d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
9e88f47c8f
freebsd-dev/sys/net/rtsock.c

2566 lines
61 KiB
C
Raw Normal View History

/* -> /*- for license, minor formatting changes
2005-01-07 01:45:51 +00:00
/*-
sys: further adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 3-Clause license. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts. Special thanks to Wind River for providing access to "The Duke of Highlander" tool: an older (2014) run over FreeBSD tree was useful as a starting point.
2017-11-20 19:43:44 +00:00
* SPDX-License-Identifier: BSD-3-Clause
*
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* Copyright (c) 1988, 1991, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
Renumber copyright clause 4 Renumber cluase 4 to 3, per what everybody else did when BSD granted them permission to remove clause 3. My insistance on keeping the same numbering for legal reasons is too pedantic, so give up on that point. Submitted by: Jan Schaumann <jschauma@stevens.edu> Pull Request: https://github.com/freebsd/freebsd/pull/96
2017-02-28 23:42:47 +00:00
* 3. Neither the name of the University nor the names of its contributors
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
*
Bring in latest CSRG revisions to this file: - Report destination address of a P2P link when servicing routing socket messages. - Report interface name, address, and destination address of a P2P link when servicing NET_RT_{DUMP,FLAGS} sysctls. Part of CSRG revision 8.6 coresponds to revision 1.12. CSRG revision 8.7 corresponds to revision 1.15.
2001-10-17 11:23:59 +00:00
* @(#)rtsock.c 8.7 (Berkeley) 10/12/95
$Id$ -> $FreeBSD$
1999-08-28 01:08:13 +00:00
* $FreeBSD$
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
#include "opt_ddb.h"
This patch provides the back end support for equal-cost multi-path (ECMP) for both IPv4 and IPv6. Previously, multipath route insertion is disallowed. For example, route add -net 192.103.54.0/24 10.9.44.1 route add -net 192.103.54.0/24 10.9.44.2 The second route insertion will trigger an error message of "add net 192.103.54.0/24: gateway 10.2.5.2: route already in table" Multiple default routes can also be inserted. Here is the netstat output: default 10.2.5.1 UGS 0 3074 bge0 => default 10.2.5.2 UGS 0 0 bge0 When multipath routes exist, the "route delete" command requires a specific gateway to be specified or else an error message would be displayed. For example, route delete default would fail and trigger the following error message: "route: writing to routing socket: No such process" "delete net default: not in table" On the other hand, route delete default 10.2.5.2 would be successful: "delete net default: gateway 10.2.5.2" One does not have to specify a gateway if there is only a single route for a particular destination. I need to perform more testings on address aliases and multiple interfaces that have the same IP prefixes. This patch as it stands today is not yet ready for prime time. Therefore, the ECMP code fragments are fully guarded by the RADIX_MPATH macro. Include the "options RADIX_MPATH" in the kernel configuration to enable this feature. Reviewed by: robert, sam, gnn, julian, kmacy
2008-04-13 05:45:14 +00:00
#include "opt_mpath.h"
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
#include "opt_inet.h"
#include "opt_inet6.h"
This patch provides the back end support for equal-cost multi-path (ECMP) for both IPv4 and IPv6. Previously, multipath route insertion is disallowed. For example, route add -net 192.103.54.0/24 10.9.44.1 route add -net 192.103.54.0/24 10.9.44.2 The second route insertion will trigger an error message of "add net 192.103.54.0/24: gateway 10.2.5.2: route already in table" Multiple default routes can also be inserted. Here is the netstat output: default 10.2.5.1 UGS 0 3074 bge0 => default 10.2.5.2 UGS 0 0 bge0 When multipath routes exist, the "route delete" command requires a specific gateway to be specified or else an error message would be displayed. For example, route delete default would fail and trigger the following error message: "route: writing to routing socket: No such process" "delete net default: not in table" On the other hand, route delete default 10.2.5.2 would be successful: "delete net default: gateway 10.2.5.2" One does not have to specify a gateway if there is only a single route for a particular destination. I need to perform more testings on address aliases and multiple interfaces that have the same IP prefixes. This patch as it stands today is not yet ready for prime time. Therefore, the ECMP code fragments are fully guarded by the RADIX_MPATH macro. Include the "options RADIX_MPATH" in the kernel configuration to enable this feature. Reviewed by: robert, sam, gnn, julian, kmacy
2008-04-13 05:45:14 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/jail.h>
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
#include <sys/kernel.h>
Introduce and use a sysinit-based initialization scheme for virtual network stacks, VNET_SYSINIT: - Add VNET_SYSINIT and VNET_SYSUNINIT macros to declare events that will occur each time a network stack is instantiated and destroyed. In the !VIMAGE case, these are simply mapped into regular SYSINIT/SYSUNINIT. For the VIMAGE case, we instead use SYSINIT's to track their order and properties on registration, using them for each vnet when created/ destroyed, or immediately on module load for already-started vnets. - Remove vnet_modinfo mechanism that existed to serve this purpose previously, as well as its dependency scheme: we now just use the SYSINIT ordering scheme. - Implement VNET_DOMAIN_SET() to allow protocol domains to declare that they want init functions to be called for each virtual network stack rather than just once at boot, compiling down to DOMAIN_SET() in the non-VIMAGE case. - Walk all virtualized kernel subsystems and make use of these instead of modinfo or DOMAIN_SET() for init/uninit events. In some cases, convert modular components from using modevent to using sysinit (where appropriate). In some cases, do minor rejuggling of SYSINIT ordering to make room for or better manage events. Portions submitted by: jhb (VNET_SYSINIT), bz (cleanup) Discussed with: jhb, bz, julian, zec Reviewed by: bz Approved by: re (VIMAGE blanket)
2009-07-23 20:46:49 +00:00
#include <sys/domain.h>
Add missing include to sys/lock.h before sys/rwlock.h
2008-12-08 00:28:21 +00:00
#include <sys/lock.h>
Added used #include - don't depend on <sys/mbuf.h> including <sys/malloc.h> (unless we only use the bogusly shared M*WAIT flags).
1997-09-02 01:19:47 +00:00
#include <sys/malloc.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/mbuf.h>
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
#include <sys/priv.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/proc.h>
#include <sys/protosw.h>
Switch RIB and RADIX_NODE_HEAD lock from rwlock(9) to rmlock(9). Using of rwlock with multiqueue NICs for IP forwarding on high pps produces high lock contention and inefficient. Rmlock fits better for such workloads. Reviewed by: melifaro, olivier Obtained from: Yandex LLC Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15789
2018-06-16 08:26:23 +00:00
#include <sys/rmlock.h>
- convert radix node head lock from mutex to rwlock - make radix node head lock not recursive - fix LOR in rtexpunge - fix LOR in rtredirect Reviewed by: sam
2008-12-07 21:15:43 +00:00
#include <sys/rwlock.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/signalvar.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/socket.h>
#include <sys/socketvar.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/sysctl.h>
#include <sys/systm.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
#ifdef DDB
#include <ddb/ddb.h>
#include <ddb/db_lex.h>
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <net/if.h>
The r48589 promised to remove implicit inclusion of if_var.h soon. Prepare to this event, adding if_var.h to files that do need it. Also, include all includes that now are included due to implicit pollution via if_var.h Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-26 17:58:36 +00:00
#include <net/if_var.h>
This checkin addresses a couple of issues: 1. The "route" command allows route insertion through the interface-direct option "-iface". During if_attach(), an sockaddr_dl{} entry is created for the interface and is part of the interface address list. This sockaddr_dl{} entry describes the interface in detail. The "route" command selects this entry as the "gateway" object when the "-iface" option is present. The "arp" and "ndp" commands also interact with the kernel through the routing socket when adding and removing static L2 entries. The static L2 information is also provided through the "gateway" object with an AF_LINK family type, similar to what is provided by the "route" command. In order to differentiate between these two types of operations, a RTF_LLDATA flag is introduced. This flag is set by the "arp" and "ndp" commands when issuing the add and delete commands. This flag is also set in each L2 entry returned by the kernel. The "arp" and "ndp" command follows a convention where a RTM_GET is issued first followed by a RTM_ADD/DELETE. This RTM_GET request fills in the fields for a "rtm" object, which is reinjected into the kernel by a subsequent RTM_ADD/DELETE command. The entry returend from RTM_GET is a prefix route, so the RTF_LLDATA flag must be specified when issuing the RTM_ADD/DELETE messages. 2. Enforce the convention that NET_RT_FLAGS with a 0 w_arg is the specification for retrieving L2 information. Also optimized the code logic. Reviewed by: julian
2008-12-26 19:45:24 +00:00
#include <net/if_dl.h>
This main goals of this project are: 1. separating L2 tables (ARP, NDP) from the L3 routing tables 2. removing as much locking dependencies among these layers as possible to allow for some parallelism in the search operations 3. simplify the logic in the routing code, The most notable end result is the obsolescent of the route cloning (RTF_CLONING) concept, which translated into code reduction in both IPv4 ARP and IPv6 NDP related modules, and size reduction in struct rtentry{}. The change in design obsoletes the semantics of RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland applications such as "arp" and "ndp" have been modified to reflect those changes. The output from "netstat -r" shows only the routing entries. Quite a few developers have contributed to this project in the past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and Andre Oppermann. And most recently: - Kip Macy revised the locking code completely, thus completing the last piece of the puzzle, Kip has also been conducting active functional testing - Sam Leffler has helped me improving/refactoring the code, and provided valuable reviews - Julian Elischer setup the perforce tree for me and has helped me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
#include <net/if_llatbl.h>
This patch fixes the problem where proxy ARP entries cannot be added over the if_ng interface. MFC after: 3 days
2010-05-25 20:42:35 +00:00
#include <net/if_types.h>
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
#include <net/netisr.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <net/raw_cb.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <net/route.h>
MFP r287070,r287073: split radix implementation and route table structure. There are number of radix consumers in kernel land (pf,ipfw,nfs,route) with different requirements. In fact, first 3 don't have _any_ requirements and first 2 does not use radix locking. On the other hand, routing structure do have these requirements (rnh_gen, multipath, custom to-be-added control plane functions, different locking). Additionally, radix should not known anything about its consumers internals. So, radix code now uses tiny 'struct radix_head' structure along with internal 'struct radix_mask_head' instead of 'struct radix_node_head'. Existing consumers still uses the same 'struct radix_node_head' with slight modifications: they need to pass pointer to (embedded) 'struct radix_head' to all radix callbacks. Routing code now uses new 'struct rib_head' with different locking macro: RADIX_NODE_HEAD prefix was renamed to RIB_ (which stands for routing information base). New net/route_var.h header was added to hold routing subsystem internal data. 'struct rib_head' was placed there. 'struct rtentry' will also be moved there soon.
2016-01-25 06:33:15 +00:00
#include <net/route_var.h>
Rather than using hidden includes (with cicular dependencies), directly include only the header files needed. This reduces the unneeded spamming of various headers into lots of files. For now, this leaves us with very few modules including vnet.h and thus needing to depend on opt_route.h. Reviewed by: brooks, gnn, des, zec, imp Sponsored by: The FreeBSD Foundation
2008-12-02 21:37:28 +00:00
#include <net/vnet.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Give jail(8) the feature to allow raw sockets from within a jail, which is less restrictive but allows for more flexible jail usage (for those who are willing to make the sacrifice). The default is off, but allowing raw sockets within jails can now be accomplished by tuning security.jail.allow_raw_sockets to 1. Turning this on will allow you to use things like ping(8) or traceroute(8) from within a jail. The patch being committed is not identical to the patch in the PR. The committed version is more friendly to APIs which pjd is working on, so it should integrate into his work quite nicely. This change has also been presented and addressed on the freebsd-hackers mailing list. Submitted by: Christian S.J. Peron <maneo@bsdpro.com> PR: kern/65800
2004-04-26 19:46:52 +00:00
#include <netinet/in.h>
The proxy arp entries could not be added into the system over the IFF_POINTOPOINT link types. The reason was due to the routing entry returned from the kernel covering the remote end is of an interface type that does not support ARP. This patch fixes this problem by providing a hint to the kernel routing code, which indicates the prefix route instead of the PPP host route should be returned to the caller. Since a host route to the local end point is also added into the routing table, and there could be multiple such instantiations due to multiple PPP links can be created with the same local end IP address, this patch also fixes the loopback route installation failure problem observed prior to this patch. The reference count of loopback route to local end would be either incremented or decremented. The first instantiation would create the entry and the last removal would delete the route entry. MFC after: 5 days
2009-12-30 21:35:34 +00:00
#include <netinet/if_ether.h>
A major overhaul of the CARP implementation. The ip_carp.c was started from scratch, copying needed functionality from the old implemenation on demand, with a thorough review of all code. The main change is that interface layer has been removed from the CARP. Now redundant addresses are configured exactly on the interfaces, they run on. The CARP configuration itself is, as before, configured and read via SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or SIOCAIFADDR_IN6 may now be configured to a particular virtual host id, which makes the prefix redundant. ifconfig(8) semantics has been changed too: now one doesn't need to clone carpXX interface, he/she should directly configure a vhid on a Ethernet interface. To supply vhid data from the kernel to an application the getifaddrs(8) function had been changed to pass ifam_data with each address. [1] The new implementation definitely closes all PRs related to carp(4) being an interface, and may close several others. It also allows to run a single redundant IP per interface. Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for idea on using ifam_data and for several rounds of reviewing! PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448 Reviewed by: bz Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
#include <netinet/ip_carp.h>
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
#ifdef INET6
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#include <netinet6/ip6_var.h>
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
#include <netinet6/scope6_var.h>
#endif
Introduce nexthop objects and new routing KPI. This is the foundational change for the routing subsytem rearchitecture. More details and goals are available in https://reviews.freebsd.org/D24141 . This patch introduces concept of nexthop objects and new nexthop-based routing KPI. Nexthops are objects, containing all necessary information for performing the packet output decision. Output interface, mtu, flags, gw address goes there. For most of the cases, these objects will serve the same role as the struct rtentry is currently serving. Typically there will be low tens of such objects for the router even with multiple BGP full-views, as these objects will be shared between routing entries. This allows to store more information in the nexthop. New KPI: struct nhop_object *fib4_lookup(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, uint32_t flowid); struct nhop_object *fib6_lookup(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, uint32_t flowid); These 2 function are intended to replace all all flavours of <in_|in6_>rtalloc[1]<_ign><_fib>, mpath functions and the previous fib[46]-generation functions. Upon successful lookup, they return nexthop object which is guaranteed to exist within current NET_EPOCH. If longer lifetime is desired, one can specify NHR_REF as a flag and get a referenced version of the nexthop. Reference semantic closely resembles rtentry one, allowing sed-style conversion. Additionally, another 2 functions are introduced to support uRPF functionality inside variety of our firewalls. Their primary goal is to hide the multipath implementation details inside the routing subsystem, greatly simplifying firewalls implementation: int fib4_lookup_urpf(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); int fib6_lookup_urpf(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); All functions have a separate scopeid argument, paving way to eliminating IPv6 scope embedding and allowing to support IPv4 link-locals in the future. Structure changes: * rtentry gets new 'rt_nhop' pointer, slightly growing the overall size. * rib_head gets new 'rnh_preadd' callback pointer, slightly growing overall sz. Old KPI: During the transition state old and new KPI will coexists. As there are another 4-5 decent-sized conversion patches, it will probably take a couple of weeks. To support both KPIs, fields not required by the new KPI (most of rtentry) has to be kept, resulting in the temporary size increase. Once conversion is finished, rtentry will notably shrink. More details: * architectural overview: https://reviews.freebsd.org/D24141 * list of the next changes: https://reviews.freebsd.org/D24232 Reviewed by: ae,glebius(initial version) Differential Revision: https://reviews.freebsd.org/D24232
2020-04-12 14:30:00 +00:00
#include <net/route/nhop.h>
Remove an duplicate definition of nhops_dump_sysctl() One of the source files included both nhop.h and shared.h, leading to this clashing. Tested with: mips-gcc cross toolchain
2020-04-16 23:28:47 +00:00
#include <net/route/shared.h>
Give jail(8) the feature to allow raw sockets from within a jail, which is less restrictive but allows for more flexible jail usage (for those who are willing to make the sacrifice). The default is off, but allowing raw sockets within jails can now be accomplished by tuning security.jail.allow_raw_sockets to 1. Turning this on will allow you to use things like ping(8) or traceroute(8) from within a jail. The patch being committed is not identical to the patch in the PR. The committed version is more friendly to APIs which pjd is working on, so it should integrate into his work quite nicely. This change has also been presented and addressed on the freebsd-hackers mailing list. Submitted by: Christian S.J. Peron <maneo@bsdpro.com> PR: kern/65800
2004-04-26 19:46:52 +00:00
Provide 32bit compat shims for sysctl net.route NET_RT_IFLIST. This allows getifaddrs(3) to work for compat32 binaries. Submitted by: jhb (6.x version) Reviewed by: emaste Tested by: emaste and <pluknet gmail com> MFC after: 2 weeks
2010-04-25 16:42:47 +00:00
#ifdef COMPAT_FREEBSD32
#include <sys/mount.h>
#include <compat/freebsd32/freebsd32.h>
struct if_msghdr32 {
uint16_t ifm_msglen;
uint8_t ifm_version;
uint8_t ifm_type;
int32_t ifm_addrs;
int32_t ifm_flags;
uint16_t ifm_index;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
uint16_t _ifm_spare1;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data ifm_data;
Provide 32bit compat shims for sysctl net.route NET_RT_IFLIST. This allows getifaddrs(3) to work for compat32 binaries. Submitted by: jhb (6.x version) Reviewed by: emaste Tested by: emaste and <pluknet gmail com> MFC after: 2 weeks
2010-04-25 16:42:47 +00:00
};
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
struct if_msghdrl32 {
uint16_t ifm_msglen;
uint8_t ifm_version;
uint8_t ifm_type;
int32_t ifm_addrs;
int32_t ifm_flags;
uint16_t ifm_index;
uint16_t _ifm_spare1;
uint16_t ifm_len;
uint16_t ifm_data_off;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
uint32_t _ifm_spare2;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data ifm_data;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
};
struct ifa_msghdrl32 {
uint16_t ifam_msglen;
uint8_t ifam_version;
uint8_t ifam_type;
int32_t ifam_addrs;
int32_t ifam_flags;
uint16_t ifam_index;
uint16_t _ifam_spare1;
uint16_t ifam_len;
uint16_t ifam_data_off;
int32_t ifam_metric;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data ifam_data;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
};
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
#define SA_SIZE32(sa) \
( (((struct sockaddr *)(sa))->sa_len == 0) ? \
sizeof(int) : \
1 + ( (((struct sockaddr *)(sa))->sa_len - 1) | (sizeof(int) - 1) ) )
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#endif /* COMPAT_FREEBSD32 */
Provide 32bit compat shims for sysctl net.route NET_RT_IFLIST. This allows getifaddrs(3) to work for compat32 binaries. Submitted by: jhb (6.x version) Reviewed by: emaste Tested by: emaste and <pluknet gmail com> MFC after: 2 weeks
2010-04-25 16:42:47 +00:00
Last major round (Unless Bruce thinks of somthing :-) of malloc changes. Distribute all but the most fundamental malloc types. This time I also remembered the trick to making things static: Put "static" in front of them. A couple of finer points by: bde
1997-10-12 20:26:33 +00:00
MALLOC_DEFINE(M_RTABLE, "routetbl", "routing tables");
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
/* NB: these are not modified */
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
static struct sockaddr route_src = { 2, PF_ROUTE, };
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
static struct sockaddr sa_zero = { sizeof(sa_zero), AF_INET, };
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
A major overhaul of the CARP implementation. The ip_carp.c was started from scratch, copying needed functionality from the old implemenation on demand, with a thorough review of all code. The main change is that interface layer has been removed from the CARP. Now redundant addresses are configured exactly on the interfaces, they run on. The CARP configuration itself is, as before, configured and read via SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or SIOCAIFADDR_IN6 may now be configured to a particular virtual host id, which makes the prefix redundant. ifconfig(8) semantics has been changed too: now one doesn't need to clone carpXX interface, he/she should directly configure a vhid on a Ethernet interface. To supply vhid data from the kernel to an application the getifaddrs(8) function had been changed to pass ifam_data with each address. [1] The new implementation definitely closes all PRs related to carp(4) being an interface, and may close several others. It also allows to run a single redundant IP per interface. Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for idea on using ifam_data and for several rounds of reviewing! PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448 Reviewed by: bz Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
/* These are external hooks for CARP. */
int (*carp_get_vhid_p)(struct ifaddr *);
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
/*
* Used by rtsock/raw_input callback code to decide whether to filter the update
* notification to a socket bound to a particular FIB.
