_shutdown_complete_chunk *)ch,
stcb, *netp);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
break;
case SCTP_ASCONF:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_ASCONF\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* He's alive so give him credit */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (stcb) {
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_asconf(m, *offset,
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
(struct sctp_asconf_chunk *)ch, stcb, asconf_cnt == 0);
asconf_cnt++;
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_ASCONF_ACK:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_ASCONF-ACK\n");
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if (chk_length < sizeof(struct sctp_asconf_ack_chunk)) {
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((stcb) && netp && *netp) {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
/* He's alive so give him credit */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_asconf_ack(m, *offset,
- Get rid of unsused constants for sysctl variables. - Fix panic from mutex unlock on freed lock when ASCONF-ACK aborts an assoc - Fix panic from addr lock recursion when ASCONFs are queued in the front states - ASCONFs "queued" in the front states should really be bundled after the COOKIE-ACK, not in front of it - Fix issue with addresses deleted in the front states from being sent with ASCONF(DELETE)-- replaced sctp_asconf_queue_add_sa() with delete specific function - Comment change in sctp.h the drafts are now RFC's Approved by: re@freebsd.org (B Mah)
2007-09-15 19:07:42 +00:00
(struct sctp_asconf_ack_chunk *)ch, stcb, *netp, &abort_no_unlock);
if (abort_no_unlock)
return (NULL);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_FORWARD_CUM_TSN:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_FWD-TSN\n");
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if (chk_length < sizeof(struct sctp_forward_tsn_chunk)) {
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* He's alive so give him credit */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (stcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int abort_flag = 0;
stcb->asoc.overall_error_count = 0;
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*fwd_tsn_seen = 1;
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/* We are not interested anymore */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(inp);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_29);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
sctp_handle_forward_tsn(stcb,
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
(struct sctp_forward_tsn_chunk *)ch, &abort_flag, m, *offset);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_flag) {
*offset = length;
return (NULL);
} else {
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
}
}
break;
case SCTP_STREAM_RESET:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_STREAM_RESET\n");
if (((stcb == NULL) || (ch == NULL) || (chk_length < sizeof(struct sctp_stream_reset_tsn_req)))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (inp->sctp_flags & SCTP_PCB_FLAGS_SOCKET_GONE) {
/* We are not interested anymore */
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
so = SCTP_INP_SO(inp);
atomic_add_int(&stcb->asoc.refcnt, 1);
SCTP_TCB_UNLOCK(stcb);
SCTP_SOCKET_LOCK(so, 1);
SCTP_TCB_LOCK(stcb);
atomic_subtract_int(&stcb->asoc.refcnt, 1);
#endif
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
(void)sctp_free_assoc(inp, stcb, SCTP_NORMAL_PROC, SCTP_FROM_SCTP_INPUT + SCTP_LOC_30);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
#if defined (__APPLE__) || defined(SCTP_SO_LOCK_TESTING)
SCTP_SOCKET_UNLOCK(so, 1);
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (stcb->asoc.peer_supports_strreset == 0) {
/*
* hmm, peer should have announced this, but
* we will turn it on since he is sending us
* a stream reset.
*/
stcb->asoc.peer_supports_strreset = 1;
}
- Fix stream reset so it limits the number of streams that can be listed - Fix fwd-tsn to use proper accessor so it does not overrun mbufs - Fix stream reset error reporting to actually work (it has always been broken if the peer rejects a stream reset) - Some 64 bit friendly changes Approved by: re(bmah@freebsd.org)
2007-06-22 13:50:56 +00:00
if (sctp_handle_stream_reset(stcb, m, *offset, (struct sctp_stream_reset_out_req *)ch)) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
/* stop processing */
*offset = length;
return (NULL);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
break;
case SCTP_PACKET_DROPPED:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_PACKET_DROPPED\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* re-get it all please */
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if (chk_length < sizeof(struct sctp_pktdrop_chunk)) {
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (ch && (stcb) && netp && (*netp)) {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
sctp_handle_packet_dropped((struct sctp_pktdrop_chunk *)ch,
- Issue one, new stack reduction left packet_drop handling still thinking it had the whole chunk. This could cause a crash if a large packet drop came in. Fixed by adjusting the trunc length down to the limit. - Large sacks with lots of segments could also have same issue. Changed duplicate and segment handling to use proper get_m_ptr function to pull each block from mbuf chains.
2007-06-15 17:59:57 +00:00
stcb, *netp,
min(chk_length, (sizeof(chunk_buf) - 4)));
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_AUTHENTICATION:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "SCTP_AUTHENTICATION\n");
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_auth_disable))
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
goto unknown_chunk;
if (stcb == NULL) {
/* save the first AUTH for later processing */
if (auth_skipped == 0) {
auth_offset = *offset;
auth_len = chk_length;
auth_skipped = 1;
}
/* skip this chunk (temporarily) */
goto next_chunk;
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
if ((chk_length < (sizeof(struct sctp_auth_chunk))) ||
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
(chk_length > (sizeof(struct sctp_auth_chunk) +
SCTP_AUTH_DIGEST_LEN_MAX))) {
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
/* Its not ours */
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Two bugs: - Locks were not being unlocked when an invalid size chunk is sent in. - When a notification comes in, we cannot use it to look up the fragment interleave stream information since its not on a stream.
2007-05-06 00:01:17 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
- Somehow the disable fragment option got lost. We could set/clear it but would not do it. Now we will. - Moved to latest socket api for extended sndrcv info struct. - Moved to support all new levels of fragment interleave (0-2). - Codenomicon security test updates - length checks and such. - Bug in stream reset (2 actually). - setpeerprimary could unlock a null pointer, fixed. - Added a flag in the pcb so netstat can see if we are listening easier. Obtained from: (some of the Listen changes from Weongyo Jeong)
2007-05-02 12:50:13 +00:00
*offset = length;
return (NULL);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (got_auth == 1) {
/* skip this chunk... it's already auth'd */
goto next_chunk;
}
got_auth = 1;
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if ((ch == NULL) || sctp_handle_auth(stcb, (struct sctp_auth_chunk *)ch,
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
m, *offset)) {
/* auth HMAC failed so dump the packet */
*offset = length;
return (stcb);
} else {
/* remaining chunks are HMAC checked */
stcb->asoc.authenticated = 1;
}
break;
default:
unknown_chunk:
/* it's an unknown chunk! */
if ((ch->chunk_type & 0x40) && (stcb != NULL)) {
struct mbuf *mm;
struct sctp_paramhdr *phd;
mm = sctp_get_mbuf_for_msg(sizeof(struct sctp_paramhdr),
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
0, M_DONTWAIT, 1, MT_DATA);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (mm) {
phd = mtod(mm, struct sctp_paramhdr *);
/*
* We cheat and use param type since
* we did not bother to define a
* error cause struct. They are the
* same basic format with different
* names.
*/
phd->param_type = htons(SCTP_CAUSE_UNRECOG_CHUNK);
phd->param_length = htons(chk_length + sizeof(*phd));
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
SCTP_BUF_LEN(mm) = sizeof(*phd);
- Macroizes the V6ONLY flag check. - Added a short time wait (not used yet) constant - Corrected the type of the crc32c table (it was unsigned long and really is a uint32_t - Got rid of the user of MHeaders until they are truely needed by lower layers. - Fixed an initialization problem in the readq structure (ordering was off). - Found yet another collision bug when the random number generator returns two numbers on one side (during a collision) that are the same. Also added some tracking of cookies that will go away when we know that we have the last collision bug gone. - Fixed an init bug for book_size_scale, that was causing Early FR code to run when it should not. - Fixed a flight size tracking bug that was associated with Early FR but due to above bug also effected all FR's - Fixed it so Max Burst also will apply to Fast Retransmit. - Fixed a bug in the temporary logging code that allowed a static log array overflow - hashinit_flags is now used. - Two last mcopym's were converted to the macro sctp_m_copym that has always been used by all other places - macro sctp_m_copym was converted to upper case. - We now validate sinfo_flags on input (we did not before). - Fixed a bug that prevented a user from sending data and immediately shuting down with one send operation. - Moved to use hashdestroy instead of free() in our macros. - Fixed an init problem in our timed_wait vtag where we did not fully initialize our time-wait blocks. - Timer stops were re-positioned. - A pcb cleanup method was added, however this probably will not be used in BSD.. unless we make module loadable protocols - I think this fixes the mysterious timer bug.. it was a ordering of locks problem in the way we did timers. It now conforms to the timeout(9) manual (except for the _drain part, we had to do this a different way due to locks). - Fixed error return code so we get either CONNREUSED or CONNRESET depending on where one is in progression - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. - Purged an unused clone macro. - Fixed a read erro code issue where we were NOT getting the proper error when the connection was reset. Approved by: gnn
2007-01-15 15:12:10 +00:00
SCTP_BUF_NEXT(mm) = SCTP_M_COPYM(m, *offset, SCTP_SIZE32(chk_length),
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
M_DONTWAIT);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
if (SCTP_BUF_NEXT(mm)) {
Use the pru_flush infrastructure to avoid a panic PR: 122710 MFC after: 1 week
2008-04-14 18:13:33 +00:00
#ifdef SCTP_MBUF_LOGGING
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MBUF_LOGGING_ENABLE) {
Use the pru_flush infrastructure to avoid a panic PR: 122710 MFC after: 1 week
2008-04-14 18:13:33 +00:00
struct mbuf *mat;
mat = SCTP_BUF_NEXT(mm);
while (mat) {
if (SCTP_BUF_IS_EXTENDED(mat)) {
sctp_log_mb(mat, SCTP_MBUF_ICOPY);
}
mat = SCTP_BUF_NEXT(mat);
}
}
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_queue_op_err(stcb, mm);
} else {
sctp_m_freem(mm);
}
}
}
if ((ch->chunk_type & 0x80) == 0) {
/* discard this packet */
*offset = length;
return (stcb);
} /* else skip this bad chunk and continue... */
break;
} /* switch (ch->chunk_type) */
next_chunk:
/* get the next chunk */
*offset += SCTP_SIZE32(chk_length);
if (*offset >= length) {
/* no more data left in the mbuf chain */
break;
}
ch = (struct sctp_chunkhdr *)sctp_m_getptr(m, *offset,
sizeof(struct sctp_chunkhdr), chunk_buf);
if (ch == NULL) {
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
if (locked_tcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(locked_tcb);
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
*offset = length;
return (NULL);
}
} /* while */
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if (asconf_cnt > 0 && stcb != NULL) {
sctp_send_asconf_ack(stcb);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
return (stcb);
}
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
#ifdef INVARIANTS
This does two changes: 1) Makes it so that the INVARIANT function validate nolocks is available anywhere. 2) Fixes a BUG where a close has been done on a collision socket and the cookie processing would return leaving a lock held. MFC after: 1 week
2010-06-05 21:20:28 +00:00
#ifdef __GNUC__
__attribute__((noinline))
#endif
void
sctp_validate_no_locks(struct sctp_inpcb *inp)
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
{
This does two changes: 1) Makes it so that the INVARIANT function validate nolocks is available anywhere. 2) Fixes a BUG where a close has been done on a collision socket and the cookie processing would return leaving a lock held. MFC after: 1 week
2010-06-05 21:20:28 +00:00
struct sctp_tcb *lstcb;
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
This does two changes: 1) Makes it so that the INVARIANT function validate nolocks is available anywhere. 2) Fixes a BUG where a close has been done on a collision socket and the cookie processing would return leaving a lock held. MFC after: 1 week
2010-06-05 21:20:28 +00:00
LIST_FOREACH(lstcb, &inp->sctp_asoc_list, sctp_tcblist) {
if (mtx_owned(&lstcb->tcb_mtx)) {
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
panic("Own lock on stcb at return from input");
}
}
1) Further enhance the INVARIANT lock validation (no locks) are held by checking the create and inp locks as well. 2) Fix a bug in that when a socket is closed an INIT-ACK is returned, we do NOT unlock the locked_tcb unless its different (an unlikely scenario). If we blindly unlock as we were doing before we can end up unlocking the actual stcb thats about to be sent down to the free function which requires the lock be held. MFC after: 1 week
2010-06-06 16:11:16 +00:00
if (mtx_owned(&inp->inp_create_mtx)) {
panic("Own create lock on inp");
}
if (mtx_owned(&inp->inp_mtx)) {
panic("Own inp lock on inp");
}
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
}
#endif
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* common input chunk processing (v4 and v6)
*/
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
void
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_common_input_processing(struct mbuf **mm, int iphlen, int offset,
int length, struct sctphdr *sh, struct sctp_chunkhdr *ch,
struct sctp_inpcb *inp, struct sctp_tcb *stcb, struct sctp_nets *net,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
uint8_t ecn_bits, uint32_t vrf_id, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
/*
* Control chunk processing
*/
uint32_t high_tsn;
int fwd_tsn_seen = 0, data_processed = 0;
struct mbuf *m = *mm;
int abort_flag = 0;
int un_sent;
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
int cnt_ctrl_ready = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR(sctps_recvdatagrams);
#ifdef SCTP_AUDITING_ENABLED
sctp_audit_log(0xE0, 1);
sctp_auditing(0, inp, stcb, net);
#endif
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "Ok, Common input processing called, m:%p iphlen:%d offset:%d length:%d stcb:%p\n",
m, iphlen, offset, length, stcb);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb) {
/* always clear this before beginning a packet */
stcb->asoc.authenticated = 0;
stcb->asoc.seen_a_sack_this_pkt = 0;
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1, "stcb:%p state:%x\n",
stcb, stcb->asoc.state);
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
if ((stcb->asoc.state & SCTP_STATE_WAS_ABORTED) ||
(stcb->asoc.state & SCTP_STATE_ABOUT_TO_BE_FREED)) {
- change number assignments for SHA225-512 (match artisync for bakeoff.. using the next sequential ones) - In cookie processing 1-2-1, we did not increment the stcb refcnt before releasing the tcb lock. We need to do this to keep the tcb from being freed by a abort or ?? unlikely but worth doing. Also get rid of unneed INP_WLOCK. - extra receive info included the rcvinfo which killed the padding/alignment. We now redefine all the fields properly so they both align properly both to 128 bytes. - A peeled off socket would not close without an error due to its misguided idea that sctp_disconnect() was not supported on it. This fixes it so it goes through the proper path. - When an assoc was being deleted after abort (via a timer) a small race condition exists where we might take a packet for the old assoc (since we are waiting for a cleanup timer). This state especially happens in mac. We now add a state in the asoc so these can properly handle the packet as OOTB. Approved by: re@freebsd.org(Ken Smith)
2007-08-06 15:46:46 +00:00
/*-
* If we hit here, we had a ref count
* up when the assoc was aborted and the
* timer is clearing out the assoc, we should
* NOT respond to any packet.. its OOTB.
*/
SCTP_TCB_UNLOCK(stcb);
sctp_handle_ootb(m, iphlen, offset, sh, inp, NULL,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- change number assignments for SHA225-512 (match artisync for bakeoff.. using the next sequential ones) - In cookie processing 1-2-1, we did not increment the stcb refcnt before releasing the tcb lock. We need to do this to keep the tcb from being freed by a abort or ?? unlikely but worth doing. Also get rid of unneed INP_WLOCK. - extra receive info included the rcvinfo which killed the padding/alignment. We now redefine all the fields properly so they both align properly both to 128 bytes. - A peeled off socket would not close without an error due to its misguided idea that sctp_disconnect() was not supported on it. This fixes it so it goes through the proper path. - When an assoc was being deleted after abort (via a timer) a small race condition exists where we might take a packet for the old assoc (since we are waiting for a cleanup timer). This state especially happens in mac. We now add a state in the asoc so these can properly handle the packet as OOTB. Approved by: re@freebsd.org(Ken Smith)
2007-08-06 15:46:46 +00:00
goto out_now;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (IS_SCTP_CONTROL(ch)) {
/* process the control portion of the SCTP packet */
- Fixed 1-2-1 model to not worry about associd in sockopts - Fixed RTOinfo for bounding. - Fixed connect() to return ECONNREFUSED when an ABORT is received. - Added comments to direct Static Analysis not to look at some things it does not understand (comments are /* sa_ignore XXXXX */) - Bind when colliding was broken, missing not_found = 1 before checking to see if the port was in use caused endless bind loop. - Cookie life needs to be in milliseconds to conform to socket api. - Cookie life is not supposed to change if its 0, On the assoc level set we changed it to 0 opps. - Two more static analysis issues identified by the cisco tool. Null checks needed. - An issue for sendfile(). Need to validate the correct input argument. - When sending failed due to a no route to host, we leaked the mbuf chain failing to call m_freem(). - Fix #ifdef issue for getting hash block len when HAVE_SHA2 is NOT defined Reviewed by: gnn
2007-05-17 12:16:24 +00:00
/* sa_ignore NO_NULL_CHK */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb = sctp_process_control(m, iphlen, &offset, length, sh, ch,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
inp, stcb, &net, &fwd_tsn_seen, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (stcb) {
/*
* This covers us if the cookie-echo was there and
* it changes our INP.