*/
#define RTS_FILTER_FIB M_PROTO8
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
typedef struct {
Fix a typo (attacked -> attached). Approved by: sam
2004-08-24 08:47:15 +00:00
int ip_count; /* attached w/ AF_INET */
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
int ip6_count; /* attached w/ AF_INET6 */
int any_count; /* total attached */
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
} route_cb_t;
Use the new VNET_DEFINE_STATIC macro when we are defining static VNET variables. Reviewed by: bz Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D16147
2018-07-24 16:35:52 +00:00
VNET_DEFINE_STATIC(route_cb_t, route_cb);
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
#define V_route_cb VNET(route_cb)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
trivial locking rtsock_cb Sponsored by: FreeBSD Foundation
2003-10-03 18:27:02 +00:00
struct mtx rtsock_mtx;
MTX_SYSINIT(rtsock, &rtsock_mtx, "rtsock route_cb lock", MTX_DEF);
#define RTSOCK_LOCK() mtx_lock(&rtsock_mtx)
#define RTSOCK_UNLOCK() mtx_unlock(&rtsock_mtx)
#define RTSOCK_LOCK_ASSERT() mtx_assert(&rtsock_mtx, MA_OWNED)
Mark more nodes as CTLFLAG_MPSAFE or CTLFLAG_NEEDGIANT (17 of many) r357614 added CTLFLAG_NEEDGIANT to make it easier to find nodes that are still not MPSAFE (or already are but aren’t properly marked). Use it in preparation for a general review of all nodes. This is non-functional change that adds annotations to SYSCTL_NODE and SYSCTL_PROC nodes using one of the soon-to-be-required flags. Mark all obvious cases as MPSAFE. All entries that haven't been marked as MPSAFE before are by default marked as NEEDGIANT Approved by: kib (mentor, blanket) Commented by: kib, gallatin, melifaro Differential Revision: https://reviews.freebsd.org/D23718
2020-02-26 14:26:36 +00:00
static SYSCTL_NODE(_net, OID_AUTO, route, CTLFLAG_RD | CTLFLAG_MPSAFE, 0,
"");
Allow the size of the routing socket netisr queue to be configured using the tunable or sysctl 'net.route.netisr_maxqlen'. Default the maximum depth to 256 rather than IFQ_MAXLEN due to the downsides of dropping routing messages. MT5 candidate. Discussed with: mdodd, mlaier, Vincent Jardin <jardin at 6wind.com>
2004-08-21 21:20:06 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct walkarg {
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
int w_tmemsize;
int w_op, w_arg;
caddr_t w_tmem;
struct sysctl_req *w_req;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
static void rts_input(struct mbuf *m);
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
static struct mbuf *rtsock_msg_mbuf(int type, struct rt_addrinfo *rtinfo);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
static int rtsock_msg_buffer(int type, struct rt_addrinfo *rtinfo,
struct walkarg *w, int *plen);
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
static int rt_xaddrs(caddr_t cp, caddr_t cplim,
struct rt_addrinfo *rtinfo);
Remove __P.
2002-03-19 21:54:18 +00:00
static int sysctl_dumpentry(struct radix_node *rn, void *vw);
static int sysctl_iflist(int af, struct walkarg *w);
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
static int sysctl_ifmalist(int af, struct walkarg *w);
Change pr_output's prototype to avoid the need for explicit casts. This is a follow up to r269699. Phabric: D564 Reviewed by: jhb
2014-08-15 02:43:02 +00:00
static int route_output(struct mbuf *m, struct socket *so, ...);
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
static void rt_getmetrics(const struct rtentry *rt, struct rt_metrics *out);
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
static void rt_dispatch(struct mbuf *, sa_family_t);
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
static struct sockaddr *rtsock_fix_netmask(struct sockaddr *dst,
struct sockaddr *smask, struct sockaddr_storage *dmask);
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
static int handle_rtm_get(struct rt_addrinfo *info, u_int fibnum,
struct rt_msghdr *rtm, struct rtentry **ret_nrt);
static int update_rtm_from_rte(struct rt_addrinfo *info,
struct rt_msghdr **prtm, int alloc_len,
struct rtentry *rt);
static void send_rtm_reply(struct socket *so, struct rt_msghdr *rtm,
struct mbuf *m, sa_family_t saf, u_int fibnum,
int rtm_errno);
static int can_export_rte(struct ucred *td_ucred, const struct rtentry *rt);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Reimplement the netisr framework in order to support parallel netisr threads: - Support up to one netisr thread per CPU, each processings its own workstream, or set of per-protocol queues. Threads may be bound to specific CPUs, or allowed to migrate, based on a global policy. In the future it would be desirable to support topology-centric policies, such as "one netisr per package". - Allow each protocol to advertise an ordering policy, which can currently be one of: NETISR_POLICY_SOURCE: packets must maintain ordering with respect to an implicit or explicit source (such as an interface or socket). NETISR_POLICY_FLOW: make use of mbuf flow identifiers to place work, as well as allowing protocols to provide a flow generation function for mbufs without flow identifers (m2flow). Falls back on NETISR_POLICY_SOURCE if now flow ID is available. NETISR_POLICY_CPU: allow protocols to inspect and assign a CPU for each packet handled by netisr (m2cpuid). - Provide utility functions for querying the number of workstreams being used, as well as a mapping function from workstream to CPU ID, which protocols may use in work placement decisions. - Add explicit interfaces to get and set per-protocol queue limits, and get and clear drop counters, which query data or apply changes across all workstreams. - Add a more extensible netisr registration interface, in which protocols declare 'struct netisr_handler' structures for each registered NETISR_ type. These include name, handler function, optional mbuf to flow ID function, optional mbuf to CPU ID function, queue limit, and ordering policy. Padding is present to allow these to be expanded in the future. If no queue limit is declared, then a default is used. - Queue limits are now per-workstream, and raised from the previous IFQ_MAXLEN default of 50 to 256. - All protocols are updated to use the new registration interface, and with the exception of netnatm, default queue limits. Most protocols register as NETISR_POLICY_SOURCE, except IPv4 and IPv6, which use NETISR_POLICY_FLOW, and will therefore take advantage of driver- generated flow IDs if present. - Formalize a non-packet based interface between interface polling and the netisr, rather than having polling pretend to be two protocols. Provide two explicit hooks in the netisr worker for start and end events for runs: netisr_poll() and netisr_pollmore(), as well as a function, netisr_sched_poll(), to allow the polling code to schedule netisr execution. DEVICE_POLLING still embeds single-netisr assumptions in its implementation, so for now if it is compiled into the kernel, a single and un-bound netisr thread is enforced regardless of tunable configuration. In the default configuration, the new netisr implementation maintains the same basic assumptions as the previous implementation: a single, un-bound worker thread processes all deferred work, and direct dispatch is enabled by default wherever possible. Performance measurement shows a marginal performance improvement over the old implementation due to the use of batched dequeue. An rmlock is used to synchronize use and registration/unregistration using the framework; currently, synchronized use is disabled (replicating current netisr policy) due to a measurable 3%-6% hit in ping-pong micro-benchmarking. It will be enabled once further rmlock optimization has taken place. However, in practice, netisrs are rarely registered or unregistered at runtime. A new man page for netisr will follow, but since one doesn't currently exist, it hasn't been updated. This change is not appropriate for MFC, although the polling shutdown handler should be merged to 7-STABLE. Bump __FreeBSD_version. Reviewed by: bz
2009-06-01 10:41:38 +00:00
static struct netisr_handler rtsock_nh = {
.nh_name = "rtsock",
.nh_handler = rts_input,
.nh_proto = NETISR_ROUTE,
.nh_policy = NETISR_POLICY_SOURCE,
};
static int
sysctl_route_netisr_maxqlen(SYSCTL_HANDLER_ARGS)
{
int error, qlimit;
netisr_getqlimit(&rtsock_nh, &qlimit);
error = sysctl_handle_int(oidp, &qlimit, 0, req);
if (error || !req->newptr)
return (error);
if (qlimit < 1)
return (EINVAL);
return (netisr_setqlimit(&rtsock_nh, qlimit));
}
Mark more nodes as CTLFLAG_MPSAFE or CTLFLAG_NEEDGIANT (17 of many) r357614 added CTLFLAG_NEEDGIANT to make it easier to find nodes that are still not MPSAFE (or already are but aren’t properly marked). Use it in preparation for a general review of all nodes. This is non-functional change that adds annotations to SYSCTL_NODE and SYSCTL_PROC nodes using one of the soon-to-be-required flags. Mark all obvious cases as MPSAFE. All entries that haven't been marked as MPSAFE before are by default marked as NEEDGIANT Approved by: kib (mentor, blanket) Commented by: kib, gallatin, melifaro Differential Revision: https://reviews.freebsd.org/D23718
2020-02-26 14:26:36 +00:00
SYSCTL_PROC(_net_route, OID_AUTO, netisr_maxqlen,
CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE,
Reimplement the netisr framework in order to support parallel netisr threads: - Support up to one netisr thread per CPU, each processings its own workstream, or set of per-protocol queues. Threads may be bound to specific CPUs, or allowed to migrate, based on a global policy. In the future it would be desirable to support topology-centric policies, such as "one netisr per package". - Allow each protocol to advertise an ordering policy, which can currently be one of: NETISR_POLICY_SOURCE: packets must maintain ordering with respect to an implicit or explicit source (such as an interface or socket). NETISR_POLICY_FLOW: make use of mbuf flow identifiers to place work, as well as allowing protocols to provide a flow generation function for mbufs without flow identifers (m2flow). Falls back on NETISR_POLICY_SOURCE if now flow ID is available. NETISR_POLICY_CPU: allow protocols to inspect and assign a CPU for each packet handled by netisr (m2cpuid). - Provide utility functions for querying the number of workstreams being used, as well as a mapping function from workstream to CPU ID, which protocols may use in work placement decisions. - Add explicit interfaces to get and set per-protocol queue limits, and get and clear drop counters, which query data or apply changes across all workstreams. - Add a more extensible netisr registration interface, in which protocols declare 'struct netisr_handler' structures for each registered NETISR_ type. These include name, handler function, optional mbuf to flow ID function, optional mbuf to CPU ID function, queue limit, and ordering policy. Padding is present to allow these to be expanded in the future. If no queue limit is declared, then a default is used. - Queue limits are now per-workstream, and raised from the previous IFQ_MAXLEN default of 50 to 256. - All protocols are updated to use the new registration interface, and with the exception of netnatm, default queue limits. Most protocols register as NETISR_POLICY_SOURCE, except IPv4 and IPv6, which use NETISR_POLICY_FLOW, and will therefore take advantage of driver- generated flow IDs if present. - Formalize a non-packet based interface between interface polling and the netisr, rather than having polling pretend to be two protocols. Provide two explicit hooks in the netisr worker for start and end events for runs: netisr_poll() and netisr_pollmore(), as well as a function, netisr_sched_poll(), to allow the polling code to schedule netisr execution. DEVICE_POLLING still embeds single-netisr assumptions in its implementation, so for now if it is compiled into the kernel, a single and un-bound netisr thread is enforced regardless of tunable configuration. In the default configuration, the new netisr implementation maintains the same basic assumptions as the previous implementation: a single, un-bound worker thread processes all deferred work, and direct dispatch is enabled by default wherever possible. Performance measurement shows a marginal performance improvement over the old implementation due to the use of batched dequeue. An rmlock is used to synchronize use and registration/unregistration using the framework; currently, synchronized use is disabled (replicating current netisr policy) due to a measurable 3%-6% hit in ping-pong micro-benchmarking. It will be enabled once further rmlock optimization has taken place. However, in practice, netisrs are rarely registered or unregistered at runtime. A new man page for netisr will follow, but since one doesn't currently exist, it hasn't been updated. This change is not appropriate for MFC, although the polling shutdown handler should be merged to 7-STABLE. Bump __FreeBSD_version. Reviewed by: bz
2009-06-01 10:41:38 +00:00
0, 0, sysctl_route_netisr_maxqlen, "I",
"maximum routing socket dispatch queue length");
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
static void
Introduce a per-VNET flag to enable/disable netisr prcessing on that VNET. Add accessor functions to toggle the state per VNET. The base system (vnet0) will always enable itself with the normal registration. We will share the registered protocol handlers in all VNETs minimising duplication and management. Upon disabling netisr processing for a VNET drain the netisr queue from packets for that VNET. Update netisr consumers to (de)register on a per-VNET start/teardown using VNET_SYS(UN)INIT functionality. The change should be transparent for non-VIMAGE kernels. Reviewed by: gnn (, hiren) Obtained from: projects/vnet MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D6691
2016-06-03 13:57:10 +00:00
vnet_rts_init(void)
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
{
If a tunable for the routing socket netisr queue max is defined, allow it to override the default value, rather than the default value overriding the tunable.