*/
inp = stcb->sctp_ep;
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if ((net) && (port)) {
if (net->port == 0) {
sctp_pathmtu_adjustment(inp, stcb, net, net->mtu - sizeof(struct udphdr));
}
net->port = port;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
} else {
/*
* no control chunks, so pre-process DATA chunks (these
* checks are taken care of by control processing)
*/
/*
* if DATA only packet, and auth is required, then punt...
* can't have authenticated without any AUTH (control)
* chunks
*/
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if ((stcb != NULL) &&
!SCTP_BASE_SYSCTL(sctp_auth_disable) &&
sctp_auth_is_required_chunk(SCTP_DATA, stcb->asoc.local_auth_chunks)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* "silently" ignore */
SCTP_STAT_INCR(sctps_recvauthmissing);
SCTP_TCB_UNLOCK(stcb);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (stcb == NULL) {
/* out of the blue DATA chunk */
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_handle_ootb(m, iphlen, offset, sh, inp, NULL,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (stcb->asoc.my_vtag != ntohl(sh->v_tag)) {
/* v_tag mismatch! */
SCTP_STAT_INCR(sctps_badvtag);
SCTP_TCB_UNLOCK(stcb);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
if (stcb == NULL) {
/*
* no valid TCB for this packet, or we found it's a bad
* packet while processing control, or we're done with this
* packet (done or skip rest of data), so we drop it...
*/
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* DATA chunk processing
*/
/* plow through the data chunks while length > offset */
/*
* Rest should be DATA only. Check authentication state if AUTH for
* DATA is required.
*/
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if ((length > offset) &&
(stcb != NULL) &&
!SCTP_BASE_SYSCTL(sctp_auth_disable) &&
sctp_auth_is_required_chunk(SCTP_DATA, stcb->asoc.local_auth_chunks) &&
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
!stcb->asoc.authenticated) {
/* "silently" ignore */
SCTP_STAT_INCR(sctps_recvauthmissing);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_AUTH1,
"Data chunk requires AUTH, skipped\n");
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
goto trigger_send;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
if (length > offset) {
int retval;
/*
* First check to make sure our state is correct. We would
* not get here unless we really did have a tag, so we don't
* abort if this happens, just dump the chunk silently.
*/
switch (SCTP_GET_STATE(&stcb->asoc)) {
case SCTP_STATE_COOKIE_ECHOED:
/*
* we consider data with valid tags in this state
* shows us the cookie-ack was lost. Imply it was
* there.
*/
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_THRESHOLD_LOGGING) {
- Fix address add handling to clear cached routes and source addresses when peer acks the add in case the routing table changes. - Fix sctp_lower_sosend to send shutdown chunk for mbuf send case when sndlen = 0 and sinfoflag = SCTP_EOF - Fix sctp_lower_sosend for SCTP_ABORT mbuf send case with null data, So that it does not send the "null" data mbuf out and cause it to get freed twice. - Fix so auto-asconf sysctl actually effect the socket's asconf state. - Do not allow SCTP_AUTO_ASCONF option to be used on subset bound sockets. - Memset bug in sctp_output.c (arguments were reversed) submitted found and reported by Dave Jones (davej@codemonkey.org.uk). - PD-API point needs to be invoked >= not just > to conform to socket api draft this fixes sctp_indata.c in the two places need to be >=. - move M_NOTIFICATION to use M_PROTO5. - PEER_ADDR_PARAMS did not fail properly if you specify an address that is not in the association with a valid assoc_id. This meant you got or set the stcb level values instead of the destination you thought you were going to get/set. Now validate if the stcb is non-null and the net is NULL that the sa_family is set and the address is unspecified otherwise return an error. - The thread based iterator could crash if associations were freed at the exact time it was running. rework the worker thread to use the increment/decrement to prevent this and no longer use the markers that the timer based iterator uses. - Fix the memleak in sctp_add_addr_to_vrf() for the case when it is detected that ifa is already pointing to a ifn. - Fix it so that if someone is so insane that they drop the send window below the minimal add mark, they still can send. - Changed all state for associations to use mask safe macro. - During front states in association freeing in sctp_inpcbfree, we had a locking problem where locks were not in place where they should have been. - Free association calls were not testing the return value in sctp_inpcb_free() properly... others should be cast void returns where we don't care about the return value. - If a reference count is held on an assoc, even from the "force free" we should not do the actual free.. but instead let the timer free it. - When we enter sctp_input(), if the SCTP_ASOC_ABOUT_TO_BE_FREED flag is set, we must NOT process the packet but handle it like ootb. This is because while freeing an assoc we release the locks to get all the higher order locks so we can purge all the hash tables. This leaves a hole if a packet comes in just at that point. Now sctp_common_input_processing() will call the ootb code in such a case. - Change MBUF M_NOTIFICATION to use M_PROTO5 (per Sam L). This makes it so we don't have a conflict (I think this is a covertity change). We made this change AFTER some conversation and looking to make sure that M_PROTO5 does not have a problem between SCTP and the 802.11 stuff (which is the only other place its used). - Fixed lock order reversal and missing atomic protection around locked_tcb during association lookup and the 1-2-1 model. - Added debug to source address selection. - V6 output must always do checksum even for loopback. - Remove more locks around inp that are not needed for an atomically added/subtracted ref count. - slight optimization in the way we zero the array in sctp_sack_check() - It was possible to respond to a ABORT() with bad checksum with a PKT-DROP. This lead to a PKT-DROP/ABORT war. Add code to NOT send a PKT-DROP to any ABORT(). - Add an option for local logging (useful for macintosh or when you need better performing during debugging). Note no commands are here to get the log info, you must just use kgdb. - The timer code needs to be aware of if it needs to call sctp_sack_check() to slide the maps and adjust the cum-ack. This is because it may be out of sync cum-ack wise. - Added threshold managment logging. - If the user picked just the right size, that just filled the send window minus one mtu, we would enter a forever loop not copying and at the same time not blocking. Change from < to <= solves this. - Sysctl added to control the fragment interleave level which defaults to 1. - My rwnd control was not being used to control the rwnd properly (we did not add and subtract to it :-() this is now fixed so we handle small messages (1 byte etc) better to bring our rwnd down more slowly. Approved by: re@freebsd.org (Bruce Mah)
2007-08-24 00:53:53 +00:00
sctp_misc_ints(SCTP_THRESHOLD_CLEAR,
stcb->asoc.overall_error_count,
0,
SCTP_FROM_SCTP_INPUT,
__LINE__);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
stcb->asoc.overall_error_count = 0;
sctp_handle_cookie_ack((struct sctp_cookie_ack_chunk *)ch, stcb, net);
break;
case SCTP_STATE_COOKIE_WAIT:
/*
* We consider OOTB any data sent during asoc setup.
*/
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
sctp_handle_ootb(m, iphlen, offset, sh, inp, NULL,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(stcb);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
- remove duplicate code from sctp_asconf.c - remove duplicate #include <sys/priv.h> that is not under #ifdef FreeBSD version to allow compile on 6.1 - static analysis changes per the cisco SA tool including: o some SA_IGNORE comments o some checks for NULL before unlock. o type corrections int -> size_t - Fix it so sctp_alloc_asoc takes a thread/proc argument. Without this we pass a NULL in to bind on implicit assoc setup and crash :-( Approved by: re@freebsd.org(Ken Smith)
2007-07-21 21:41:32 +00:00
/* sa_ignore NOTREACHED */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_STATE_EMPTY: /* should not happen */
case SCTP_STATE_INUSE: /* should not happen */
case SCTP_STATE_SHUTDOWN_RECEIVED: /* This is a peer error */
case SCTP_STATE_SHUTDOWN_ACK_SENT:
default:
SCTP_TCB_UNLOCK(stcb);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
- remove duplicate code from sctp_asconf.c - remove duplicate #include <sys/priv.h> that is not under #ifdef FreeBSD version to allow compile on 6.1 - static analysis changes per the cisco SA tool including: o some SA_IGNORE comments o some checks for NULL before unlock. o type corrections int -> size_t - Fix it so sctp_alloc_asoc takes a thread/proc argument. Without this we pass a NULL in to bind on implicit assoc setup and crash :-( Approved by: re@freebsd.org(Ken Smith)
2007-07-21 21:41:32 +00:00
/* sa_ignore NOTREACHED */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
break;
case SCTP_STATE_OPEN:
case SCTP_STATE_SHUTDOWN_SENT:
break;
}
/* plow through the data chunks while length > offset */
retval = sctp_process_data(mm, iphlen, &offset, length, sh,
inp, stcb, net, &high_tsn);
if (retval == 2) {
/*
* The association aborted, NO UNLOCK needed since
* the association is destroyed.
*/
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
data_processed = 1;
/*
* Anything important needs to have been m_copy'ed in
* process_data
*/
}
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
/* take care of ecn */
Make sure that changing the ECN sysctl does not affect exisiting associations and endpoints. MFC after: 3 months.
2011-02-03 19:59:00 +00:00
if ((stcb->asoc.ecn_allowed == 1) &&
((ecn_bits & SCTP_CE_BITS) == SCTP_CE_BITS)) {
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
/* Yep, we need to add a ECNE */
sctp_send_ecn_echo(stcb, net, high_tsn);
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((data_processed == 0) && (fwd_tsn_seen)) {
Bugfix: Take also the nr-mapping array into account when detecting gaps. Reviewed by: rrs@ MFC after: 3 days.
2010-12-16 21:01:02 +00:00
int was_a_gap;
uint32_t highest_tsn;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
if (SCTP_TSN_GT(stcb->asoc.highest_tsn_inside_nr_map, stcb->asoc.highest_tsn_inside_map)) {
Bugfix: Take also the nr-mapping array into account when detecting gaps. Reviewed by: rrs@ MFC after: 3 days.
2010-12-16 21:01:02 +00:00
highest_tsn = stcb->asoc.highest_tsn_inside_nr_map;
} else {
highest_tsn = stcb->asoc.highest_tsn_inside_map;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Define and use SCTP_SSN_GE, SCTP_SSN_GT, SCTP_TSN_GE, SCTP_TSN_GT macros and use them instead of the generic compare_with_wrap. Retire compare_with_wrap. MFC after: 3 months.
2010-12-30 21:32:35 +00:00
was_a_gap = SCTP_TSN_GT(highest_tsn, stcb->asoc.cumulative_tsn);
Many bug fixes (from the IETF hack-fest): - PR-SCTP had major issues when skipping through a multi-part message. o Did not look at socket buffer. o Did not properly handle the reassmebly queue. o The MARKED segments could interfere and un-skip a chunk causing a problem with the proper FWD-TSN. o No FR of FWD-TSN's was being done. - NR-Sack code was basically disabled. It needed fixes that never got into the real code. - CMT code had issues when the two paths were NOT the same b/w. We found a few small bugs, but also the critcal one here was not dividing the rwnd amongst the paths. Obtained from: Michael Tuexen and myself at the IETF hack-fest ;-)
2009-04-04 11:43:32 +00:00
stcb->asoc.send_sack = 1;
* Fix some race condition in SACK/NR-SACK processing. * Fix handling of mapping arrays when draining mbufs or processing FORWARD-TSN chunks. * Cleanup code (no duplicate code anymore for SACKs and NR-SACKs). Part of this code was developed together with rrs. MFC after: 2 weeks.
2010-04-03 15:40:14 +00:00
sctp_sack_check(stcb, was_a_gap, &abort_flag);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (abort_flag) {
/* Again, we aborted so NO UNLOCK needed */
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
goto out_now;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Many bug fixes (from the IETF hack-fest): - PR-SCTP had major issues when skipping through a multi-part message. o Did not look at socket buffer. o Did not properly handle the reassmebly queue. o The MARKED segments could interfere and un-skip a chunk causing a problem with the proper FWD-TSN. o No FR of FWD-TSN's was being done. - NR-Sack code was basically disabled. It needed fixes that never got into the real code. - CMT code had issues when the two paths were NOT the same b/w. We found a few small bugs, but also the critcal one here was not dividing the rwnd amongst the paths. Obtained from: Michael Tuexen and myself at the IETF hack-fest ;-)
2009-04-04 11:43:32 +00:00
} else if (fwd_tsn_seen) {
stcb->asoc.send_sack = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/* trigger send of any chunks in queue... */
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
trigger_send:
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
#ifdef SCTP_AUDITING_ENABLED
sctp_audit_log(0xE0, 2);
sctp_auditing(1, inp, stcb, net);
#endif
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"Check for chunk output prw:%d tqe:%d tf=%d\n",
stcb->asoc.peers_rwnd,
TAILQ_EMPTY(&stcb->asoc.control_send_queue),
stcb->asoc.total_flight);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
un_sent = (stcb->asoc.total_output_queue_size - stcb->asoc.total_flight);
More ECN fixes: 1) We now remove ECN-Nonce since it will no longer continue as a I-D 2) Eliminate last_tsn_echo, this tied us to an assoc not the net and thus we were not doing m-homing on the ECN-Echo senders side right. 3) Increment the count going out even if the TSN in lower in the pending ECN-Echo, this way the receiver knows exactly how many packets were marked even with network re-ordering 4) Fix so we DO NOT stop doing delayed sack if a ECN Echo is in queue MFC after: 1 month
2011-01-31 11:50:11 +00:00
if (!TAILQ_EMPTY(&stcb->asoc.control_send_queue)) {
cnt_ctrl_ready = stcb->asoc.ctrl_queue_cnt - stcb->asoc.ecn_echo_cnt_onq;
}
if (cnt_ctrl_ready ||
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
((un_sent) &&
(stcb->asoc.peers_rwnd > 0 ||
(stcb->asoc.peers_rwnd <= 0 && stcb->asoc.total_flight == 0)))) {
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "Calling chunk OUTPUT\n");
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_chunk_output(inp, stcb, SCTP_OUTPUT_FROM_CONTROL_PROC, SCTP_SO_NOT_LOCKED);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT3, "chunk OUTPUT returns\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
#ifdef SCTP_AUDITING_ENABLED
sctp_audit_log(0xE0, 3);
sctp_auditing(2, inp, stcb, net);
#endif
SCTP_TCB_UNLOCK(stcb);
- Modular congestion control, with RFC2581 being the default. - CMT_PF states added (w/sysctl to turn the PF version on) - sctp_input.c had a missing incr of cookie case when the auth was bad. This meant a free was called without an increment to refcnt, added increment like rest of code. - There was a case, unlikely, when the scope of the destination changed (this is a TSNH case). In that case, it would not free the alloc'ed asoc (in sctp_input.c). - When listed addresses found a colliding cookie/Init, then the collided upon tcb was not unlocked in sctp_pcb.c - Add error checking on arguments of sctp_sendx(3) to prevent it from referencing a NULL pointer. - Fix an error return of sctp_sendx(3), it was returing ENOMEM not -1. - Get assoc id was changed to use the sanctified socket api method for getting a assoc id (PEER_ADDR_INFO instead of PEER_ADDR_PARAMS). - Fix it so a peeled off socket will get a proper error return if it trys to send to a different address then it is connected to. - Fix so that select_a_stream can avoid an endless loop that could hang a caller. - time_entered (state set time) was not being set in all cases to the time we went established. Approved by: re(ken smith)
2007-07-14 09:36:28 +00:00
out_now:
#ifdef INVARIANTS
sctp_validate_no_locks(inp);
#endif
- Static analyisis fixes for cisco's commit (this is equivilant to the coverity tool.. may even be the same one.. not sure). - A bug in the way sctp_abort() and friends were setting the IP_CLOSE flag.. and NOT passing the last argument as a (,1)... so that things would get freed..