2004-08-21 21:45:40 +00:00
int tmp;
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
Introduce a per-VNET flag to enable/disable netisr prcessing on that VNET. Add accessor functions to toggle the state per VNET. The base system (vnet0) will always enable itself with the normal registration. We will share the registered protocol handlers in all VNETs minimising duplication and management. Upon disabling netisr processing for a VNET drain the netisr queue from packets for that VNET. Update netisr consumers to (de)register on a per-VNET start/teardown using VNET_SYS(UN)INIT functionality. The change should be transparent for non-VIMAGE kernels. Reviewed by: gnn (, hiren) Obtained from: projects/vnet MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D6691
2016-06-03 13:57:10 +00:00
if (IS_DEFAULT_VNET(curvnet)) {
if (TUNABLE_INT_FETCH("net.route.netisr_maxqlen", &tmp))
rtsock_nh.nh_qlimit = tmp;
netisr_register(&rtsock_nh);
}
#ifdef VIMAGE
else
netisr_register_vnet(&rtsock_nh);
#endif
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
}
Introduce a per-VNET flag to enable/disable netisr prcessing on that VNET. Add accessor functions to toggle the state per VNET. The base system (vnet0) will always enable itself with the normal registration. We will share the registered protocol handlers in all VNETs minimising duplication and management. Upon disabling netisr processing for a VNET drain the netisr queue from packets for that VNET. Update netisr consumers to (de)register on a per-VNET start/teardown using VNET_SYS(UN)INIT functionality. The change should be transparent for non-VIMAGE kernels. Reviewed by: gnn (, hiren) Obtained from: projects/vnet MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D6691
2016-06-03 13:57:10 +00:00
VNET_SYSINIT(vnet_rtsock, SI_SUB_PROTO_DOMAIN, SI_ORDER_THIRD,
vnet_rts_init, 0);
#ifdef VIMAGE
static void
vnet_rts_uninit(void)
{
netisr_unregister_vnet(&rtsock_nh);
}
VNET_SYSUNINIT(vnet_rts_uninit, SI_SUB_PROTO_DOMAIN, SI_ORDER_THIRD,
vnet_rts_uninit, 0);
#endif
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
static int
raw_input_rts_cb(struct mbuf *m, struct sockproto *proto, struct sockaddr *src,
struct rawcb *rp)
{
int fibnum;
KASSERT(m != NULL, ("%s: m is NULL", __func__));
KASSERT(proto != NULL, ("%s: proto is NULL", __func__));
KASSERT(rp != NULL, ("%s: rp is NULL", __func__));
/* No filtering requested. */
if ((m->m_flags & RTS_FILTER_FIB) == 0)
return (0);
/* Check if it is a rts and the fib matches the one of the socket. */
fibnum = M_GETFIB(m);
if (proto->sp_family != PF_ROUTE ||
rp->rcb_socket == NULL ||
rp->rcb_socket->so_fibnum == fibnum)
return (0);
/* Filtering requested and no match, the socket shall be skipped. */
return (1);
}
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
static void
rts_input(struct mbuf *m)
{
struct sockproto route_proto;
unsigned short *family;
struct m_tag *tag;
route_proto.sp_family = PF_ROUTE;
tag = m_tag_find(m, PACKET_TAG_RTSOCKFAM, NULL);
if (tag != NULL) {
family = (unsigned short *)(tag + 1);
route_proto.sp_protocol = *family;
m_tag_delete(m, tag);
} else
route_proto.sp_protocol = 0;
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
raw_input_ext(m, &route_proto, &route_src, raw_input_rts_cb);
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
}
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
/*
* It really doesn't make any sense at all for this code to share much
* with raw_usrreq.c, since its functionality is so restricted. XXX
*/
Change protocol switch pru_abort() API so that it returns void rather than an int, as an error here is not meaningful. Modify soabort() to unconditionally free the socket on the return of pru_abort(), and modify most protocols to no longer conditionally free the socket, since the caller will do this. This commit likely leaves parts of netinet and netinet6 in a situation where they may panic or leak memory, as they have not are not fully updated by this commit. This will be corrected shortly in followup commits to these components. MFC after: 3 months
2006-04-01 15:15:05 +00:00
static void
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
rts_abort(struct socket *so)
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
Change protocol switch pru_abort() API so that it returns void rather than an int, as an error here is not meaningful. Modify soabort() to unconditionally free the socket on the return of pru_abort(), and modify most protocols to no longer conditionally free the socket, since the caller will do this. This commit likely leaves parts of netinet and netinet6 in a situation where they may panic or leak memory, as they have not are not fully updated by this commit. This will be corrected shortly in followup commits to these components. MFC after: 3 months
2006-04-01 15:15:05 +00:00
raw_usrreqs.pru_abort(so);
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
Change semantics of socket close and detach. Add a new protocol switch function, pru_close, to notify protocols that the file descriptor or other consumer of a socket is closing the socket. pru_abort is now a notification of close also, and no longer detaches. pru_detach is no longer used to notify of close, and will be called during socket tear-down by sofree() when all references to a socket evaporate after an earlier call to abort or close the socket. This means detach is now an unconditional teardown of a socket, whereas previously sockets could persist after detach of the protocol retained a reference. This faciliates sharing mutexes between layers of the network stack as the mutex is required during the checking and removal of references at the head of sofree(). With this change, pru_detach can now assume that the mutex will no longer be required by the socket layer after completion, whereas before this was not necessarily true. Reviewed by: gnn
2006-07-21 17:11:15 +00:00
static void
rts_close(struct socket *so)
{
raw_usrreqs.pru_close(so);
}
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
/* pru_accept is EOPNOTSUPP */
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
rts_attach(struct socket *so, int proto, struct thread *td)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
struct rawcb *rp;
Mechanically remove the last stray remains of spl* calls from net*/*. They have been Noop's for a long time now.
2012-10-18 13:57:24 +00:00
int error;
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
KASSERT(so->so_pcb == NULL, ("rts_attach: so_pcb != NULL"));
Convert more malloc+bzero to malloc+M_ZERO. Submitted by: josh@zipperup.org Submitted by: Robert Drehmel <robd@gmx.net>
2000-12-08 21:51:06 +00:00
/* XXX */
Retire the MALLOC and FREE macros. They are an abomination unto style(9). MFC after: 3 months
2008-10-23 15:53:51 +00:00
rp = malloc(sizeof *rp, M_PCB, M_WAITOK | M_ZERO);
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
so->so_pcb = (caddr_t)rp;
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco Reviewed by: several including rwatson, bz and mlair (parts each) Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
so->so_fibnum = td->td_proc->p_fibnum;
Synch with NetBSD and OpenBSD. Allow non-superuser to open, listen to, and send safe commands on the routing socket. Superuser priviledge is required for all commands but RTM_GET. Lose `setuid root' bit of route(8). Reviewed by: wollman, dd
2001-08-31 12:31:09 +00:00
error = raw_attach(so, proto);
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
rp = sotorawcb(so);
if (error) {
fix memory leak when error during opening of routing socket PR: kern/29336 Submitted by: Richard Andrades <richard@xebeo.com> MFC after: 1 month
2001-08-02 19:56:29 +00:00
so->so_pcb = NULL;
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
free(rp, M_PCB);
return error;
}
trivial locking rtsock_cb Sponsored by: FreeBSD Foundation
2003-10-03 18:27:02 +00:00
RTSOCK_LOCK();
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
switch(rp->rcb_proto.sp_protocol) {
case AF_INET:
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.ip_count++;
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
break;
Count AF_INET6 attachement to routing socket. Obtained from: KAME project
2000-01-28 20:10:51 +00:00
case AF_INET6:
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.ip6_count++;
Count AF_INET6 attachement to routing socket. Obtained from: KAME project
2000-01-28 20:10:51 +00:00
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.any_count++;
trivial locking rtsock_cb Sponsored by: FreeBSD Foundation
2003-10-03 18:27:02 +00:00
RTSOCK_UNLOCK();
Remove so*_locked(), which were backed out by mistake.
2002-06-18 07:42:02 +00:00
soisconnected(so);
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
so->so_options |= SO_USELOOPBACK;
return 0;
}
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
rts_bind(struct socket *so, struct sockaddr *nam, struct thread *td)
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_bind(so, nam, td)); /* xxx just EINVAL */
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
rts_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_connect(so, nam, td)); /* XXX just EINVAL */
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
/* pru_connect2 is EOPNOTSUPP */
/* pru_control is EOPNOTSUPP */
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
static void
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
rts_detach(struct socket *so)
{
struct rawcb *rp = sotorawcb(so);
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
KASSERT(rp != NULL, ("rts_detach: rp == NULL"));
RTSOCK_LOCK();
switch(rp->rcb_proto.sp_protocol) {
case AF_INET:
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.ip_count--;
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
break;
case AF_INET6:
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.ip6_count--;
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
V_route_cb.any_count--;
Chance protocol switch method pru_detach() so that it returns void rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months
2006-04-01 15:42:02 +00:00
RTSOCK_UNLOCK();
raw_usrreqs.pru_detach(so);
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
static int
rts_disconnect(struct socket *so)
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_disconnect(so));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
/* pru_listen is EOPNOTSUPP */
static int
Fix all areas of the system (or at least all those in LINT) to avoid storing socket addresses in mbufs. (Socket buffers are the one exception.) A number of kernel APIs needed to get fixed in order to make this happen. Also, fix three protocol families which kept PCBs in mbufs to not malloc them instead. Delete some old compatibility cruft while we're at it, and add some new routines in the in_cksum family.
1997-08-16 19:16:27 +00:00
rts_peeraddr(struct socket *so, struct sockaddr **nam)
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_peeraddr(so, nam));
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
/* pru_rcvd is EOPNOTSUPP */
/* pru_rcvoob is EOPNOTSUPP */
static int
Fix all areas of the system (or at least all those in LINT) to avoid storing socket addresses in mbufs. (Socket buffers are the one exception.) A number of kernel APIs needed to get fixed in order to make this happen. Also, fix three protocol families which kept PCBs in mbufs to not malloc them instead. Delete some old compatibility cruft while we're at it, and add some new routines in the in_cksum family.
1997-08-16 19:16:27 +00:00
rts_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
struct mbuf *control, struct thread *td)
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_send(so, flags, m, nam, control, td));
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
/* pru_sense is null */
static int
rts_shutdown(struct socket *so)
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_shutdown(so));
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
static int
Fix all areas of the system (or at least all those in LINT) to avoid storing socket addresses in mbufs. (Socket buffers are the one exception.) A number of kernel APIs needed to get fixed in order to make this happen. Also, fix three protocol families which kept PCBs in mbufs to not malloc them instead. Delete some old compatibility cruft while we're at it, and add some new routines in the in_cksum family.
1997-08-16 19:16:27 +00:00
rts_sockaddr(struct socket *so, struct sockaddr **nam)
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
{
De-spl parts of the routing socket code now generally protected through locking; leave some spl references around code where there are open questions about global variable references. Also, add an XXX regarding locking in sysctl. MFC after: 3 days
2005-08-25 13:30:04 +00:00
return (raw_usrreqs.pru_sockaddr(so, nam));
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
}
static struct pr_usrreqs route_usrreqs = {
Initialize struct pr_userreqs in new/sparse style and fill in common default elements in net_init_domain(). This makes it possible to grep these structures and see any bogosities.
2004-11-08 14:44:54 +00:00
.pru_abort = rts_abort,
.pru_attach = rts_attach,
.pru_bind = rts_bind,
.pru_connect = rts_connect,
.pru_detach = rts_detach,
.pru_disconnect = rts_disconnect,
.pru_peeraddr = rts_peeraddr,
.pru_send = rts_send,
.pru_shutdown = rts_shutdown,
.pru_sockaddr = rts_sockaddr,
Change semantics of socket close and detach. Add a new protocol switch function, pru_close, to notify protocols that the file descriptor or other consumer of a socket is closing the socket. pru_abort is now a notification of close also, and no longer detaches. pru_detach is no longer used to notify of close, and will be called during socket tear-down by sofree() when all references to a socket evaporate after an earlier call to abort or close the socket. This means detach is now an unconditional teardown of a socket, whereas previously sockets could persist after detach of the protocol retained a reference. This faciliates sharing mutexes between layers of the network stack as the mutex is required during the checking and removal of references at the head of sofree(). With this change, pru_detach can now assume that the mutex will no longer be required by the socket layer after completion, whereas before this was not necessarily true. Reviewed by: gnn
2006-07-21 17:11:15 +00:00
.pru_close = rts_close,
The long-awaited mega-massive-network-code- cleanup. Part I. This commit includes the following changes: 1) Old-style (pr_usrreq()) protocols are no longer supported, the compatibility glue for them is deleted, and the kernel will panic on boot if any are compiled in. 2) Certain protocol entry points are modified to take a process structure, so they they can easily tell whether or not it is possible to sleep, and also to access credentials. 3) SS_PRIV is no more, and with it goes the SO_PRIVSTATE setsockopt() call. Protocols should use the process pointer they are now passed. 4) The PF_LOCAL and PF_ROUTE families have been updated to use the new style, as has the `raw' skeleton family. 5) PF_LOCAL sockets now obey the process's umask when creating a socket in the filesystem. As a result, LINT is now broken. I'm hoping that some enterprising hacker with a bit more time will either make the broken bits work (should be easy for netipx) or dike them out.
1997-04-27 20:01:29 +00:00
};
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
#ifndef _SOCKADDR_UNION_DEFINED
#define _SOCKADDR_UNION_DEFINED
/*
* The union of all possible address formats we handle.
*/
union sockaddr_union {
struct sockaddr sa;
struct sockaddr_in sin;
struct sockaddr_in6 sin6;
};
#endif /* _SOCKADDR_UNION_DEFINED */
static int
rtm_get_jailed(struct rt_addrinfo *info, struct ifnet *ifp,
struct rtentry *rt, union sockaddr_union *saun, struct ucred *cred)
{
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
#if defined(INET) || defined(INET6)
struct epoch_tracker et;
#endif
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
/* First, see if the returned address is part of the jail. */
if (prison_if(cred, rt->rt_ifa->ifa_addr) == 0) {
info->rti_info[RTAX_IFA] = rt->rt_ifa->ifa_addr;
return (0);
}
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
switch (info->rti_info[RTAX_DST]->sa_family) {
#ifdef INET
case AF_INET:
{
struct in_addr ia;
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
struct ifaddr *ifa;
int found;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
found = 0;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
/*
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
* Try to find an address on the given outgoing interface
* that belongs to the jail.
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
*/
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
NET_EPOCH_ENTER(et);
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
struct sockaddr *sa;
sa = ifa->ifa_addr;
if (sa->sa_family != AF_INET)
continue;
ia = ((struct sockaddr_in *)sa)->sin_addr;
if (prison_check_ip4(cred, &ia) == 0) {
found = 1;
break;
}
}
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
NET_EPOCH_EXIT(et);
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
if (!found) {
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
/*
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
* As a last resort return the 'default' jail address.
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
*/
Add hierarchical jails. A jail may further virtualize its environment by creating a child jail, which is visible to that jail and to any parent jails. Child jails may be restricted more than their parents, but never less. Jail names reflect this hierarchy, being MIB-style dot-separated strings. Every thread now points to a jail, the default being prison0, which contains information about the physical system. Prison0's root directory is the same as rootvnode; its hostname is the same as the global hostname, and its securelevel replaces the global securelevel. Note that the variable "securelevel" has actually gone away, which should not cause any problems for code that properly uses securelevel_gt() and securelevel_ge(). Some jail-related permissions that were kept in global variables and set via sysctls are now per-jail settings. The sysctls still exist for backward compatibility, used only by the now-deprecated jail(2) system call. Approved by: bz (mentor)
2009-05-27 14:11:23 +00:00
ia = ((struct sockaddr_in *)rt->rt_ifa->ifa_addr)->
sin_addr;
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
if (prison_get_ip4(cred, &ia) != 0)
return (ESRCH);
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
}
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
bzero(&saun->sin, sizeof(struct sockaddr_in));
saun->sin.sin_len = sizeof(struct sockaddr_in);
saun->sin.sin_family = AF_INET;
saun->sin.sin_addr.s_addr = ia.s_addr;
info->rti_info[RTAX_IFA] = (struct sockaddr *)&saun->sin;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
break;
}
#endif
#ifdef INET6
case AF_INET6:
{
struct in6_addr ia6;
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
struct ifaddr *ifa;
int found;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
found = 0;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
/*
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
* Try to find an address on the given outgoing interface
* that belongs to the jail.
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
*/
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
NET_EPOCH_ENTER(et);
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
struct sockaddr *sa;
sa = ifa->ifa_addr;
if (sa->sa_family != AF_INET6)
continue;
bcopy(&((struct sockaddr_in6 *)sa)->sin6_addr,
&ia6, sizeof(struct in6_addr));
if (prison_check_ip6(cred, &ia6) == 0) {
found = 1;
break;
}
}
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
NET_EPOCH_EXIT(et);
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
if (!found) {
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
/*
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
* As a last resort return the 'default' jail address.