2007-05-08 14:32:53 +00:00
return;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
Code from the hack-session known as the IETF (and a bit of debugging afterwards): - Fix protection code for notification generation. - Decouple associd from vtag - Allow vtags to have less strigent requirements in non-uniqueness. o don't pre-hash them when you issue one in a cookie. o Allow duplicates and use addresses and ports to discriminate amongst the duplicates during lookup. - Add support for the NAT draft draft-ietf-behave-sctpnat-00, this is still experimental and needs more extensive testing with the Jason Butt ipfw changes. - Support for the SENDER_DRY event to get DTLS in OpenSSL working with a set of patches from Michael Tuexen (hopefully heading to OpenSSL soon). - Update the support of SCTP-AUTH by Peter Lei. - Use macros for refcounting. - Fix MTU for UDP encapsulation. - Fix reporting back of unsent data. - Update assoc send counter handling to be consistent with endpoint sent counter. - Fix a bug in PR-SCTP. - Fix so we only send another FWD-TSN when a SACK arrives IF and only if the adv-peer-ack point progressed. However we still make sure a timer is running if we do have an adv_peer_ack point. - Fix PR-SCTP bug where chunks were retransmitted if they are sent unreliable but not abandoned yet. With the help of: Michael Teuxen and Peter Lei :-) MFC after: 4 weeks
2008-12-06 13:19:54 +00:00
#if 0
static void
sctp_print_mbuf_chain(struct mbuf *m)
{
for (; m; m = SCTP_BUF_NEXT(m)) {
printf("%p: m_len = %ld\n", m, SCTP_BUF_LEN(m));
if (SCTP_BUF_IS_EXTENDED(m))
printf("%p: extend_size = %d\n", m, SCTP_BUF_EXTEND_SIZE(m));
}
}
#endif
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#ifdef INET
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
void
Change the functions to ANSI in those cases where it breaks promotion to int rule. See ISO C Standard: SS6.7.5.3:15. Approved by: kib (mentor) Reviewed by: warner Tested by: silence on -current
2009-02-24 18:09:31 +00:00
sctp_input_with_port(struct mbuf *i_pak, int off, uint16_t port)
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
{
#ifdef SCTP_MBUF_LOGGING
struct mbuf *mat;
#endif
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
struct mbuf *m;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
int iphlen;
- Take out the broken table-id concept. Panda Routers have a M-VRF concept that is NOT well thought out for a multi-homed transport protocol. So the useless table-id entries passed around need to be removed. - Add a event timer for the zero copy api. - Fix a bug in sctp_timer.c when searching for an alternate with the largest ssthresh (the compare was wrong).
2007-06-01 11:19:54 +00:00
uint32_t vrf_id = 0;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
uint8_t ecn_bits;
struct ip *ip;
struct sctphdr *sh;
struct sctp_inpcb *inp = NULL;
struct sctp_nets *net;
struct sctp_tcb *stcb = NULL;
struct sctp_chunkhdr *ch;
int refcount_up = 0;
int length, mlen, offset;
Fix the the SCTP_WITH_NO_CSUM option when used in combination with interface supporting CRC offload. While at it, make use of the feature that the loopback interface provides CRC offloading. MFC after: 4 weeks
2010-08-29 18:50:30 +00:00
#if !defined(SCTP_WITH_NO_CSUM)
uint32_t check, calc_check;
#endif
- More macros for OS compatabilty - PR-SCTP would ignore FWD-TSN's above a rwnd's worth of TSN's (1 byte msgs).. this left the peer hopelessly out of sync.. or an attacker. So now we abort the assoc. - New IFN hash, also rename hashes to match addr/ifn now that the vrf has multiple. - Do not enable SCTP_PCB_FLAGS_RECVDATAIOEVNT per default as defined in the Socket API ID. - Export MTU information via sysctl. - Vrf's need table id's. This is default for BSD, but may be other things later when BSD fully supports VRFs. - Additional stream reset bug (caught by cisco dev-test). - Additional validations for the address in sending a message (socket api). -------- and ----- - Fix association notifications not to give the active open side false notifications. - Fix so sendfile and SENDALL will work properly (missing flag to say socket sender is done). - Fix Bug that prevented COOKIES from being retransmitted. - Break out connectx into helper sub-models so that iox routines can reuse the helpers. - When an address is added during system init (non-dynamic mode) make sure that the "defer use" flag is not set. ** its compiling on XR now :-D ** Reviewed by: gnn
2007-05-08 00:21:05 +00:00
if (SCTP_GET_PKT_VRFID(i_pak, vrf_id)) {
SCTP_RELEASE_PKT(i_pak);
return;
}
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
mlen = SCTP_HEADER_LEN(i_pak);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
iphlen = off;
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
m = SCTP_HEADER_TO_CHAIN(i_pak);
- fixed autclose to not allow setting on 1-2-1 model. - bounded cookie-life to 1 second minimum in socket option set. - Delayed_ack_time becomes delayed_ack per new socket api document. - Improve port number selection, we now use low/high bounds and no chance of a endless loop. Only one call to random per bind as well. - fixes so set_peer_primary pre-screens addresses to be valid to this host. - maxseg did not allow setting on an assoc basis. We needed to thus track and use an association value instead of a inp value. - Fixed ep get of HB status to report back properly. - use settings flag to tell if assoc level hb is on off not the timer.. since the timer may still run if unconf address are present. - check for crazy ENABLE/DISABLE conditions. - set and get of pmtud (fixed path mtu) not always taking into account ovh. - Getting PMTU info on stcb only needs to return PMTUD_ENABLED if any net is doing PMTU discovery. - Panic or warning fixed to not do so when a valid ip frag is taking place. - sndrcvinfo appearing in both inp and stcb was full size, instead of the non-pad version. This saves about 92 bytes from each struct by carefully converting to use the smaller version. - one-2-one model get(maxseg) would always get ep value, never the tcb's value. - The delayed ack time could be under a tick, this fixes so it bounds it to at least 1 tick for platforms whos tick is more than a ms. - Fragment interleave level set to wrong default value. - Fragment interleave could not set level 0. - Defered stream reset was broken due to a guard check and ntohl issue. - Found two lock order reversals and fixed. - Tighten up address checking, if the user gives an address the sa_len had better be set properly. - Get asoc by assoc-id would return a locked tcb when it was asked not to if the tcb was in the restart hash. - sysctl to dig down and get more association details Reviewed by: gnn
2007-05-28 11:17:24 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
net = NULL;
SCTP_STAT_INCR(sctps_recvpackets);
SCTP_STAT_INCR_COUNTER64(sctps_inpackets);
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
#ifdef SCTP_MBUF_LOGGING
/* Log in any input mbufs */
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_MBUF_LOGGING_ENABLE) {
- Fix so ifn's are properly deleted when the ref count goes to 0. - Fix so VRF's will clean themselves up when no references are around. - Allow sctp_ifa to be passed into inpcb_bind, addr_mgmt_ep_sa to bypass normal validation checks. - turn auto-asconf off for subset bound sockets - Moves all logging to use KTR. This gets rid of most of the logging #ifdef's with a few exceptions reducing the number of config options for SCTP.
2007-06-14 22:59:04 +00:00
mat = m;
while (mat) {
if (SCTP_BUF_IS_EXTENDED(mat)) {
sctp_log_mb(mat, SCTP_MBUF_INPUT);
}
mat = SCTP_BUF_NEXT(mat);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
}
#endif
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
#ifdef SCTP_PACKET_LOGGING
- Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-)
2008-06-14 07:58:05 +00:00
if (SCTP_BASE_SYSCTL(sctp_logging_level) & SCTP_LAST_PACKET_TRACING)
Add additional logging level mask for packet_logging too.
2007-06-18 13:57:37 +00:00
sctp_packet_log(m, mlen);
- Fixes so we won't try to start a timer when we hold a wq lock for the iterator. Panda uses a silly recursive lock they hold through the timer. - Add poor mans wireshark compile option.. - Allocate and start using SCTP_M_XXX for all SCTP_MALLOC() calls. - sysctl now will get back the refcnt for viewing by onlookers. Reviewed by: gnn
2007-05-29 09:29:03 +00:00
#endif
/*
* Must take out the iphlen, since mlen expects this (only effect lb
* case)
*/
mlen -= iphlen;
- Found bug in min split point bundling which caused incorrect, non-bundlable fragmentation. - Added min residual to better control split points for both how big a msg must be as well as how much needs to be left over. - With our new algo in place, we need to implicitly set "end of msg" on the sp-> structure otherwise we end up with "hung" associations. - Room reserved up front in IP header by pushing IP header to back of mbuf. - Fix so FR's peg count of retransmissions needed. - Fix so an unlucky chunk that never gets across will kill the assoc via the kill timer and send an abort too. - Fix bug in sctp_input which can result in a crash. - Do not strip off IP options anymore. - Clean up sctp_calculate_rto(). - Get rid of unused sysctl. - Fixed so we discard all M-Cast - Fixed so port check done AFTER checksum - Fixed bug in fragmentation code that prevented us from fragmenting a small complete message when we needed to. - Window probes were not marked back to unsent and flight adjusted when a sack came in with no window change or accepting of the probe data. We now fix this with having a mark on the net and the chunk so we can clear it out when the sack arrives forcing it to retran just like it was "new" this improves the handling of window probes, which were dropped by the receiver. - Tighten AUTH protocol error checks during INIT/INIT-ACK exchange
2007-03-31 11:47:30 +00:00
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* Get IP, SCTP, and first chunk header together in first mbuf.