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
*/
Add hierarchical jails. A jail may further virtualize its environment by creating a child jail, which is visible to that jail and to any parent jails. Child jails may be restricted more than their parents, but never less. Jail names reflect this hierarchy, being MIB-style dot-separated strings. Every thread now points to a jail, the default being prison0, which contains information about the physical system. Prison0's root directory is the same as rootvnode; its hostname is the same as the global hostname, and its securelevel replaces the global securelevel. Note that the variable "securelevel" has actually gone away, which should not cause any problems for code that properly uses securelevel_gt() and securelevel_ge(). Some jail-related permissions that were kept in global variables and set via sysctls are now per-jail settings. The sysctls still exist for backward compatibility, used only by the now-deprecated jail(2) system call. Approved by: bz (mentor)
2009-05-27 14:11:23 +00:00
ia6 = ((struct sockaddr_in6 *)rt->rt_ifa->ifa_addr)->
sin6_addr;
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
if (prison_get_ip6(cred, &ia6) != 0)
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
return (ESRCH);
}
Call prison_if from rtm_get_jailed, instead of splitting it out into prison_check_ip4 and prison_check_ip6. As prison_if includes a jailed() check, remove that check before calling rtm_get_jailed. Approved by: bz (mentor)
2009-02-05 14:58:16 +00:00
bzero(&saun->sin6, sizeof(struct sockaddr_in6));
saun->sin6.sin6_len = sizeof(struct sockaddr_in6);
saun->sin6.sin6_family = AF_INET6;
bcopy(&ia6, &saun->sin6.sin6_addr, sizeof(struct in6_addr));
if (sa6_recoverscope(&saun->sin6) != 0)
return (ESRCH);
info->rti_info[RTAX_IFA] = (struct sockaddr *)&saun->sin6;
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
break;
}
#endif
default:
return (ESRCH);
}
return (0);
}
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
/*
* Fills in @info based on userland-provided @rtm message.
*
* Returns 0 on success.
*/
static int
fill_addrinfo(struct rt_msghdr *rtm, int len, u_int fibnum, struct rt_addrinfo *info)
{
int error;
sa_family_t saf;
rtm->rtm_pid = curproc->p_pid;
info->rti_addrs = rtm->rtm_addrs;
info->rti_mflags = rtm->rtm_inits;
info->rti_rmx = &rtm->rtm_rmx;
/*
* rt_xaddrs() performs s6_addr[2] := sin6_scope_id for AF_INET6
* link-local address because rtrequest requires addresses with
* embedded scope id.
*/
if (rt_xaddrs((caddr_t)(rtm + 1), len + (caddr_t)rtm, info))
return (EINVAL);
if (rtm->rtm_flags & RTF_RNH_LOCKED)
return (EINVAL);
info->rti_flags = rtm->rtm_flags;
if (info->rti_info[RTAX_DST] == NULL ||
info->rti_info[RTAX_DST]->sa_family >= AF_MAX ||
(info->rti_info[RTAX_GATEWAY] != NULL &&
info->rti_info[RTAX_GATEWAY]->sa_family >= AF_MAX))
return (EINVAL);
saf = info->rti_info[RTAX_DST]->sa_family;
/*
* Verify that the caller has the appropriate privilege; RTM_GET
* is the only operation the non-superuser is allowed.
*/
if (rtm->rtm_type != RTM_GET) {
error = priv_check(curthread, PRIV_NET_ROUTE);
if (error != 0)
return (error);
}
/*
* The given gateway address may be an interface address.
* For example, issuing a "route change" command on a route
* entry that was created from a tunnel, and the gateway
* address given is the local end point. In this case the
* RTF_GATEWAY flag must be cleared or the destination will
* not be reachable even though there is no error message.
*/
if (info->rti_info[RTAX_GATEWAY] != NULL &&
info->rti_info[RTAX_GATEWAY]->sa_family != AF_LINK) {
struct rt_addrinfo ginfo;
struct sockaddr *gdst;
struct sockaddr_storage ss;
bzero(&ginfo, sizeof(ginfo));
bzero(&ss, sizeof(ss));
ss.ss_len = sizeof(ss);
ginfo.rti_info[RTAX_GATEWAY] = (struct sockaddr *)&ss;
gdst = info->rti_info[RTAX_GATEWAY];
/*
* A host route through the loopback interface is
* installed for each interface adddress. In pre 8.0
* releases the interface address of a PPP link type
* is not reachable locally. This behavior is fixed as
* part of the new L2/L3 redesign and rewrite work. The
* signature of this interface address route is the
* AF_LINK sa_family type of the rt_gateway, and the
* rt_ifp has the IFF_LOOPBACK flag set.
*/
if (rib_lookup_info(fibnum, gdst, NHR_REF, 0, &ginfo) == 0) {
if (ss.ss_family == AF_LINK &&
ginfo.rti_ifp->if_flags & IFF_LOOPBACK) {
info->rti_flags &= ~RTF_GATEWAY;
info->rti_flags |= RTF_GWFLAG_COMPAT;
}
rib_free_info(&ginfo);
}
}
return (0);
}
/*
* Handles RTM_GET message from routing socket, returning matching rt.
*
* Returns:
* 0 on success, with locked and referenced matching rt in @rt_nrt
* errno of failure
*/
static int
handle_rtm_get(struct rt_addrinfo *info, u_int fibnum,
struct rt_msghdr *rtm, struct rtentry **ret_nrt)
{
RIB_RLOCK_TRACKER;
struct rtentry *rt;
struct rib_head *rnh;
sa_family_t saf;
saf = info->rti_info[RTAX_DST]->sa_family;
rnh = rt_tables_get_rnh(fibnum, saf);
if (rnh == NULL)
return (EAFNOSUPPORT);
RIB_RLOCK(rnh);
if (info->rti_info[RTAX_NETMASK] == NULL) {
/*
* Provide longest prefix match for
* address lookup (no mask).
* 'route -n get addr'
*/
rt = (struct rtentry *) rnh->rnh_matchaddr(
info->rti_info[RTAX_DST], &rnh->head);
} else
rt = (struct rtentry *) rnh->rnh_lookup(
info->rti_info[RTAX_DST],
info->rti_info[RTAX_NETMASK], &rnh->head);
if (rt == NULL) {
RIB_RUNLOCK(rnh);
return (ESRCH);
}
#ifdef RADIX_MPATH
/*
* for RTM_GET, gate is optional even with multipath.
* if gate == NULL the first match is returned.
* (no need to call rt_mpath_matchgate if gate == NULL)
*/
if (rt_mpath_capable(rnh) && info->rti_info[RTAX_GATEWAY]) {
rt = rt_mpath_matchgate(rt, info->rti_info[RTAX_GATEWAY]);
if (!rt) {
RIB_RUNLOCK(rnh);
return (ESRCH);
}
}
#endif
/*
* If performing proxied L2 entry insertion, and
* the actual PPP host entry is found, perform
* another search to retrieve the prefix route of
* the local end point of the PPP link.
*/
if (rtm->rtm_flags & RTF_ANNOUNCE) {
struct sockaddr laddr;
if (rt->rt_ifp != NULL &&
rt->rt_ifp->if_type == IFT_PROPVIRTUAL) {
struct epoch_tracker et;
struct ifaddr *ifa;
NET_EPOCH_ENTER(et);
ifa = ifa_ifwithnet(info->rti_info[RTAX_DST], 1,
RT_ALL_FIBS);
NET_EPOCH_EXIT(et);
if (ifa != NULL)
rt_maskedcopy(ifa->ifa_addr,
&laddr,
ifa->ifa_netmask);
} else
rt_maskedcopy(rt->rt_ifa->ifa_addr,
&laddr,
rt->rt_ifa->ifa_netmask);
/*
* refactor rt and no lock operation necessary
*/
rt = (struct rtentry *)rnh->rnh_matchaddr(&laddr,
&rnh->head);
if (rt == NULL) {
RIB_RUNLOCK(rnh);
return (ESRCH);
}
}
RT_LOCK(rt);
RT_ADDREF(rt);
RIB_RUNLOCK(rnh);
*ret_nrt = rt;
return (0);
}
/*
* Update sockaddrs, flags, etc in @prtm based on @rt data.
* Assumes @rt is locked.
* rtm can be reallocated.
*
* Returns 0 on success, along with pointer to (potentially reallocated)
* rtm.
*
*/
static int
update_rtm_from_rte(struct rt_addrinfo *info, struct rt_msghdr **prtm,
int alloc_len, struct rtentry *rt)
{
struct sockaddr_storage netmask_ss;
struct walkarg w;
union sockaddr_union saun;
struct rt_msghdr *rtm, *orig_rtm = NULL;
struct ifnet *ifp;
int error, len;
RT_LOCK_ASSERT(rt);
rtm = *prtm;
info->rti_info[RTAX_DST] = rt_key(rt);
info->rti_info[RTAX_GATEWAY] = rt->rt_gateway;
info->rti_info[RTAX_NETMASK] = rtsock_fix_netmask(rt_key(rt),
rt_mask(rt), &netmask_ss);
info->rti_info[RTAX_GENMASK] = 0;
ifp = rt->rt_ifp;
if (rtm->rtm_addrs & (RTA_IFP | RTA_IFA)) {
if (ifp) {
info->rti_info[RTAX_IFP] =
ifp->if_addr->ifa_addr;
error = rtm_get_jailed(info, ifp, rt,
&saun, curthread->td_ucred);
if (error != 0)
return (error);
if (ifp->if_flags & IFF_POINTOPOINT)
info->rti_info[RTAX_BRD] =
rt->rt_ifa->ifa_dstaddr;
rtm->rtm_index = ifp->if_index;
} else {
info->rti_info[RTAX_IFP] = NULL;
info->rti_info[RTAX_IFA] = NULL;
}
} else if (ifp != NULL)
rtm->rtm_index = ifp->if_index;
/* Check if we need to realloc storage */
rtsock_msg_buffer(rtm->rtm_type, info, NULL, &len);
if (len > alloc_len) {
struct rt_msghdr *tmp_rtm;
tmp_rtm = malloc(len, M_TEMP, M_NOWAIT);
if (tmp_rtm == NULL)
return (ENOBUFS);
bcopy(rtm, tmp_rtm, rtm->rtm_msglen);
orig_rtm = rtm;
rtm = tmp_rtm;
alloc_len = len;
/*
* Delay freeing original rtm as info contains
* data referencing it.
*/
}
w.w_tmem = (caddr_t)rtm;
w.w_tmemsize = alloc_len;
rtsock_msg_buffer(rtm->rtm_type, info, &w, &len);
if (rt->rt_flags & RTF_GWFLAG_COMPAT)
rtm->rtm_flags = RTF_GATEWAY |
(rt->rt_flags & ~RTF_GWFLAG_COMPAT);
else
rtm->rtm_flags = rt->rt_flags;
rt_getmetrics(rt, &rtm->rtm_rmx);
rtm->rtm_addrs = info->rti_addrs;
if (orig_rtm != NULL)
free(orig_rtm, M_TEMP);
*prtm = rtm;
return (0);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*ARGSUSED*/
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
static int
Change pr_output's prototype to avoid the need for explicit casts. This is a follow up to r269699. Phabric: D564 Reviewed by: jhb
2014-08-15 02:43:02 +00:00
route_output(struct mbuf *m, struct socket *so, ...)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
struct rt_msghdr *rtm = NULL;
struct rtentry *rt = NULL;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct rt_addrinfo info;
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
struct epoch_tracker et;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
#ifdef INET6
Fix NOINET6 build broken by r356236. MFC after: 2 weeks
2019-12-31 17:57:12 +00:00
struct sockaddr_storage ss;
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
struct sockaddr_in6 *sin6;
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
int i, rti_need_deembed = 0;
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#endif
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
int alloc_len = 0, len, error = 0, fibnum;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
sa_family_t saf = AF_UNSPEC;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
struct walkarg w;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Move up fibnum to ensure it is always defined. Found by: ian MFC with: r264987
2014-04-27 02:20:09 +00:00
fibnum = so->so_fibnum;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define senderr(e) { error = e; goto flush;}
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
if (m == NULL || ((m->m_len < sizeof(long)) &&
(m = m_pullup(m, sizeof(long))) == NULL))
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (ENOBUFS);
if ((m->m_flags & M_PKTHDR) == 0)
panic("route_output");
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
NET_EPOCH_ENTER(et);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
len = m->m_pkthdr.len;
if (len < sizeof(*rtm) ||
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
len != mtod(m, struct rt_msghdr *)->rtm_msglen)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
senderr(EINVAL);
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
/*
* Most of current messages are in range 200-240 bytes,
Partially revert r265019 - allocating 512 bytes on stack can be too much for architectures like ARM. Always use rounded malloc instead. Discussed with: jmallett MFC after: 4 weeks
2014-04-29 19:48:11 +00:00
* minimize possible re-allocation on reply using larger size
* buffer aligned on 1k boundaty.
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
*/
Partially revert r265019 - allocating 512 bytes on stack can be too much for architectures like ARM. Always use rounded malloc instead. Discussed with: jmallett MFC after: 4 weeks
2014-04-29 19:48:11 +00:00
alloc_len = roundup2(len, 1024);
if ((rtm = malloc(alloc_len, M_TEMP, M_NOWAIT)) == NULL)
senderr(ENOBUFS);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
m_copydata(m, 0, len, (caddr_t)rtm);
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
bzero(&info, sizeof(info));
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
bzero(&w, sizeof(w));
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (rtm->rtm_version != RTM_VERSION) {
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
/* Do not touch message since format is unknown */
Partially revert r265019 - allocating 512 bytes on stack can be too much for architectures like ARM. Always use rounded malloc instead. Discussed with: jmallett MFC after: 4 weeks
2014-04-29 19:48:11 +00:00
free(rtm, M_TEMP);
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
rtm = NULL;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
senderr(EPROTONOSUPPORT);
}
Remove useless zeroing of RTAX_DST on error. Cleanup a bit. MFC after: 1 month
2014-04-27 10:43:48 +00:00
/*
* Starting from here, it is possible
* to alter original message and insert
* caller PID and error value.