*/
ip = mtod(m, struct ip *);
offset = iphlen + sizeof(*sh) + sizeof(*ch);
a) macro-ization of all mbuf and random number access plus timers. This makes the code more portable and able to change out the mbuf or timer system used more easily ;-) b) removal of all use of pkt-hdr's until only the places we need them (before ip_output routines). c) remove a bunch of code not needed due to <b> aka worrying about pkthdr's :-) d) There was one last reorder problem it looks where if a restart occur's and we release and relock (at the point where we setup our alias vtag) we would end up possibly getting the wrong TSN in place. The code that fixed the TSN's just needed to be shifted around BEFORE the release of the lock.. also code that set the state (since this also could contribute). Approved by: gnn
2006-12-29 20:21:42 +00:00
if (SCTP_BUF_LEN(m) < offset) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((m = m_pullup(m, offset)) == 0) {
SCTP_STAT_INCR(sctps_hdrops);
return;
}
ip = mtod(m, struct ip *);
}
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
/* validate mbuf chain length with IP payload length */
if (mlen < (SCTP_GET_IPV4_LENGTH(ip) - iphlen)) {
SCTP_STAT_INCR(sctps_hdrops);
goto bad;
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sh = (struct sctphdr *)((caddr_t)ip + iphlen);
ch = (struct sctp_chunkhdr *)((caddr_t)sh + sizeof(*sh));
- For sctp_input/sctp6_input add announcment when a packet arrives (debug) - re-factor the packet drop in sctp_output a bit more, we don't need the trim after all, but the size calc is now corrected. - When a assoc is in the COOKIE-ECHO/COOKIE-WAIT state and the user closes, it should not matter if data is queued, the assoc should be purged. - In error leg a missing free_chunk when iph comes in NULL (should not happen but just in case).
2007-06-17 01:36:02 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"sctp_input() length:%d iphlen:%d\n", mlen, iphlen);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* SCTP does not allow broadcasts or multicasts */
if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr))) {
goto bad;
}
- Found bug in min split point bundling which caused incorrect, non-bundlable fragmentation. - Added min residual to better control split points for both how big a msg must be as well as how much needs to be left over. - With our new algo in place, we need to implicitly set "end of msg" on the sp-> structure otherwise we end up with "hung" associations. - Room reserved up front in IP header by pushing IP header to back of mbuf. - Fix so FR's peg count of retransmissions needed. - Fix so an unlucky chunk that never gets across will kill the assoc via the kill timer and send an abort too. - Fix bug in sctp_input which can result in a crash. - Do not strip off IP options anymore. - Clean up sctp_calculate_rto(). - Get rid of unused sysctl. - Fixed so we discard all M-Cast - Fixed so port check done AFTER checksum - Fixed bug in fragmentation code that prevented us from fragmenting a small complete message when we needed to. - Window probes were not marked back to unsent and flight adjusted when a sack came in with no window change or accepting of the probe data. We now fix this with having a mark on the net and the chunk so we can clear it out when the sack arrives forcing it to retran just like it was "new" this improves the handling of window probes, which were dropped by the receiver. - Tighten AUTH protocol error checks during INIT/INIT-ACK exchange
2007-03-31 11:47:30 +00:00
if (SCTP_IS_IT_BROADCAST(ip->ip_dst, m)) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* We only look at broadcast if its a front state, All
* others we will not have a tcb for anyway.
*/
goto bad;
}
/* validate SCTP checksum */
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
SCTPDBG(SCTP_DEBUG_CRCOFFLOAD,
"sctp_input(): Packet of length %d received on %s with csum_flags 0x%x.\n",
m->m_pkthdr.len,
if_name(m->m_pkthdr.rcvif),
m->m_pkthdr.csum_flags);
Fix the the SCTP_WITH_NO_CSUM option when used in combination with interface supporting CRC offload. While at it, make use of the feature that the loopback interface provides CRC offloading. MFC after: 4 weeks
2010-08-29 18:50:30 +00:00
#if defined(SCTP_WITH_NO_CSUM)
SCTP_STAT_INCR(sctps_recvnocrc);
#else
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
if (m->m_pkthdr.csum_flags & CSUM_SCTP_VALID) {
SCTP_STAT_INCR(sctps_recvhwcrc);
goto sctp_skip_csum_4;
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
check = sh->checksum; /* save incoming checksum */
sh->checksum = 0; /* prepare for calc */
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
calc_check = sctp_calculate_cksum(m, iphlen);
Restore the checksum received before processing the packet. MFC after: 1 week
2010-02-04 21:02:29 +00:00
sh->checksum = check;
- Cleanup checksum code. - Prepare for CRC offloading, add MIB counters (RS/MT). - Bugfix: Disable CRC computation for IPv6 addresses with local scope (MT). - Bugfix: Handle close() with SO_LINGER correctly when notifications are generated during the close() call(MT). - Bugfix: Generate DRY event when sender is dry during subscription. Only for 1-to-1 style sockets (RS/MT) - Bugfix: Put vtags for the correct amount of time into time-wait (MT). - Bugfix: Clear vtag entries correctly on expiration (MT). - Bugfix: shutdown() indicates ENOTCONN when called for unconnected 1-to-1 style sockets (MT). - Bugfix: In sctp Auth code (PL). - Add support for devices that support SCTP csum offload (igb). - Add missing sctp_associd to mib sysctl xsctp_tcb structure (RS) Obtained from: With help from Peter Lei and Michael Tuexen
2009-02-03 11:04:03 +00:00
SCTP_STAT_INCR(sctps_recvswcrc);
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if (calc_check != check) {
SCTPDBG(SCTP_DEBUG_INPUT1, "Bad CSUM on SCTP packet calc_check:%x check:%x m:%p mlen:%d iphlen:%d\n",
calc_check, check, m, mlen, iphlen);
stcb = sctp_findassociation_addr(m, iphlen,
offset - sizeof(*ch),
sh, ch, &inp, &net,
vrf_id);
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
if ((net) && (port)) {
if (net->port == 0) {
sctp_pathmtu_adjustment(inp, stcb, net, net->mtu - sizeof(struct udphdr));
}
net->port = port;
}
Fix bugs related to M_FLOWID: * Store the flowid when receiving an SCTP/IPv6 packet. * Store the flowid when receiving an SCTP packet with wrong CRC. * Initilize flowid correctly. * Put test code under INVARIANTS. MFC after: 3 months.