*/
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
if ((error = fill_addrinfo(rtm, len, fibnum, &info)) != 0) {
senderr(error);
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
}
Synch with NetBSD and OpenBSD. Allow non-superuser to open, listen to, and send safe commands on the routing socket. Superuser priviledge is required for all commands but RTM_GET. Lose `setuid root' bit of route(8). Reviewed by: wollman, dd
2001-08-31 12:31:09 +00:00
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
saf = info.rti_info[RTAX_DST]->sa_family;
In ip_output(), the flow-table module must not try to cache L2/L3 information for interface of IFF_POINTOPOINT or IFF_LOOPBACK type. Since the L2 information (rt_lle) is invalid for these interface types, accidental caching attempt will trigger panic when the invalid rt_lle reference is accessed. When installing a new route, or when updating an existing route, the user supplied gateway address may be an interface address (this is particularly true for point-to-point interface related modules such as ppp, if_tun, if_gif). Currently the routing command handler always set the RTF_GATEWAY flag if the gateway address is given as part of the command paramters. Therefore the gateway address must be verified against interface addresses or else the route would be treated as an indirect route, thus making that route unusable. Reviewed by: kmacy, julia, rwatson Verified by: marcus MFC after: 3 days
2009-08-28 07:01:09 +00:00
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
/* support for new ARP code */
if (rtm->rtm_flags & RTF_LLDATA) {
error = lla_rt_output(rtm, &info);
#ifdef INET6
if (error == 0)
rti_need_deembed = (V_deembed_scopeid) ? 1 : 0;
#endif
goto flush;
In ip_output(), the flow-table module must not try to cache L2/L3 information for interface of IFF_POINTOPOINT or IFF_LOOPBACK type. Since the L2 information (rt_lle) is invalid for these interface types, accidental caching attempt will trigger panic when the invalid rt_lle reference is accessed. When installing a new route, or when updating an existing route, the user supplied gateway address may be an interface address (this is particularly true for point-to-point interface related modules such as ppp, if_tun, if_gif). Currently the routing command handler always set the RTF_GATEWAY flag if the gateway address is given as part of the command paramters. Therefore the gateway address must be verified against interface addresses or else the route would be treated as an indirect route, thus making that route unusable. Reviewed by: kmacy, julia, rwatson Verified by: marcus MFC after: 3 days
2009-08-28 07:01:09 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
switch (rtm->rtm_type) {
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct rtentry *saved_nrt;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
case RTM_ADD:
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
case RTM_CHANGE:
Allow route change requests to not specify the gateway. Only require a gateway to be specified on a route add request. On a route change request that does not specify the gateway, the gateway will remain the same. This allows changing other route parameters without having to re-specifying the gateway, like in "route change 10.0.0.0/8 -mtu 9000". Update the route(8) manpage to explicitly call out this usage as being supported. MFC after: 2 weeks Sponsored by: Dell EMC Isilon Reviewed By: eugen (rtsock.c change), rgrimes Differential Revision: https://reviews.freebsd.org/D14291
2018-02-21 19:13:23 +00:00
if (rtm->rtm_type == RTM_ADD) {
if (info.rti_info[RTAX_GATEWAY] == NULL)
senderr(EINVAL);
}
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
saved_nrt = NULL;
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
error = rtrequest1_fib(rtm->rtm_type, &info, &saved_nrt,
Determine fibnum once in the beginning of route_output(). MFC after: 1 month
2014-04-26 22:32:04 +00:00
fibnum);
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
if (error == 0 && saved_nrt != NULL) {
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#ifdef INET6
rti_need_deembed = (V_deembed_scopeid) ? 1 : 0;
#endif
Locking for updates to routing table entries. Each rtentry gets a mutex that covers updates to the contents. Note this is separate from holding a reference and/or locking the routing table itself. Other/related changes: o rtredirect loses the final parameter by which an rtentry reference may be returned; this was never used and added unwarranted complexity for locking. o minor style cleanups to routing code (e.g. ansi-fy function decls) o remove the logic to bump the refcnt on the parent of cloned routes, we assume the parent will remain as long as the clone; doing this avoids a circularity in locking during delete o convert some timeouts to MPSAFE callouts Notes: 1. rt_mtx in struct rtentry is guarded by #ifdef _KERNEL as user-level applications cannot/do-no know about mutex's. Doing this requires that the mutex be the last element in the structure. A better solution is to introduce an externalized version of struct rtentry but this is a major task because of the intertwining of rtentry and other data structures that are visible to user applications. 2. There are known LOR's that are expected to go away with forthcoming work to eliminate many held references. If not these will be resolved prior to release. 3. ATM changes are untested. Sponsored by: FreeBSD Foundation Obtained from: BSD/OS (partly)
2003-10-04 03:44:50 +00:00
RT_LOCK(saved_nrt);
- Fill in the correct rtm_index for RTM_ADD and RTM_CHANGE messages. - Allow RTM_CHANGE to change a number of route flags as specified by RTF_FMASK. - The unused rtm_use field in struct rt_msghdr is redesignated as rtm_fmask field to communicate route flag changes in RTM_CHANGE messages from userland. The use count of a route was moved to rtm_rmx a long time ago. For source code compatibility reasons a define of rtm_use to rtm_fmask is provided. These changes faciliate running of multiple cooperating routing daemons at the same time without causing undesired interference. Open[BGP|OSPF]D make use of these features to have IGP routes override EGP ones. Obtained from: OpenBSD (claudio@) MFC after: 3 days
2006-03-15 19:39:09 +00:00
rtm->rtm_index = saved_nrt->rt_ifp->if_index;
replace explicit changes to rt_refcnt by RT_ADDREF and RT_REMREF macros that expand to include assertions when the system is built with INVARIANTS Supported by: FreeBSD Foundation
2003-11-08 23:36:32 +00:00
RT_REMREF(saved_nrt);
Locking for updates to routing table entries. Each rtentry gets a mutex that covers updates to the contents. Note this is separate from holding a reference and/or locking the routing table itself. Other/related changes: o rtredirect loses the final parameter by which an rtentry reference may be returned; this was never used and added unwarranted complexity for locking. o minor style cleanups to routing code (e.g. ansi-fy function decls) o remove the logic to bump the refcnt on the parent of cloned routes, we assume the parent will remain as long as the clone; doing this avoids a circularity in locking during delete o convert some timeouts to MPSAFE callouts Notes: 1. rt_mtx in struct rtentry is guarded by #ifdef _KERNEL as user-level applications cannot/do-no know about mutex's. Doing this requires that the mutex be the last element in the structure. A better solution is to introduce an externalized version of struct rtentry but this is a major task because of the intertwining of rtentry and other data structures that are visible to user applications. 2. There are known LOR's that are expected to go away with forthcoming work to eliminate many held references. If not these will be resolved prior to release. 3. ATM changes are untested. Sponsored by: FreeBSD Foundation Obtained from: BSD/OS (partly)
2003-10-04 03:44:50 +00:00
RT_UNLOCK(saved_nrt);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
break;
case RTM_DELETE:
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
saved_nrt = NULL;
Determine fibnum once in the beginning of route_output(). MFC after: 1 month
2014-04-26 22:32:04 +00:00
error = rtrequest1_fib(RTM_DELETE, &info, &saved_nrt, fibnum);
Updated routing-socket code from Berkeley Obtained from: Keith Bostic by way of Paul Traina
1995-05-10 16:48:52 +00:00
if (error == 0) {
Locking for updates to routing table entries. Each rtentry gets a mutex that covers updates to the contents. Note this is separate from holding a reference and/or locking the routing table itself. Other/related changes: o rtredirect loses the final parameter by which an rtentry reference may be returned; this was never used and added unwarranted complexity for locking. o minor style cleanups to routing code (e.g. ansi-fy function decls) o remove the logic to bump the refcnt on the parent of cloned routes, we assume the parent will remain as long as the clone; doing this avoids a circularity in locking during delete o convert some timeouts to MPSAFE callouts Notes: 1. rt_mtx in struct rtentry is guarded by #ifdef _KERNEL as user-level applications cannot/do-no know about mutex's. Doing this requires that the mutex be the last element in the structure. A better solution is to introduce an externalized version of struct rtentry but this is a major task because of the intertwining of rtentry and other data structures that are visible to user applications. 2. There are known LOR's that are expected to go away with forthcoming work to eliminate many held references. If not these will be resolved prior to release. 3. ATM changes are untested. Sponsored by: FreeBSD Foundation Obtained from: BSD/OS (partly)
2003-10-04 03:44:50 +00:00
RT_LOCK(saved_nrt);
If the caller of rtrequest*(RTM_DELETE, ...) asked for a copy of the entry being removed (ret_nrt != NULL), increment the entry's rt_refcnt like we do it for RTM_ADD and RTM_RESOLVE, rather than messing around with 1->0 transitions for rtfree() all over.
2002-12-25 10:21:02 +00:00
rt = saved_nrt;
Updated routing-socket code from Berkeley Obtained from: Keith Bostic by way of Paul Traina
1995-05-10 16:48:52 +00:00
goto report;
}
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#ifdef INET6
/* rt_msg2() will not be used when RTM_DELETE fails. */
rti_need_deembed = (V_deembed_scopeid) ? 1 : 0;
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
break;
case RTM_GET:
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
error = handle_rtm_get(&info, fibnum, rtm, &rt);
if (error != 0)
senderr(error);
SMP locking for radix nodes.
2002-12-24 03:03:39 +00:00
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
report:
RT_LOCK_ASSERT(rt);
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
if (!can_export_rte(curthread->td_ucred, rt)) {
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
RT_UNLOCK(rt);
senderr(ESRCH);
}
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
error = update_rtm_from_rte(&info, &rtm, alloc_len, rt);
/*
* Note that some sockaddr pointers may have changed to
* point to memory outsize @rtm. Some may be pointing
* to the on-stack variables.
* Given that, any pointer in @info CANNOT BE USED.
*/
Decouple RTM_CHANGE from RTM_GET handling in rtsock.c:route_output(). RTM_CHANGE is now handled inside route.c:rtrequest1_fib() as it should be. Note change change handler is a separate function rtrequest1_fib_change(). MFC after: 1 month
2014-04-26 21:03:41 +00:00
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
/*
* scopeid deembedding has been performed while
* writing updated rtm in rtsock_msg_buffer().
* With that in mind, skip deembedding procedure below.
*/
Fix NOINET6 build broken by r356236. MFC after: 2 weeks
2019-12-31 17:57:12 +00:00
#ifdef INET6
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
rti_need_deembed = 0;
Fix NOINET6 build broken by r356236. MFC after: 2 weeks
2019-12-31 17:57:12 +00:00
#endif
Locking for updates to routing table entries. Each rtentry gets a mutex that covers updates to the contents. Note this is separate from holding a reference and/or locking the routing table itself. Other/related changes: o rtredirect loses the final parameter by which an rtentry reference may be returned; this was never used and added unwarranted complexity for locking. o minor style cleanups to routing code (e.g. ansi-fy function decls) o remove the logic to bump the refcnt on the parent of cloned routes, we assume the parent will remain as long as the clone; doing this avoids a circularity in locking during delete o convert some timeouts to MPSAFE callouts Notes: 1. rt_mtx in struct rtentry is guarded by #ifdef _KERNEL as user-level applications cannot/do-no know about mutex's. Doing this requires that the mutex be the last element in the structure. A better solution is to introduce an externalized version of struct rtentry but this is a major task because of the intertwining of rtentry and other data structures that are visible to user applications. 2. There are known LOR's that are expected to go away with forthcoming work to eliminate many held references. If not these will be resolved prior to release. 3. ATM changes are untested. Sponsored by: FreeBSD Foundation Obtained from: BSD/OS (partly)
2003-10-04 03:44:50 +00:00
RT_UNLOCK(rt);
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
if (error != 0)
senderr(error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
break;
default:
senderr(EOPNOTSUPP);
}
flush:
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
NET_EPOCH_EXIT(et);
Cleanup route_output() a bit. MFC after: 1 month
2014-04-27 10:20:37 +00:00
if (rt != NULL)
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
RTFREE(rt);
Cleanup route_output() a bit. MFC after: 1 month
2014-04-27 10:20:37 +00:00
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#ifdef INET6
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
if (rtm != NULL) {
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
if (rti_need_deembed) {
/* sin6_scope_id is recovered before sending rtm. */
sin6 should be assigned before the loop.
2013-07-28 20:02:41 +00:00
sin6 = (struct sockaddr_in6 *)&ss;
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
for (i = 0; i < RTAX_MAX; i++) {
if (info.rti_info[i] == NULL)
continue;
if (info.rti_info[i]->sa_family != AF_INET6)
continue;
bcopy(info.rti_info[i], sin6, sizeof(*sin6));
if (sa6_recoverscope(sin6) == 0)
bcopy(sin6, info.rti_info[i],
sizeof(*sin6));
}
}
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
}
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#endif
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
send_rtm_reply(so, rtm, m, saf, fibnum, error);
return (error);
}
/*
* Sends the prepared reply message in @rtm to all rtsock clients.
* Frees @m and @rtm.
*
*/
static void
send_rtm_reply(struct socket *so, struct rt_msghdr *rtm, struct mbuf *m,
sa_family_t saf, u_int fibnum, int rtm_errno)
{
struct rawcb *rp = NULL;
/*
* Check to see if we don't want our own messages.
*/
if ((so->so_options & SO_USELOOPBACK) == 0) {
if (V_route_cb.any_count <= 1) {
if (rtm != NULL)
free(rtm, M_TEMP);
m_freem(m);
return;
}
/* There is another listener, so construct message */
rp = sotorawcb(so);
}
if (rtm != NULL) {
if (rtm_errno!= 0)
rtm->rtm_errno = rtm_errno;
Cleanup route_output() a bit. MFC after: 1 month
2014-04-27 10:20:37 +00:00
else
rtm->rtm_flags |= RTF_DONE;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
m_copyback(m, 0, rtm->rtm_msglen, (caddr_t)rtm);
adjust mbuf length right in route_output(). Obtained from: KAME MFC after: 1 week
2001-07-04 11:29:55 +00:00
if (m->m_pkthdr.len < rtm->rtm_msglen) {
m_freem(m);
m = NULL;
} else if (m->m_pkthdr.len > rtm->rtm_msglen)
m_adj(m, rtm->rtm_msglen - m->m_pkthdr.len);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
Partially revert r265019 - allocating 512 bytes on stack can be too much for architectures like ARM. Always use rounded malloc instead. Discussed with: jmallett MFC after: 4 weeks
2014-04-29 19:48:11 +00:00
free(rtm, M_TEMP);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Cleanup route_output() a bit. MFC after: 1 month
2014-04-27 10:20:37 +00:00
if (m != NULL) {
Determine fibnum once in the beginning of route_output(). MFC after: 1 month
2014-04-26 22:32:04 +00:00
M_SETFIB(m, fibnum);
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
m->m_flags |= RTS_FILTER_FIB;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
if (rp) {
/*
* XXX insure we don't get a copy by
* invalidating our protocol
*/
unsigned short family = rp->rcb_proto.sp_family;
rp->rcb_proto.sp_family = 0;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, saf);
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
rp->rcb_proto.sp_family = family;
} else
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, saf);
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
Introduce tcp_hostcache and remove the tcp specific metrics from the routing table. Move all usage and references in the tcp stack from the routing table metrics to the tcp hostcache. It caches measured parameters of past tcp sessions to provide better initial start values for following connections from or to the same source or destination. Depending on the network parameters to/from the remote host this can lead to significant speedups for new tcp connections after the first one because they inherit and shortcut the learning curve. tcp_hostcache is designed for multiple concurrent access in SMP environments with high contention and is hash indexed by remote ip address. It removes significant locking requirements from the tcp stack with regard to the routing table. Reviewed by: sam (mentor), bms Reviewed by: -net, -current, core@kame.net (IPv6 parts) Approved by: re (scottl)
2003-11-20 20:07:39 +00:00
static void
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
rt_getmetrics(const struct rtentry *rt, struct rt_metrics *out)
Introduce tcp_hostcache and remove the tcp specific metrics from the routing table. Move all usage and references in the tcp stack from the routing table metrics to the tcp hostcache. It caches measured parameters of past tcp sessions to provide better initial start values for following connections from or to the same source or destination. Depending on the network parameters to/from the remote host this can lead to significant speedups for new tcp connections after the first one because they inherit and shortcut the learning curve. tcp_hostcache is designed for multiple concurrent access in SMP environments with high contention and is hash indexed by remote ip address. It removes significant locking requirements from the tcp stack with regard to the routing table. Reviewed by: sam (mentor), bms Reviewed by: -net, -current, core@kame.net (IPv6 parts) Approved by: re (scottl)
2003-11-20 20:07:39 +00:00
{
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
Introduce tcp_hostcache and remove the tcp specific metrics from the routing table. Move all usage and references in the tcp stack from the routing table metrics to the tcp hostcache. It caches measured parameters of past tcp sessions to provide better initial start values for following connections from or to the same source or destination. Depending on the network parameters to/from the remote host this can lead to significant speedups for new tcp connections after the first one because they inherit and shortcut the learning curve. tcp_hostcache is designed for multiple concurrent access in SMP environments with high contention and is hash indexed by remote ip address. It removes significant locking requirements from the tcp stack with regard to the routing table. Reviewed by: sam (mentor), bms Reviewed by: -net, -current, core@kame.net (IPv6 parts) Approved by: re (scottl)
2003-11-20 20:07:39 +00:00
bzero(out, sizeof(*out));
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
out->rmx_mtu = rt->rt_mtu;
out->rmx_weight = rt->rt_weight;
out->rmx_pksent = counter_u64_fetch(rt->rt_pksent);
Introduce nexthop objects and new routing KPI. This is the foundational change for the routing subsytem rearchitecture. More details and goals are available in https://reviews.freebsd.org/D24141 . This patch introduces concept of nexthop objects and new nexthop-based routing KPI. Nexthops are objects, containing all necessary information for performing the packet output decision. Output interface, mtu, flags, gw address goes there. For most of the cases, these objects will serve the same role as the struct rtentry is currently serving. Typically there will be low tens of such objects for the router even with multiple BGP full-views, as these objects will be shared between routing entries. This allows to store more information in the nexthop. New KPI: struct nhop_object *fib4_lookup(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, uint32_t flowid); struct nhop_object *fib6_lookup(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, uint32_t flowid); These 2 function are intended to replace all all flavours of <in_|in6_>rtalloc[1]<_ign><_fib>, mpath functions and the previous fib[46]-generation functions. Upon successful lookup, they return nexthop object which is guaranteed to exist within current NET_EPOCH. If longer lifetime is desired, one can specify NHR_REF as a flag and get a referenced version of the nexthop. Reference semantic closely resembles rtentry one, allowing sed-style conversion. Additionally, another 2 functions are introduced to support uRPF functionality inside variety of our firewalls. Their primary goal is to hide the multipath implementation details inside the routing subsystem, greatly simplifying firewalls implementation: int fib4_lookup_urpf(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); int fib6_lookup_urpf(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); All functions have a separate scopeid argument, paving way to eliminating IPv6 scope embedding and allowing to support IPv4 link-locals in the future. Structure changes: * rtentry gets new 'rt_nhop' pointer, slightly growing the overall size. * rib_head gets new 'rnh_preadd' callback pointer, slightly growing overall sz. Old KPI: During the transition state old and new KPI will coexists. As there are another 4-5 decent-sized conversion patches, it will probably take a couple of weeks. To support both KPIs, fields not required by the new KPI (most of rtentry) has to be kept, resulting in the temporary size increase. Once conversion is finished, rtentry will notably shrink. More details: * architectural overview: https://reviews.freebsd.org/D24141 * list of the next changes: https://reviews.freebsd.org/D24232 Reviewed by: ae,glebius(initial version) Differential Revision: https://reviews.freebsd.org/D24232
2020-04-12 14:30:00 +00:00
out->rmx_nhidx = nhop_get_idx(rt->rt_nhop);
Adjust rt_(set|get)metrics() to do kernel <-> userland timebase conversion. We need it since kernel timebase has changed (time_second -> time_uptime). Approved by: glebius (mentor)
2006-07-06 00:24:36 +00:00
/* Kernel -> userland timebase conversion. */
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
out->rmx_expire = rt->rt_expire ?
rt->rt_expire - time_uptime + time_second : 0;
Introduce tcp_hostcache and remove the tcp specific metrics from the routing table. Move all usage and references in the tcp stack from the routing table metrics to the tcp hostcache. It caches measured parameters of past tcp sessions to provide better initial start values for following connections from or to the same source or destination. Depending on the network parameters to/from the remote host this can lead to significant speedups for new tcp connections after the first one because they inherit and shortcut the learning curve. tcp_hostcache is designed for multiple concurrent access in SMP environments with high contention and is hash indexed by remote ip address. It removes significant locking requirements from the tcp stack with regard to the routing table. Reviewed by: sam (mentor), bms Reviewed by: -net, -current, core@kame.net (IPv6 parts) Approved by: re (scottl)
2003-11-20 20:07:39 +00:00
}
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
/*
* Extract the addresses of the passed sockaddrs.