2011-02-07 15:04:23 +00:00
if ((net != NULL) && (m->m_flags & M_FLOWID)) {
net->flowid = m->m_pkthdr.flowid;
#ifdef INVARIANTS
net->flowidset = 1;
#endif
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
if ((inp) && (stcb)) {
sctp_send_packet_dropped(stcb, net, m, iphlen, 1);
- Locking compatiability changes. This involves adding additional flags to many function calls. The flags only get used in BSD when we compile with lock testing. These flags allow apple to escape the "giant" lock it holds on the socket and have more fine-grained locking in the NKE. It also allows us to test (with witness) the locking used by apple via a compile switch (manually applied). Approved by: re@freebsd.org(B Mah)
2007-09-08 11:35:11 +00:00
sctp_chunk_output(inp, stcb, SCTP_OUTPUT_FROM_INPUT_ERROR, SCTP_SO_NOT_LOCKED);
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
} else if ((inp != NULL) && (stcb == NULL)) {
refcount_up = 1;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
- During shutdown pending, when the last sack came in and the last message on the send stream was "null" but still there, a state we allow, we could get hung and not clean it up and wait for the shutdown guard timer to clear the association without a graceful close. Fix this so that that we properly clean up. - Added support for Multiple ASCONF per new RFC. We only (so far) accept input of these and cannot yet generate a multi-asconf. - Sysctl'd support for experimental Fast Handover feature. Always disabled unless sysctl or socket option changes to enable. - Error case in add-ip where the peer supports AUTH and ADD-IP but does NOT require AUTH of ASCONF/ASCONF-ACK. We need to ABORT in this case. - According to the Kyoto summit of socket api developers (Solaris, Linux, BSD). We need to have: o non-eeor mode messages be atomic - Fixed o Allow implicit setup of an assoc in 1-2-1 model if using the sctp_**() send calls - Fixed o Get rid of HAVE_XXX declarations - Done o add a sctp_pr_policy in hole in sndrcvinfo structure - Done o add a PR_SCTP_POLICY_VALID type flag - yet to-do in a future patch! - Optimize sctp6 calls to reuse code in sctp_usrreq. Also optimize when we close sending out the data and disabling Nagle. - Change key concatenation order to match the auth RFC - When sending OOTB shutdown_complete always do csum. - Don't send PKT-DROP to a PKT-DROP - For abort chunks just always checksums same for shutdown-complete. - inpcb_free front state had a bug where in queue data could wedge an assoc. We need to just abandon ones in front states (free_assoc). - If a peer sends us a 64k abort, we would try to assemble a response packet which may be larger than 64k. This then would be dropped by IP. Instead make a "minimum" size for us 64k-2k (we want at least 2k for our initack). If we receive such an init discard it early without all the processing. - When we peel off we must increment the tcb ref count to keep it from being freed from underneath us. - handling fwd-tsn had bugs that caused memory overwrites when given faulty data, fixed so can't happen and we also stop at the first bad stream no. - Fixed so comm-up generates the adaption indication. - peeloff did not get the hmac params copied. - fix it so we lock the addr list when doing src-addr selection (in future we need to use a multi-reader/one writer lock here) - During lowlevel output, we could end up with a _l_addr set to null if the iterator is calling the output routine. This means we would possibly crash when we gather the MTU info. Fix so we only do the gather where we have a src address cached. - we need to be sure to set abort flag on conn state when we receive an abort. - peeloff could leak a socket. Moved code so the close will find the socket if the peeloff fails (uipc_syscalls.c) Approved by: re@freebsd.org(Ken Smith)
2007-08-27 05:19:48 +00:00
SCTP_STAT_INCR(sctps_badsum);
SCTP_STAT_INCR_COUNTER32(sctps_checksumerrors);
goto bad;
- Found bug in min split point bundling which caused incorrect, non-bundlable fragmentation. - Added min residual to better control split points for both how big a msg must be as well as how much needs to be left over. - With our new algo in place, we need to implicitly set "end of msg" on the sp-> structure otherwise we end up with "hung" associations. - Room reserved up front in IP header by pushing IP header to back of mbuf. - Fix so FR's peg count of retransmissions needed. - Fix so an unlucky chunk that never gets across will kill the assoc via the kill timer and send an abort too. - Fix bug in sctp_input which can result in a crash. - Do not strip off IP options anymore. - Clean up sctp_calculate_rto(). - Get rid of unused sysctl. - Fixed so we discard all M-Cast - Fixed so port check done AFTER checksum - Fixed bug in fragmentation code that prevented us from fragmenting a small complete message when we needed to. - Window probes were not marked back to unsent and flight adjusted when a sack came in with no window change or accepting of the probe data. We now fix this with having a mark on the net and the chunk so we can clear it out when the sack arrives forcing it to retran just like it was "new" this improves the handling of window probes, which were dropped by the receiver. - Tighten AUTH protocol error checks during INIT/INIT-ACK exchange
2007-03-31 11:47:30 +00:00
}
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
sctp_skip_csum_4:
Fix the the SCTP_WITH_NO_CSUM option when used in combination with interface supporting CRC offload. While at it, make use of the feature that the loopback interface provides CRC offloading. MFC after: 4 weeks
2010-08-29 18:50:30 +00:00
#endif
- Found bug in min split point bundling which caused incorrect, non-bundlable fragmentation. - Added min residual to better control split points for both how big a msg must be as well as how much needs to be left over. - With our new algo in place, we need to implicitly set "end of msg" on the sp-> structure otherwise we end up with "hung" associations. - Room reserved up front in IP header by pushing IP header to back of mbuf. - Fix so FR's peg count of retransmissions needed. - Fix so an unlucky chunk that never gets across will kill the assoc via the kill timer and send an abort too. - Fix bug in sctp_input which can result in a crash. - Do not strip off IP options anymore. - Clean up sctp_calculate_rto(). - Get rid of unused sysctl. - Fixed so we discard all M-Cast - Fixed so port check done AFTER checksum - Fixed bug in fragmentation code that prevented us from fragmenting a small complete message when we needed to. - Window probes were not marked back to unsent and flight adjusted when a sack came in with no window change or accepting of the probe data. We now fix this with having a mark on the net and the chunk so we can clear it out when the sack arrives forcing it to retran just like it was "new" this improves the handling of window probes, which were dropped by the receiver. - Tighten AUTH protocol error checks during INIT/INIT-ACK exchange
2007-03-31 11:47:30 +00:00
/* destination port of 0 is illegal, based on RFC2960. */
if (sh->dest_port == 0) {
SCTP_STAT_INCR(sctps_hdrops);
goto bad;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
}
/*
* Locate pcb and tcb for datagram sctp_findassociation_addr() wants
* IP/SCTP/first chunk header...
*/
stcb = sctp_findassociation_addr(m, iphlen, offset - sizeof(*ch),
- fixed several places where we did not release INP locks. - fixed a refcount bug in the new ifa structures. - use vrf's from default stcb or inp whenever possible. - Address limits raised to account for a full IP fragmented packet (1000 addresses). - flight size correcting updated to include one message only and to handle case where the peer does not cumack the next segment aka lists 1/1 in sack blocks.. - Various bad init/init-ack handling could cause a panic since we tried to unlock the destroyed mutex. Fixes so we properly exit when we need to destroy an assoc. (Found by Cisco DevTest team :D) - name rename in src-addr-selection from pass to sifa. - route structure typedef'd to allow different platforms and updated into sctp_os_bsd file. - Max retransmissions a chunk can be made added. Reviewed by: gnn
2007-04-03 11:15:32 +00:00
sh, ch, &inp, &net, vrf_id);
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
if ((net) && (port)) {
if (net->port == 0) {
sctp_pathmtu_adjustment(inp, stcb, net, net->mtu - sizeof(struct udphdr));
}
net->port = port;
}
Add support for M_FLOWID.
2011-02-05 19:13:38 +00:00
if ((net != NULL) && (m->m_flags & M_FLOWID)) {
net->flowid = m->m_pkthdr.flowid;
Fix bugs related to M_FLOWID: * Store the flowid when receiving an SCTP/IPv6 packet. * Store the flowid when receiving an SCTP packet with wrong CRC. * Initilize flowid correctly. * Put test code under INVARIANTS. MFC after: 3 months.
2011-02-07 15:04:23 +00:00
#ifdef INVARIANTS
1) Track when flowid does get set. MFC after: 3 months
2011-02-07 08:10:29 +00:00
net->flowidset = 1;
Fix bugs related to M_FLOWID: * Store the flowid when receiving an SCTP/IPv6 packet. * Store the flowid when receiving an SCTP packet with wrong CRC. * Initilize flowid correctly. * Put test code under INVARIANTS. MFC after: 3 months.
2011-02-07 15:04:23 +00:00
#endif
Add support for M_FLOWID.
2011-02-05 19:13:38 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* inp's ref-count increased && stcb locked */
if (inp == NULL) {
struct sctp_init_chunk *init_chk, chunk_buf;
SCTP_STAT_INCR(sctps_noport);
#ifdef ICMP_BANDLIM
/*
* we use the bandwidth limiting to protect against sending
* too many ABORTS all at once. In this case these count the
* same as an ICMP message.
*/
if (badport_bandlim(0) < 0)
goto bad;
#endif /* ICMP_BANDLIM */
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
SCTPDBG(SCTP_DEBUG_INPUT1,
"Sending a ABORT from packet entry!\n");
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if (ch->chunk_type == SCTP_INITIATION) {
/*
* we do a trick here to get the INIT tag, dig in
* and get the tag from the INIT and put it in the
* common header.
*/
init_chk = (struct sctp_init_chunk *)sctp_m_getptr(m,
iphlen + sizeof(*sh), sizeof(*init_chk),
(uint8_t *) & chunk_buf);
if (init_chk != NULL)
sh->v_tag = init_chk->init.initiate_tag;
}
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
if (ch->chunk_type == SCTP_SHUTDOWN_ACK) {
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sctp_send_shutdown_complete2(m, iphlen, sh, vrf_id, port);
1) Fixes on a number of different collision case LOR's. 2) Fix all "magic numbers" to be constants. 3) A collision case that would generate two associations to the same peer due to a missing lock is fixed. 4) Added tracking of where timers are stopped. Approved by: gnn
2006-12-14 17:02:55 +00:00
goto bad;
}
if (ch->chunk_type == SCTP_SHUTDOWN_COMPLETE) {
goto bad;
}
if (ch->chunk_type != SCTP_ABORT_ASSOCIATION)
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
sctp_send_abort(m, iphlen, sh, 0, NULL, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
goto bad;
} else if (stcb == NULL) {
refcount_up = 1;
}
Commit the change from FAST_IPSEC to IPSEC. The FAST_IPSEC option is now deprecated, as well as the KAME IPsec code. What was FAST_IPSEC is now IPSEC. Approved by: re Sponsored by: Secure Computing
2007-07-03 12:13:45 +00:00
#ifdef IPSEC
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/*
* I very much doubt any of the IPSEC stuff will work but I have no
* idea, so I will leave it in place.