* Do a little sanity checking so as to avoid bad memory references.
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
* This data is derived straight from userland.
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
*/
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
static int
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
rt_xaddrs(caddr_t cp, caddr_t cplim, struct rt_addrinfo *rtinfo)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
route.h: introduce a macro, SA_SIZE(struct sockaddr *) which returns the space occupied by a struct sockaddr when passed through a routing socket. Use it to replace the macro ROUNDUP(int), that does the same but is redefined by every file which uses it, courtesy of the School of Cut'n'Paste Programming(TM). (partial) userland changes to follow.
2004-04-13 11:22:22 +00:00
struct sockaddr *sa;
int i;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
for (i = 0; i < RTAX_MAX && cp < cplim; i++) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if ((rtinfo->rti_addrs & (1 << i)) == 0)
continue;
Bungled cut/paste leaves kernel with page faults.. (read all about it!)
1997-07-16 14:55:14 +00:00
sa = (struct sockaddr *)cp;
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
/*
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
* It won't fit.
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
*/
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
if (cp + sa->sa_len > cplim)
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
return (EINVAL);
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
/*
* there are no more.. quit now
* If there are more bits, they are in error.
* I've seen this. route(1) can evidently generate these.
* This causes kernel to core dump.
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
* for compatibility, If we see this, point to a safe address.
Finally track down the reason for some of my occasional kernel crashes. Route(1) has a bug that sends a bad message to the kernel. The kernel trusts it and crashes. Add some sanity checks so that we don't trust the user quite as much any more. (also add a comment in if_ethersubr.c)
1997-07-15 23:25:32 +00:00
*/
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
if (sa->sa_len == 0) {
rtinfo->rti_info[i] = &sa_zero;
return (0); /* should be EINVAL but for compat */
}
/* accept it */
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
#ifdef INET6
if (sa->sa_family == AF_INET6)
sa6_embedscope((struct sockaddr_in6 *)sa,
V_ip6_use_defzone);
#endif
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
rtinfo->rti_info[i] = sa;
route.h: introduce a macro, SA_SIZE(struct sockaddr *) which returns the space occupied by a struct sockaddr when passed through a routing socket. Use it to replace the macro ROUNDUP(int), that does the same but is redefined by every file which uses it, courtesy of the School of Cut'n'Paste Programming(TM). (partial) userland changes to follow.
2004-04-13 11:22:22 +00:00
cp += SA_SIZE(sa);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
An actual fix for the routing default crashes that 1/ is compatible with the old route(1) in case needed. 2/ actually fixes the problem while vetting bad user input. note: I have already fixed route(1) so the problem shouldn't occur. if it does. use 0.0.0.0/0 instead of the word 'default' :)
1997-07-18 11:44:24 +00:00
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
/*
* Fill in @dmask with valid netmask leaving original @smask
* intact. Mostly used with radix netmasks.
*/
static struct sockaddr *
rtsock_fix_netmask(struct sockaddr *dst, struct sockaddr *smask,
struct sockaddr_storage *dmask)
{
if (dst == NULL || smask == NULL)
return (NULL);
memset(dmask, 0, dst->sa_len);
memcpy(dmask, smask, smask->sa_len);
dmask->ss_len = dst->sa_len;
dmask->ss_family = dst->sa_family;
return ((struct sockaddr *)dmask);
}
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
/*
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
* Writes information related to @rtinfo object to newly-allocated mbuf.
* Assumes MCLBYTES is enough to construct any message.
* Used for OS notifications of vaious events (if/ifa announces,etc)
*
* Returns allocated mbuf or NULL on failure.
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
*/
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
static struct mbuf *
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
rtsock_msg_mbuf(int type, struct rt_addrinfo *rtinfo)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
struct rt_msghdr *rtm;
struct mbuf *m;
int i;
struct sockaddr *sa;
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#ifdef INET6
struct sockaddr_storage ss;
struct sockaddr_in6 *sin6;
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int len, dlen;
switch (type) {
case RTM_DELADDR:
case RTM_NEWADDR:
len = sizeof(struct ifa_msghdr);
break;
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
case RTM_DELMADDR:
case RTM_NEWMADDR:
len = sizeof(struct ifma_msghdr);
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
case RTM_IFINFO:
len = sizeof(struct if_msghdr);
break;
Introduce an interface announcement message for the routing socket so that routing daemons and other interested parties know when an interface is attached/detached. PR: kern/33747 Obtained from: NetBSD MFC after: 2 weeks
2002-01-18 14:33:04 +00:00
case RTM_IFANNOUNCE:
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
case RTM_IEEE80211:
Introduce an interface announcement message for the routing socket so that routing daemons and other interested parties know when an interface is attached/detached. PR: kern/33747 Obtained from: NetBSD MFC after: 2 weeks
2002-01-18 14:33:04 +00:00
len = sizeof(struct if_announcemsghdr);
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
default:
len = sizeof(struct rt_msghdr);
}
- Use m_getcl() instead of hand allocating. - Convert panic() to KASSERT. - Remove superfluous cleaning of mbuf fields after allocation. - Add comment on possible use of m_get2() here. Sponsored by: Nginx, Inc.
2013-03-15 12:52:59 +00:00
/* XXXGL: can we use MJUMPAGESIZE cluster here? */
KASSERT(len <= MCLBYTES, ("%s: message too big", __func__));
if (len > MHLEN)
m = m_getcl(M_NOWAIT, MT_DATA, M_PKTHDR);
else
m = m_gethdr(M_NOWAIT, MT_DATA);
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
if (m == NULL)
Sync with recent KAME. This work was based on kame-20010528-freebsd43-snap.tgz and some critical problem after the snap was out were fixed. There are many many changes since last KAME merge. TODO: - The definitions of SADB_* in sys/net/pfkeyv2.h are still different from RFC2407/IANA assignment because of binary compatibility issue. It should be fixed under 5-CURRENT. - ip6po_m member of struct ip6_pktopts is no longer used. But, it is still there because of binary compatibility issue. It should be removed under 5-CURRENT. Reviewed by: itojun Obtained from: KAME MFC after: 3 weeks
2001-06-11 12:39:29 +00:00
return (m);
- Use m_getcl() instead of hand allocating. - Convert panic() to KASSERT. - Remove superfluous cleaning of mbuf fields after allocation. - Add comment on possible use of m_get2() here. Sponsored by: Nginx, Inc.
2013-03-15 12:52:59 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
m->m_pkthdr.len = m->m_len = len;
rtm = mtod(m, struct rt_msghdr *);
bzero((caddr_t)rtm, len);
for (i = 0; i < RTAX_MAX; i++) {
if ((sa = rtinfo->rti_info[i]) == NULL)
continue;
rtinfo->rti_addrs |= (1 << i);
route.h: introduce a macro, SA_SIZE(struct sockaddr *) which returns the space occupied by a struct sockaddr when passed through a routing socket. Use it to replace the macro ROUNDUP(int), that does the same but is redefined by every file which uses it, courtesy of the School of Cut'n'Paste Programming(TM). (partial) userland changes to follow.
2004-04-13 11:22:22 +00:00
dlen = SA_SIZE(sa);
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#ifdef INET6
- Fix LOR in sa6_recoverscope() in rt_msg2()[1]. - Check V_deembed_scopeid before checking if sa_family == AF_INET6. - Fix scope id handing in route(8)[2] and ifconfig(8). Reported by: rpaulo[1], Mateusz Guzik[1], peter[2]
2012-12-04 17:12:23 +00:00
if (V_deembed_scopeid && sa->sa_family == AF_INET6) {
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
sin6 = (struct sockaddr_in6 *)&ss;
bcopy(sa, sin6, sizeof(*sin6));
if (sa6_recoverscope(sin6) == 0)
sa = (struct sockaddr *)sin6;
}
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
m_copyback(m, len, dlen, (caddr_t)sa);
len += dlen;
}
if (m->m_pkthdr.len != len) {
m_freem(m);
return (NULL);
}
rtm->rtm_msglen = len;
rtm->rtm_version = RTM_VERSION;
rtm->rtm_type = type;
return (m);
}
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
/*
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
* Writes information related to @rtinfo object to preallocated buffer.
* Stores needed size in @plen. If @w is NULL, calculates size without
* writing.
* Used for sysctl dumps and rtsock answers (RTM_DEL/RTM_GET) generation.
*
* Returns 0 on success.
*
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
*/
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
static int
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
rtsock_msg_buffer(int type, struct rt_addrinfo *rtinfo, struct walkarg *w, int *plen)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
int i;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
int len, buflen = 0, dlen;
Fix build Found by: ian Pointyhat to: me
2014-04-27 21:17:54 +00:00
caddr_t cp = NULL;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
struct rt_msghdr *rtm = NULL;
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#ifdef INET6
struct sockaddr_storage ss;
struct sockaddr_in6 *sin6;
#endif
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
#ifdef COMPAT_FREEBSD32
bool compat32 = false;
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
switch (type) {
case RTM_DELADDR:
case RTM_NEWADDR:
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
if (w != NULL && w->w_op == NET_RT_IFLISTL) {
#ifdef COMPAT_FREEBSD32
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
if (w->w_req->flags & SCTL_MASK32) {
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
len = sizeof(struct ifa_msghdrl32);
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
compat32 = true;
} else
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#endif
len = sizeof(struct ifa_msghdrl);
} else
len = sizeof(struct ifa_msghdr);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
break;
case RTM_IFINFO:
Provide 32bit compat shims for sysctl net.route NET_RT_IFLIST. This allows getifaddrs(3) to work for compat32 binaries. Submitted by: jhb (6.x version) Reviewed by: emaste Tested by: emaste and <pluknet gmail com> MFC after: 2 weeks
2010-04-25 16:42:47 +00:00
#ifdef COMPAT_FREEBSD32
if (w != NULL && w->w_req->flags & SCTL_MASK32) {
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
if (w->w_op == NET_RT_IFLISTL)
len = sizeof(struct if_msghdrl32);
else
len = sizeof(struct if_msghdr32);
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
compat32 = true;
Provide 32bit compat shims for sysctl net.route NET_RT_IFLIST. This allows getifaddrs(3) to work for compat32 binaries. Submitted by: jhb (6.x version) Reviewed by: emaste Tested by: emaste and <pluknet gmail com> MFC after: 2 weeks
2010-04-25 16:42:47 +00:00
break;
}
#endif
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
if (w != NULL && w->w_op == NET_RT_IFLISTL)
len = sizeof(struct if_msghdrl);
else
len = sizeof(struct if_msghdr);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
break;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
case RTM_NEWMADDR:
len = sizeof(struct ifma_msghdr);
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
default:
len = sizeof(struct rt_msghdr);
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if (w != NULL) {
rtm = (struct rt_msghdr *)w->w_tmem;
buflen = w->w_tmemsize - len;
cp = (caddr_t)w->w_tmem + len;
}
rtinfo->rti_addrs = 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
for (i = 0; i < RTAX_MAX; i++) {
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
struct sockaddr *sa;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
if ((sa = rtinfo->rti_info[i]) == NULL)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
continue;
rtinfo->rti_addrs |= (1 << i);
Fix compat32 for sysctl net.PF_ROUTE...NET_RT_IFLISTL. Route messages are aligned to the host long type alignment, which breaks 32bit. Reported and tested by: lwhsu Diagnosed by: Yuri Pankov <yuripv@icloud.com> Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-01-22 20:49:17 +00:00
#ifdef COMPAT_FREEBSD32
if (compat32)
dlen = SA_SIZE32(sa);
else
#endif
dlen = SA_SIZE(sa);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if (cp != NULL && buflen >= dlen) {
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#ifdef INET6
- Move definition of V_deembed_scopeid to scope6_var.h. - Deembed scope id in L3 address in in6_lltable_dump(). - Simplify scope id recovery in rtsock routines. - Remove embedded scope id handling in ndp(8) and route(8) completely.
2012-12-05 19:45:24 +00:00
if (V_deembed_scopeid && sa->sa_family == AF_INET6) {
sin6 = (struct sockaddr_in6 *)&ss;
bcopy(sa, sin6, sizeof(*sin6));
if (sa6_recoverscope(sin6) == 0)
sa = (struct sockaddr *)sin6;
}
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to userland via routing socket or sysctl. This eliminates the following KAME-specific sin6_scope_id handling routine from each userland utility: sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]); This behavior can be controlled by net.inet6.ip6.deembed_scopeid. This is set to 1 by default (sin6_scope_id will be filled in the kernel). Reviewed by: bz
2012-11-17 20:19:00 +00:00
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
bcopy((caddr_t)sa, cp, (unsigned)dlen);
cp += dlen;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
buflen -= dlen;
} else if (cp != NULL) {
/*
* Buffer too small. Count needed size
* and return with error.
*/
cp = NULL;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
len += dlen;
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if (cp != NULL) {
dlen = ALIGN(len) - len;
if (buflen < dlen)
cp = NULL;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
else {
bzero(cp, dlen);
cp += dlen;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
buflen -= dlen;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
len = ALIGN(len);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if (cp != NULL) {
/* fill header iff buffer is large enough */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
rtm->rtm_version = RTM_VERSION;
rtm->rtm_type = type;
rtm->rtm_msglen = len;
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
*plen = len;
if (w != NULL && cp == NULL)
return (ENOBUFS);
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* This routine is called to generate a message from the routing
sys/net*: minor spelling fixes. No functional change.
2016-05-03 18:05:43 +00:00
* socket indicating that a redirect has occurred, a routing lookup
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* has failed, or that a protocol has detected timeouts to a particular
* destination.
*/
void
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
rt_missmsg_fib(int type, struct rt_addrinfo *rtinfo, int flags, int error,
int fibnum)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct rt_msghdr *rtm;
struct mbuf *m;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct sockaddr *sa = rtinfo->rti_info[RTAX_DST];
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return;
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
m = rtsock_msg_mbuf(type, rtinfo);
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
if (m == NULL)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return;
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
Constanly use RT_ALL_FIBS everywhere instead of -1. MFC after: 2 weeks
2014-01-08 23:09:02 +00:00
if (fibnum != RT_ALL_FIBS) {
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
KASSERT(fibnum >= 0 && fibnum < rt_numfibs, ("%s: fibnum out "
"of range 0 <= %d < %d", __func__, fibnum, rt_numfibs));
M_SETFIB(m, fibnum);
m->m_flags |= RTS_FILTER_FIB;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
rtm = mtod(m, struct rt_msghdr *);
rtm->rtm_flags = RTF_DONE | flags;
rtm->rtm_errno = error;
rtm->rtm_addrs = rtinfo->rti_addrs;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, sa ? sa->sa_family : AF_UNSPEC);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
void
rt_missmsg(int type, struct rt_addrinfo *rtinfo, int flags, int error)
{
Constanly use RT_ALL_FIBS everywhere instead of -1. MFC after: 2 weeks
2014-01-08 23:09:02 +00:00
rt_missmsg_fib(type, rtinfo, flags, error, RT_ALL_FIBS);
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* This routine is called to generate a message from the routing
* socket indicating that the status of a network interface has changed.
*/
void
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
rt_ifmsg(struct ifnet *ifp)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct if_msghdr *ifm;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct mbuf *m;
struct rt_addrinfo info;
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return;
bzero((caddr_t)&info, sizeof(info));
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
m = rtsock_msg_mbuf(RTM_IFINFO, &info);
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
if (m == NULL)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return;
ifm = mtod(m, struct if_msghdr *);
ifm->ifm_index = ifp->if_index;
Rename IFF_RUNNING to IFF_DRV_RUNNING, IFF_OACTIVE to IFF_DRV_OACTIVE, and move both flags from ifnet.if_flags to ifnet.if_drv_flags, making and documenting the locking of these flags the responsibility of the device driver, not the network stack. The flags for these two fields will be mutually exclusive so that they can be exposed to user space as though they were stored in the same variable. Provide #defines to provide the old names #ifndef _KERNEL, so that user applications (such as ifconfig) can use the old flag names. Using the old names in a device driver will result in a compile error in order to help device driver writers adopt the new model. When exposing the interface flags to user space, via interface ioctls or routing sockets, or the two fields together. Since the driver flags cannot currently be set for user space, no new logic is currently required to handle this case. Add some assertions that general purpose network stack routines, such as if_setflags(), are not improperly used on driver-owned flags. With this change, a large number of very minor network stack races are closed, subject to correct device driver locking. Most were likely never triggered. Driver sweep to follow; many thanks to pjd and bz for the line-by-line review they gave this patch. Reviewed by: pjd, bz MFC after: 7 days
2005-08-09 10:16:17 +00:00
ifm->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
o Remove struct if_data from struct ifnet. Now it is merely API structure for route(4) socket and ifmib(4) sysctl. o Move fields from if_data to ifnet, but keep all statistic counters separate, since they should disappear later. o Provide function if_data_copy() to fill if_data, utilize it in routing socket and ifmib handler. o Provide overridable ifnet(9) method to fetch counters. If no provided, if_get_counters_compat() would be used, that returns old counters. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-08-31 06:46:21 +00:00
if_data_copy(ifp, &ifm->ifm_data);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
ifm->ifm_addrs = 0;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, AF_UNSPEC);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
* Announce interface address arrival/withdraw.
* Please do not call directly, use rt_addrmsg().
* Assume input data to be valid.
* Returns 0 on success.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
int
rtsock_addrmsg(int cmd, struct ifaddr *ifa, int fibnum)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct rt_addrinfo info;
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
struct sockaddr *sa;
int ncmd;
struct mbuf *m;
struct ifa_msghdr *ifam;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct ifnet *ifp = ifa->ifa_ifp;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
struct sockaddr_storage ss;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
return (0);
Remove trailing whitespace.
1995-05-30 08:16:23 +00:00
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
ncmd = cmd == RTM_ADD ? RTM_NEWADDR : RTM_DELADDR;
bzero((caddr_t)&info, sizeof(info));
info.rti_info[RTAX_IFA] = sa = ifa->ifa_addr;
info.rti_info[RTAX_IFP] = ifp->if_addr->ifa_addr;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
info.rti_info[RTAX_NETMASK] = rtsock_fix_netmask(
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
info.rti_info[RTAX_IFA], ifa->ifa_netmask, &ss);
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
info.rti_info[RTAX_BRD] = ifa->ifa_dstaddr;
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
if ((m = rtsock_msg_mbuf(ncmd, &info)) == NULL)
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
return (ENOBUFS);
ifam = mtod(m, struct ifa_msghdr *);
ifam->ifam_index = ifp->if_index;
Garbage collect couple of unused fields from struct ifaddr: - ifa_claim_addr() unused since removal of NetAtalk - ifa_metric seems to be never utilized, always a copy of if_metric
2014-07-29 15:01:29 +00:00
ifam->ifam_metric = ifa->ifa_ifp->if_metric;
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
ifam->ifam_flags = ifa->ifa_flags;
ifam->ifam_addrs = info.rti_addrs;
if (fibnum != RT_ALL_FIBS) {
M_SETFIB(m, fibnum);
m->m_flags |= RTS_FILTER_FIB;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
rt_dispatch(m, sa ? sa->sa_family : AF_UNSPEC);
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
/*
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
* Announce route addition/removal to rtsock based on @rt data.
* Callers are advives to use rt_routemsg() instead of using this
* function directly.
* Assume @rt data is consistent.
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
*
* Returns 0 on success.
*/
int
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
rtsock_routemsg(int cmd, struct rtentry *rt, struct ifnet *ifp, int rti_addrs,
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
int fibnum)
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
{
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
struct sockaddr_storage ss;
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
struct rt_addrinfo info;
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
return (0);
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
bzero((caddr_t)&info, sizeof(info));
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
info.rti_info[RTAX_DST] = rt_key(rt);
info.rti_info[RTAX_NETMASK] = rtsock_fix_netmask(rt_key(rt), rt_mask(rt), &ss);
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
info.rti_info[RTAX_GATEWAY] = rt->rt_gateway;
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
info.rti_flags = rt->rt_flags;
info.rti_ifp = ifp;
return (rtsock_routemsg_info(cmd, &info, fibnum));
}
int
rtsock_routemsg_info(int cmd, struct rt_addrinfo *info, int fibnum)
{
struct rt_msghdr *rtm;
struct sockaddr *sa;
struct mbuf *m;
if (V_route_cb.any_count == 0)
return (0);
if (info->rti_flags & RTF_HOST)
info->rti_info[RTAX_NETMASK] = NULL;
m = rtsock_msg_mbuf(cmd, info);
if (m == NULL)
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
return (ENOBUFS);
if (fibnum != RT_ALL_FIBS) {
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
KASSERT(fibnum >= 0 && fibnum < rt_numfibs, ("%s: fibnum out "
"of range 0 <= %d < %d", __func__, fibnum, rt_numfibs));
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
M_SETFIB(m, fibnum);
m->m_flags |= RTS_FILTER_FIB;
}
Fix rtsock route message generation for interface addresses. Reviewed by: olivier MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D22974
2020-01-07 21:16:30 +00:00
rtm = mtod(m, struct rt_msghdr *);
rtm->rtm_addrs = info->rti_addrs;
if (info->rti_ifp != NULL)
rtm->rtm_index = info->rti_ifp->if_index;
/* Add RTF_DONE to indicate command 'completion' required by API */
info->rti_flags |= RTF_DONE;
/* Reported routes has to be up */
if (cmd == RTM_ADD || cmd == RTM_CHANGE)
info->rti_flags |= RTF_UP;
rtm->rtm_flags = info->rti_flags;
sa = info->rti_info[RTAX_DST];
Split rt_newaddrmsg_fib() into two different functions. Adding/deleting interface addresses involves access to 3 different subsystems, int different parts of code. Each call can fail, so reporting successful operation by rtsock in the middle of the process error-prone. Further split routing notification API and actual rtsock calls via creating public-available rt_addrmsg() / rt_routemsg() functions with "private" rtsock_* backend. MFC after: 2 weeks
2014-01-09 18:13:25 +00:00
rt_dispatch(m, sa ? sa->sa_family : AF_UNSPEC);
return (0);
Pass the fibnum where we need filtering of the message on the rtsock allowing routing daemons to filter routing updates on an rtsock per FIB. Adjust raw_input() and split it into wrapper and a new function taking an optional callback argument even though we only have one consumer [1] to keep the hackish flags local to rtsock.c. PR: kern/134931 Submitted by: multiple (see PR) Suggested by: rwatson [1] Reviewed by: rwatson MFC after: 3 days
2011-09-28 13:48:36 +00:00
}
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
/*
* This is the analogue to the rt_newaddrmsg which performs the same
* function but for multicast group memberhips. This is easier since
* there is no route state to worry about.
*/
void
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
rt_newmaddrmsg(int cmd, struct ifmultiaddr *ifma)
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
{
struct rt_addrinfo info;
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
struct mbuf *m = NULL;
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
struct ifnet *ifp = ifma->ifma_ifp;
struct ifma_msghdr *ifmam;
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
return;
bzero((caddr_t)&info, sizeof(info));
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_IFA] = ifma->ifma_addr;
The ifnet pointer (ifp) in rt_newaddrmsg can be valid without ifp->if_addr being set if if the ifnet is still live by way of a reference but in line for deletion. Check ifp->if_addr before dereferencing. Approved by: sbruno
2018-05-06 20:32:47 +00:00
if (ifp && ifp->if_addr)
info.rti_info[RTAX_IFP] = ifp->if_addr->ifa_addr;
else
info.rti_info[RTAX_IFP] = NULL;
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
/*
* If a link-layer address is present, present it as a ``gateway''
* (similarly to how ARP entries, e.g., are presented).
*/
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_GATEWAY] = ifma->ifma_lladdr;
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
m = rtsock_msg_mbuf(cmd, &info);
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
if (m == NULL)
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
return;
ifmam = mtod(m, struct ifma_msghdr *);
Fix a case where hardware removal of an interface caused an attempt to announce an ll_ifma which has gone away. Add a KASSERT to catch regressions. Bug found by: Tom Uffner
2007-03-27 16:11:28 +00:00
KASSERT(ifp != NULL, ("%s: link-layer multicast address w/o ifp\n",
__func__));
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
ifmam->ifmam_index = ifp->if_index;
ifmam->ifmam_addrs = info.rti_addrs;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, ifma->ifma_addr ? ifma->ifma_addr->sa_family : AF_UNSPEC);
Use the new if_multiaddrs list for multicast addresses rather than the previous hackery involving struct in_ifaddr and arpcom. Get rid of the abominable multi_kludge. Update all network interfaces to use the new machanism. Distressingly few Ethernet drivers program the multicast filter properly (assuming the hardware has one, which it usually does).
1997-01-13 21:26:53 +00:00
}
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
static struct mbuf *
rt_makeifannouncemsg(struct ifnet *ifp, int type, int what,
struct rt_addrinfo *info)
{
struct if_announcemsghdr *ifan;
struct mbuf *m;
V_irtualize rtsock refcounting, which reduces the chances for panics on teardown of vnets without active routing sockets while at least one routing socket is active elsewhere. Tested by: Vijay Singh MFC after: 3 days
2014-02-19 08:29:07 +00:00
if (V_route_cb.any_count == 0)
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
return NULL;
bzero((caddr_t)info, sizeof(*info));
Rename rt_msg1() to more handy rtsock_msg_mbuf(). (Just for history purposes: rt_msg2() was renamed to rtsock_msg_buffer() in r265019). Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 13:54:57 +00:00
m = rtsock_msg_mbuf(type, info);
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
if (m != NULL) {
ifan = mtod(m, struct if_announcemsghdr *);
ifan->ifan_index = ifp->if_index;
strlcpy(ifan->ifan_name, ifp->if_xname,
sizeof(ifan->ifan_name));
ifan->ifan_what = what;
}
return m;
}
/*
* This is called to generate routing socket messages indicating
* IEEE80211 wireless events.
* XXX we piggyback on the RTM_IFANNOUNCE msg format in a clumsy way.
*/
void
rt_ieee80211msg(struct ifnet *ifp, int what, void *data, size_t data_len)
{
struct mbuf *m;
struct rt_addrinfo info;
m = rt_makeifannouncemsg(ifp, RTM_IEEE80211, what, &info);
if (m != NULL) {
/*
* Append the ieee80211 data. Try to stick it in the
* mbuf containing the ifannounce msg; otherwise allocate
* a new mbuf and append.
*
* NB: we assume m is a single mbuf.
*/
if (data_len > M_TRAILINGSPACE(m)) {
struct mbuf *n = m_get(M_NOWAIT, MT_DATA);
if (n == NULL) {
m_freem(m);
return;
}
bcopy(data, mtod(n, void *), data_len);
n->m_len = data_len;
m->m_next = n;
} else if (data_len > 0) {
bcopy(data, mtod(m, u_int8_t *) + m->m_len, data_len);
m->m_len += data_len;
}
if (m->m_flags & M_PKTHDR)
m->m_pkthdr.len += data_len;
mtod(m, struct if_announcemsghdr *)->ifan_msglen += data_len;
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, AF_UNSPEC);
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
}
}
Introduce an interface announcement message for the routing socket so that routing daemons and other interested parties know when an interface is attached/detached. PR: kern/33747 Obtained from: NetBSD MFC after: 2 weeks
2002-01-18 14:33:04 +00:00
/*
* This is called to generate routing socket messages indicating
* network interface arrival and departure.
*/
void
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
rt_ifannouncemsg(struct ifnet *ifp, int what)
Introduce an interface announcement message for the routing socket so that routing daemons and other interested parties know when an interface is attached/detached. PR: kern/33747 Obtained from: NetBSD MFC after: 2 weeks
2002-01-18 14:33:04 +00:00
{
struct mbuf *m;
struct rt_addrinfo info;
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
m = rt_makeifannouncemsg(ifp, RTM_IFANNOUNCE, what, &info);
if (m != NULL)
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(m, AF_UNSPEC);
Add 802.11-specific events that are dispatched through the routing socket. This really doesn't belong here but is preferred (for the moment) over adding yet another mechanism for sending msgs from the kernel to user apps. Reviewed by: imp
2004-10-05 19:48:33 +00:00
}
Introduce an interface announcement message for the routing socket so that routing daemons and other interested parties know when an interface is attached/detached. PR: kern/33747 Obtained from: NetBSD MFC after: 2 weeks
2002-01-18 14:33:04 +00:00
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
static void
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
rt_dispatch(struct mbuf *m, sa_family_t saf)
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
{
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
struct m_tag *tag;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
/*
* Preserve the family from the sockaddr, if any, in an m_tag for
* use when injecting the mbuf into the routing socket buffer from
* the netisr.
*/
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
if (saf != AF_UNSPEC) {
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
tag = m_tag_get(PACKET_TAG_RTSOCKFAM, sizeof(unsigned short),
M_NOWAIT);
if (tag == NULL) {
m_freem(m);
return;
}
Fix a use-after-free/redzone issue in the routing code. Reported by (repeatedly): Mike Tancsa Prodded by (repeatedly): bz Forgotten by (repeatedly): mlaier MFC after: 2 weeks
2011-11-03 18:33:30 +00:00
*(unsigned short *)(tag + 1) = saf;
Introduce a netisr to deliver kernel-generated routing, avoiding recursive entering of the socket code from the routing code: - Modify rt_dispatch() to bundle up the sockaddr family, if any, associated with a pending mbuf to dispatch to routing sockets, in an m_tag on the mbuf. - Allocate NETISR_ROUTE for use by routing sockets. - Introduce rtsintrq, an ifqueue to be used by the netisr, and introduce rts_input(), a function to unbundle the tagged sockaddr and inject the mbuf and address into raw_input(), which previously occurred in rt_dispatch(). - Introduce rts_init() to initialize rtsintrq, its mutex, and register the netisr. Perform this at the same point in system initialization as setup of the domains. This change introduces asynchrony between the generation of a pending routing socket message and delivery to sockets for use by userspace. It avoids socket->routing->rtsock->socket use and helps to avoid lock order reversals between the routing code and socket code (in particular, raw socket control blocks), as route locks are held over calls to rt_dispatch(). Reviewed by: "George V.Neville-Neil" <gnn@neville-neil.com> Conceptual head nod by: sam
2004-06-09 02:48:23 +00:00
m_tag_prepend(m, tag);
}
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
#ifdef VIMAGE
if (V_loif)
m->m_pkthdr.rcvif = V_loif;
else {
m_freem(m);
return;
}
#endif
Apply error and success logic consistently to the function netisr_queue() and its users. netisr_queue() now returns (0) on success and ERRNO on failure. At the moment ENXIO (netisr queue not functional) and ENOBUFS (netisr queue full) are supported. Previously it would return (1) on success but the return value of IF_HANDOFF() was interpreted wrongly and (0) was actually returned on success. Due to this schednetisr() was never called to kick the scheduling of the isr. However this was masked by other normal packets coming through netisr_dispatch() causing the dequeueing of waiting packets. PR: kern/70988 Found by: MOROHOSHI Akihiko <moro@remus.dti.ne.jp> MFC after: 3 days
2004-08-27 18:33:08 +00:00
netisr_queue(NETISR_ROUTE, m); /* mbuf is free'd on failure. */
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
}
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
/*
* Checks if rte can be exported v.r.t jails/vnets.
*
* Returns 1 if it can, 0 otherwise.
*/
static int
can_export_rte(struct ucred *td_ucred, const struct rtentry *rt)
{
if ((rt->rt_flags & RTF_HOST) == 0
? jailed_without_vnet(td_ucred)
: prison_if(td_ucred, rt_key_const(rt)) != 0)
return (0);
return (1);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* This is used in dumping the kernel table via sysctl().