*/
Tons of fixes to get all the 64bit issues removed. This also moves two 16 bit int's to become 32 bit values so we do not have to use atomic_add_16. Most of the changes are %p, casts and other various nasty's that were in the orignal code base. With this commit my machine will now do a build universe.. however I as yet have not tested on a 64bit machine .. it may not work :-(
2006-11-05 13:25:18 +00:00
if (inp && ipsec4_in_reject(m, &inp->ip_inp.inp)) {
Support for VNET in SCTP (hopefully)
2009-09-17 15:11:12 +00:00
MODULE_GLOBAL(ipsec4stat).in_polvio++;
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_STAT_INCR(sctps_hdrops);
goto bad;
}
#endif /* IPSEC */
/*
* common chunk processing
*/
length = ip->ip_len + iphlen;
offset -= sizeof(struct sctp_chunkhdr);
ecn_bits = ip->ip_tos;
- Fixed 1-2-1 model to not worry about associd in sockopts - Fixed RTOinfo for bounding. - Fixed connect() to return ECONNREFUSED when an ABORT is received. - Added comments to direct Static Analysis not to look at some things it does not understand (comments are /* sa_ignore XXXXX */) - Bind when colliding was broken, missing not_found = 1 before checking to see if the port was in use caused endless bind loop. - Cookie life needs to be in milliseconds to conform to socket api. - Cookie life is not supposed to change if its 0, On the assoc level set we changed it to 0 opps. - Two more static analysis issues identified by the cisco tool. Null checks needed. - An issue for sendfile(). Need to validate the correct input argument. - When sending failed due to a no route to host, we leaked the mbuf chain failing to call m_freem(). - Fix #ifdef issue for getting hash block len when HAVE_SHA2 is NOT defined Reviewed by: gnn
2007-05-17 12:16:24 +00:00
/* sa_ignore NO_NULL_CHK */
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
sctp_common_input_processing(&m, iphlen, offset, length, sh, ch,
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
inp, stcb, net, ecn_bits, vrf_id, port);
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
/* inp's ref-count reduced && stcb unlocked */
if (m) {
sctp_m_freem(m);
}
if ((inp) && (refcount_up)) {
/* reduce ref-count */
SCTP_INP_DECR_REF(inp);
}
return;
bad:
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
if (stcb) {
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
SCTP_TCB_UNLOCK(stcb);
Two major items here: - All printf that was surrounded by #ifdef SCTP_DEBUG moves to a macro that does all of this. This removes all printfs from the code and makes the code more portable and easier to read. - Static Analysis (cisco) - found a few bugs, but mostly we add checks for NULL pointers and such to make the tool happy. We now pass the Cisco SA tools checks except for where it does not understand tailq/lists. We still need to look at the coverity tools output too (this is like the cisco SA tool) and see if it wants us to fix any other items. Hopefully this will be the last major churn in the code other than bug fixes.
2007-05-09 13:30:06 +00:00
}
Ok, here it is, we finally add SCTP to current. Note that this work is not just mine, but it is also the works of Peter Lei and Michael Tuexen. They both are my two key other developers working on the project.. and they need ata-boy's too: **** peterlei@cisco.com tuexen@fh-muenster.de **** I did do a make sysent which updated the syscall's and sysproto.. I hope that is correct... without it you don't build since we have new syscalls for SCTP :-0 So go out and look at the NOTES, add option SCTP (make sure inet and inet6 are present too) and play with SCTP. I will see about comitting some test tools I have after I figure out where I should place them. I also have a lib (libsctp.a) that adds some of the missing socketapi functions that I need to put into lib's.. I will talk to George about this :-) There may still be some 64 bit issues in here, none of us have a 64 bit processor to test with yet.. Michael may have a MAC but thats another beast too.. If you have a mac and want to use SCTP contact Michael he maintains a web site with a loadable module with this code :-) Reviewed by: gnn Approved by: gnn
2006-11-03 15:23:16 +00:00
if ((inp) && (refcount_up)) {
/* reduce ref-count */
SCTP_INP_DECR_REF(inp);
}
if (m) {
sctp_m_freem(m);
}
return;
}
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
#if defined(__FreeBSD__) && defined(SCTP_MCORE_INPUT) && defined(SMP)
extern int *sctp_cpuarry;
#endif
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
void
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
sctp_input(struct mbuf *m, int off)
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
{
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
#if defined(__FreeBSD__) && defined(SCTP_MCORE_INPUT) && defined(SMP)
struct ip *ip;
struct sctphdr *sh;
int offset;
int cpu_to_use;
1) Use same scheme Michael and I discussed for a selected for a flowid 2) If flowid is not set, arrange so it is stored. 3) If flowid is set by lower layer, use it. MFC after: 3 Months
2011-02-06 13:17:40 +00:00
uint32_t flowid, tag;
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
if (mp_ncpus > 1) {
1) Use same scheme Michael and I discussed for a selected for a flowid 2) If flowid is not set, arrange so it is stored. 3) If flowid is set by lower layer, use it. MFC after: 3 Months
2011-02-06 13:17:40 +00:00
if (m->m_flags & M_FLOWID) {
flowid = m->m_pkthdr.flowid;
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
} else {
/*
1) Use same scheme Michael and I discussed for a selected for a flowid 2) If flowid is not set, arrange so it is stored. 3) If flowid is set by lower layer, use it. MFC after: 3 Months
2011-02-06 13:17:40 +00:00
* No flow id built by lower layers fix it so we
* create one.
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
*/
1) Use same scheme Michael and I discussed for a selected for a flowid 2) If flowid is not set, arrange so it is stored. 3) If flowid is set by lower layer, use it. MFC after: 3 Months
2011-02-06 13:17:40 +00:00
ip = mtod(m, struct ip *);
offset = off + sizeof(*sh);
if (SCTP_BUF_LEN(m) < offset) {
if ((m = m_pullup(m, offset)) == 0) {
SCTP_STAT_INCR(sctps_hdrops);
return;
}
ip = mtod(m, struct ip *);
}
sh = (struct sctphdr *)((caddr_t)ip + off);
tag = htonl(sh->v_tag);
flowid = tag ^ ntohs(sh->dest_port) ^ ntohs(sh->src_port);
m->m_pkthdr.flowid = flowid;
m->m_flags |= M_FLOWID;
1) Fix cpu mapping per JB's suggestions 2) Fix it so INIT's don't always end up on CPU0 MFC after: 3 months
2011-02-04 13:50:30 +00:00
}
1) Use same scheme Michael and I discussed for a selected for a flowid 2) If flowid is not set, arrange so it is stored. 3) If flowid is set by lower layer, use it. MFC after: 3 Months
2011-02-06 13:17:40 +00:00
cpu_to_use = sctp_cpuarry[flowid % mp_ncpus];
Adds an experimental option to create a pool of threads. These serve as input threads and are queued packets based on the V-tag number. This is similar to what a modern card can do with queue's for TCP... but alas modern cards know nothing about SCTP. MFC after: 3 months (maybe)
2011-02-03 10:05:30 +00:00
sctp_queue_to_mcore(m, off, cpu_to_use);
return;
}
#endif
sctp_input_with_port(m, off, 0);
- Adds support for the multi-asconf (From Kozuka-san) - Adds some prepwork (Not all yet) for vimage in particular support the delete the sctppcbinfo.xx structs. There is still a leak in here if it were to be called plus we stil need the regrouping (From Me and Michael Tuexen) - Adds support for UDP tunneling. For BSD there is no socket yet setup so its disabled, but major argument changes are in here to emcompass the passing of the port number (zero when you don't have a udp tunnel, the default for BSD). Will add some hooks in UDP here shortly (discussed with Robert) that will allow easy tunneling. (Mainly from Peter Lei and Michael Tuexen with some BSD work from me :-D) - Some ease for windows, evidently leave is reserved by their compile move label leave: -> out: MFC after: 1 week
2008-05-20 13:47:46 +00:00
}
Improve compilation of SCTP code without INET support. Some bugs where fixed while doing this: * ASCONF-ACK messages might use wrong port number when using IPv6. * Checking for additional addresses takes the correct address into account and also does not do more comparisons than necessary. This patch is based on one received from bz@ who was sponsored by The FreeBSD Foundation and iXsystems. MFC after: 1 week
2011-04-30 11:18:16 +00:00
#endif
Reference in New Issue Copy Permalink