*/
Be consistent about "static" functions: if the function is marked static in its prototype, mark it static at the definition too. Inspired by: FlexeLint warning #512
2002-09-28 17:15:38 +00:00
static int
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
sysctl_dumpentry(struct radix_node *rn, void *vw)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct walkarg *w = vw;
struct rtentry *rt = (struct rtentry *)rn;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error = 0, size;
struct rt_addrinfo info;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
struct sockaddr_storage ss;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Mechanical cleanup of epoch(9) usage in network stack. - Remove macros that covertly create epoch_tracker on thread stack. Such macros a quite unsafe, e.g. will produce a buggy code if same macro is used in embedded scopes. Explicitly declare epoch_tracker always. - Unmask interface list IFNET_RLOCK_NOSLEEP(), interface address list IF_ADDR_RLOCK() and interface AF specific data IF_AFDATA_RLOCK() read locking macros to what they actually are - the net_epoch. Keeping them as is is very misleading. They all are named FOO_RLOCK(), while they no longer have lock semantics. Now they allow recursion and what's more important they now no longer guarantee protection against their companion WLOCK macros. Note: INP_HASH_RLOCK() has same problems, but not touched by this commit. This is non functional mechanical change. The only functionally changed functions are ni6_addrs() and ni6_store_addrs(), where we no longer enter epoch recursively. Discussed with: jtl, gallatin
2019-01-09 01:11:19 +00:00
NET_EPOCH_ASSERT();
Fix possible panic during ifnet detach in rtsock. The panic can happen, when some application does dump of routing table using sysctl interface. To prevent this, set IFF_DYING flag in if_detach_internal() function, when ifnet under lock is removed from the chain. In sysctl_rtsock() take IFNET_RLOCK_NOSLEEP() to prevent ifnet detach during routes enumeration. In case, if some interface was detached in the time before we take the lock, add the check, that ifnet is not DYING. This prevents access to memory that could be freed after ifnet is unlinked. PR: 227720, 230498, 233306 Reviewed by: bz, eugen MFC after: 1 week Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D18338
2018-11-27 09:04:06 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (w->w_op == NET_RT_FLAGS && !(rt->rt_flags & w->w_arg))
return 0;
Split gigantic rtsock route_output() into smaller functions. Amount of changes to the original code has been intentionally minimised to ease diffing. The changes are mostly mechanical, with the following exceptions: * lltable handler is now called directly based of RTF_LLINFO flag presense. * "report" logic for updating rtm in RTM_GET/RTM_DELETE has been simplified, fixing several potential use-after-free cases in rt_addrinfo. * llable asserts has been replaced with error-returning, preventing kernel crashes when lltable gw af family is invalid (root required). MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D22864
2019-12-31 17:26:53 +00:00
if (!can_export_rte(w->w_req->td->td_ucred, rt))
Restrict arp, ndp and theoretically the FIB listing (if not read with libkvm) to the addresses of a prison, when inside a jail. [1] As the patch from the PR was pre-'new-arp', add checks to the llt_dump handlers as well. While touching RTM_GET in route_output(), consistently use curthread credentials rather than the creds from the socket there. [2] PR: kern/68189 Submitted by: Mark Delany <sxcg2-fuwxj@qmda.emu.st> [1] Discussed with: rwatson [2] Reviewed by: rwatson MFC after: 4 weeks
2009-01-09 21:57:49 +00:00
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
bzero((caddr_t)&info, sizeof(info));
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_DST] = rt_key(rt);
info.rti_info[RTAX_GATEWAY] = rt->rt_gateway;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
info.rti_info[RTAX_NETMASK] = rtsock_fix_netmask(rt_key(rt),
rt_mask(rt), &ss);
This main goals of this project are: 1. separating L2 tables (ARP, NDP) from the L3 routing tables 2. removing as much locking dependencies among these layers as possible to allow for some parallelism in the search operations 3. simplify the logic in the routing code, The most notable end result is the obsolescent of the route cloning (RTF_CLONING) concept, which translated into code reduction in both IPv4 ARP and IPv6 NDP related modules, and size reduction in struct rtentry{}. The change in design obsoletes the semantics of RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland applications such as "arp" and "ndp" have been modified to reflect those changes. The output from "netstat -r" shows only the routing entries. Quite a few developers have contributed to this project in the past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and Andre Oppermann. And most recently: - Kip Macy revised the locking code completely, thus completing the last piece of the puzzle, Kip has also been conducting active functional testing - Sam Leffler has helped me improving/refactoring the code, and provided valuable reviews - Julian Elischer setup the perforce tree for me and has helped me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
info.rti_info[RTAX_GENMASK] = 0;
Fix possible panic during ifnet detach in rtsock. The panic can happen, when some application does dump of routing table using sysctl interface. To prevent this, set IFF_DYING flag in if_detach_internal() function, when ifnet under lock is removed from the chain. In sysctl_rtsock() take IFNET_RLOCK_NOSLEEP() to prevent ifnet detach during routes enumeration. In case, if some interface was detached in the time before we take the lock, add the check, that ifnet is not DYING. This prevents access to memory that could be freed after ifnet is unlinked. PR: 227720, 230498, 233306 Reviewed by: bz, eugen MFC after: 1 week Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D18338
2018-11-27 09:04:06 +00:00
if (rt->rt_ifp && !(rt->rt_ifp->if_flags & IFF_DYING)) {
- Store pointer to the link-level address right in "struct ifnet" rather than in ifindex_table[]; all (except one) accesses are through ifp anyway. IF_LLADDR() works faster, and all (except one) ifaddr_byindex() users were converted to use ifp->if_addr. - Stop storing a (pointer to) Ethernet address in "struct arpcom", and drop the IFP2ENADDR() macro; all users have been converted to use IF_LLADDR() instead.
2005-11-11 16:04:59 +00:00
info.rti_info[RTAX_IFP] = rt->rt_ifp->if_addr->ifa_addr;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_IFA] = rt->rt_ifa->ifa_addr;
Bring in latest CSRG revisions to this file: - Report destination address of a P2P link when servicing routing socket messages. - Report interface name, address, and destination address of a P2P link when servicing NET_RT_{DUMP,FLAGS} sysctls. Part of CSRG revision 8.6 coresponds to revision 1.12. CSRG revision 8.7 corresponds to revision 1.15.
2001-10-17 11:23:59 +00:00
if (rt->rt_ifp->if_flags & IFF_POINTOPOINT)
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_BRD] = rt->rt_ifa->ifa_dstaddr;
Bring in latest CSRG revisions to this file: - Report destination address of a P2P link when servicing routing socket messages. - Report interface name, address, and destination address of a P2P link when servicing NET_RT_{DUMP,FLAGS} sysctls. Part of CSRG revision 8.6 coresponds to revision 1.12. CSRG revision 8.7 corresponds to revision 1.15.
2001-10-17 11:23:59 +00:00
}
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if ((error = rtsock_msg_buffer(RTM_GET, &info, w, &size)) != 0)
return (error);
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
if (w->w_req && w->w_tmem) {
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct rt_msghdr *rtm = (struct rt_msghdr *)w->w_tmem;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
bzero(&rtm->rtm_index,
sizeof(*rtm) - offsetof(struct rt_msghdr, rtm_index));
Due to the routing related networking kernel redesign work in FBSD 8.0, interface routes have been returened to the applications without the RTF_GATEWAY bit. This incompatibility has caused some issues with Zebra, Qugga and the like. This patch provides the RTF_GATEWAY flag bit in returned interface routes so to behave similarly to pre 8.0 systems. Reviewed by: hrs Verified by: mackn at opendns dot com
2013-06-25 00:10:49 +00:00
if (rt->rt_flags & RTF_GWFLAG_COMPAT)
rtm->rtm_flags = RTF_GATEWAY |
(rt->rt_flags & ~RTF_GWFLAG_COMPAT);
else
rtm->rtm_flags = rt->rt_flags;
- Remove rt_metrics_lite and simply put its members into rtentry. - Use counter(9) for rt_pksent (former rt_rmx.rmx_pksent). This removes another cache trashing ++ from packet forwarding path. - Create zini/fini methods for the rtentry UMA zone. Via initialize mutex and counter in them. - Fix reporting of rmx_pksent to routing socket. - Fix netstat(1) to report "Use" both in kvm(3) and sysctl(3) mode. The change is mostly targeted for stable/10 merge. For head, rt_pksent is expected to just disappear. Discussed with: melifaro Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-05 01:17:47 +00:00
rt_getmetrics(rt, &rtm->rtm_rmx);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
rtm->rtm_index = rt->rt_ifp->if_index;
rtm->rtm_addrs = info.rti_addrs;
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
error = SYSCTL_OUT(w->w_req, (caddr_t)rtm, size);
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
return (error);
}
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
static int
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
sysctl_iflist_ifml(struct ifnet *ifp, const struct if_data *src_ifd,
struct rt_addrinfo *info, struct walkarg *w, int len)
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
{
struct if_msghdrl *ifm;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data *ifd;
ifm = (struct if_msghdrl *)w->w_tmem;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#ifdef COMPAT_FREEBSD32
if (w->w_req->flags & SCTL_MASK32) {
struct if_msghdrl32 *ifm32;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifm32 = (struct if_msghdrl32 *)ifm;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
ifm32->ifm_addrs = info->rti_addrs;
ifm32->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
ifm32->ifm_index = ifp->if_index;
ifm32->_ifm_spare1 = 0;
ifm32->ifm_len = sizeof(*ifm32);
ifm32->ifm_data_off = offsetof(struct if_msghdrl32, ifm_data);
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifm32->_ifm_spare2 = 0;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifm32->ifm_data;
} else
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#endif
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
{
ifm->ifm_addrs = info->rti_addrs;
ifm->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
ifm->ifm_index = ifp->if_index;
ifm->_ifm_spare1 = 0;
ifm->ifm_len = sizeof(*ifm);
ifm->ifm_data_off = offsetof(struct if_msghdrl, ifm_data);
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifm->_ifm_spare2 = 0;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifm->ifm_data;
}
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
memcpy(ifd, src_ifd, sizeof(*ifd));
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
return (SYSCTL_OUT(w->w_req, (caddr_t)ifm, len));
}
static int
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
sysctl_iflist_ifm(struct ifnet *ifp, const struct if_data *src_ifd,
struct rt_addrinfo *info, struct walkarg *w, int len)
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
{
struct if_msghdr *ifm;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data *ifd;
ifm = (struct if_msghdr *)w->w_tmem;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#ifdef COMPAT_FREEBSD32
if (w->w_req->flags & SCTL_MASK32) {
struct if_msghdr32 *ifm32;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifm32 = (struct if_msghdr32 *)ifm;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
ifm32->ifm_addrs = info->rti_addrs;
ifm32->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
ifm32->ifm_index = ifp->if_index;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifm32->_ifm_spare1 = 0;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifm32->ifm_data;
} else
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#endif
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
{
ifm->ifm_addrs = info->rti_addrs;
ifm->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
ifm->ifm_index = ifp->if_index;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifm->_ifm_spare1 = 0;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifm->ifm_data;
}
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
memcpy(ifd, src_ifd, sizeof(*ifd));
When exporting ifnet via sysctl, add ifqueue(9) drop count to the ifi_oqdrops. This is a temporary workaround until ifqueue(9) vanishes. While here, remove the pointless ifi_vhid assignment. It has sense only when we are exporting ifaddrs, not ifnets. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-19 06:08:03 +00:00
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
return (SYSCTL_OUT(w->w_req, (caddr_t)ifm, len));
}
static int
sysctl_iflist_ifaml(struct ifaddr *ifa, struct rt_addrinfo *info,
struct walkarg *w, int len)
{
struct ifa_msghdrl *ifam;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
struct if_data *ifd;
ifam = (struct ifa_msghdrl *)w->w_tmem;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#ifdef COMPAT_FREEBSD32
if (w->w_req->flags & SCTL_MASK32) {
struct ifa_msghdrl32 *ifam32;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifam32 = (struct ifa_msghdrl32 *)ifam;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
ifam32->ifam_addrs = info->rti_addrs;
ifam32->ifam_flags = ifa->ifa_flags;
ifam32->ifam_index = ifa->ifa_ifp->if_index;
ifam32->_ifam_spare1 = 0;
ifam32->ifam_len = sizeof(*ifam32);
ifam32->ifam_data_off =
offsetof(struct ifa_msghdrl32, ifam_data);
Garbage collect couple of unused fields from struct ifaddr: - ifa_claim_addr() unused since removal of NetAtalk - ifa_metric seems to be never utilized, always a copy of if_metric
2014-07-29 15:01:29 +00:00
ifam32->ifam_metric = ifa->ifa_ifp->if_metric;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifam32->ifam_data;
} else
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
#endif
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
{
ifam->ifam_addrs = info->rti_addrs;
ifam->ifam_flags = ifa->ifa_flags;
ifam->ifam_index = ifa->ifa_ifp->if_index;
ifam->_ifam_spare1 = 0;
ifam->ifam_len = sizeof(*ifam);
ifam->ifam_data_off = offsetof(struct ifa_msghdrl, ifam_data);
Garbage collect couple of unused fields from struct ifaddr: - ifa_claim_addr() unused since removal of NetAtalk - ifa_metric seems to be never utilized, always a copy of if_metric
2014-07-29 15:01:29 +00:00
ifam->ifam_metric = ifa->ifa_ifp->if_metric;
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd = &ifam->ifam_data;
}
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
bzero(ifd, sizeof(*ifd));
ifd->ifi_datalen = sizeof(struct if_data);
ifd->ifi_ipackets = counter_u64_fetch(ifa->ifa_ipackets);
ifd->ifi_opackets = counter_u64_fetch(ifa->ifa_opackets);
ifd->ifi_ibytes = counter_u64_fetch(ifa->ifa_ibytes);
ifd->ifi_obytes = counter_u64_fetch(ifa->ifa_obytes);
- Utilize counter(9) to accumulate statistics on interface addresses. Add four counters to struct ifaddr. This kills '+=' on a variables shared between processors for every packet. - Nuke struct if_data from struct ifaddr. - In ip_input() do not put a reference on ifaddr, instead update statistics right now in place and do IN_IFADDR_RUNLOCK(). These removes atomic(9) for every packet. [1] - To properly support NET_RT_IFLISTL sysctl used by getifaddrs(3), in rtsock.c fill if_data fields using counter_u64_fetch(). - Accidentially fix bug in COMPAT_32 version of NET_RT_IFLISTL, which took if_data not from the ifaddr, but from ifaddr's ifnet. [2] Submitted by: melifaro [1], pluknet[2] Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-15 11:37:57 +00:00
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
/* Fixup if_data carp(4) vhid. */
if (carp_get_vhid_p != NULL)
Since 32-bit if_baudrate isn't enough to describe a baud rate of a 10 Gbit interface, in the r241616 a crutch was provided. It didn't work well, and finally we decided that it is time to break ABI and simply make if_baudrate a 64-bit value. Meanwhile, the entire struct if_data was reviewed. o Remove the if_baudrate_pf crutch. o Make all fields of struct if_data fixed machine independent size. The notion of data (packet counters, etc) are by no means MD. And it is a bug that on amd64 we've got a 64-bit counters, while on i386 32-bit, which at modern speeds overflow within a second. This also removes quite a lot of COMPAT_FREEBSD32 code. o Give 16 bit for the ifi_datalen field. This field was provided to make future changes to if_data less ABI breaking. Unfortunately the 8 bit size of it had effectively limited sizeof if_data to 256 bytes. o Give 32 bits to ifi_mtu and ifi_metric. o Give 64 bits to the rest of fields, since they are counters. __FreeBSD_version bumped. Discussed with: emax Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-03-13 03:42:24 +00:00
ifd->ifi_vhid = (*carp_get_vhid_p)(ifa);
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
return (SYSCTL_OUT(w->w_req, w->w_tmem, len));
}
static int
sysctl_iflist_ifam(struct ifaddr *ifa, struct rt_addrinfo *info,
struct walkarg *w, int len)
{
struct ifa_msghdr *ifam;
ifam = (struct ifa_msghdr *)w->w_tmem;
ifam->ifam_addrs = info->rti_addrs;
ifam->ifam_flags = ifa->ifa_flags;
ifam->ifam_index = ifa->ifa_ifp->if_index;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifam->_ifam_spare1 = 0;
Garbage collect couple of unused fields from struct ifaddr: - ifa_claim_addr() unused since removal of NetAtalk - ifa_metric seems to be never utilized, always a copy of if_metric
2014-07-29 15:01:29 +00:00
ifam->ifam_metric = ifa->ifa_ifp->if_metric;
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
return (SYSCTL_OUT(w->w_req, w->w_tmem, len));
}
Be consistent about "static" functions: if the function is marked static in its prototype, mark it static at the definition too. Inspired by: FlexeLint warning #512
2002-09-28 17:15:38 +00:00
static int
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
sysctl_iflist(int af, struct walkarg *w)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct ifnet *ifp;
struct ifaddr *ifa;
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
struct if_data ifd;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
struct rt_addrinfo info;
int len, error = 0;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
struct sockaddr_storage ss;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
bzero((caddr_t)&info, sizeof(info));
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
bzero(&ifd, sizeof(ifd));
UDP: further performance improvements on tx Cumulative throughput while running 64 netperf -H $DUT -t UDP_STREAM -- -m 1 on a 2x8x2 SKL went from 1.1Mpps to 2.5Mpps Single stream throughput increases from 910kpps to 1.18Mpps Baseline: https://people.freebsd.org/~mmacy/2018.05.11/udpsender2.svg - Protect read access to global ifnet list with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender3.svg - Protect short lived ifaddr references with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender4.svg - Convert if_afdata read lock path to epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender5.svg A fix for the inpcbhash contention is pending sufficient time on a canary at LLNW. Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15409
2018-05-23 21:02:14 +00:00
CK_STAILQ_FOREACH(ifp, &V_ifnet, if_link) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (w->w_arg && w->w_arg != ifp->if_index)
continue;
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
if_data_copy(ifp, &ifd);
- Store pointer to the link-level address right in "struct ifnet" rather than in ifindex_table[]; all (except one) accesses are through ifp anyway. IF_LLADDR() works faster, and all (except one) ifaddr_byindex() users were converted to use ifp->if_addr. - Stop storing a (pointer to) Ethernet address in "struct arpcom", and drop the IFP2ENADDR() macro; all users have been converted to use IF_LLADDR() instead.
2005-11-11 16:04:59 +00:00
ifa = ifp->if_addr;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_IFP] = ifa->ifa_addr;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
error = rtsock_msg_buffer(RTM_IFINFO, &info, w, &len);
if (error != 0)
goto done;
Minor changes to improve code readability (no actual code changes): + replace 0 with NULL where appropriate (not complete) + remove register declaration while there + add argument names to function prototypes to have a better idea of what they are used for + add 'const' qualifiers in 3 places
2004-04-18 00:56:44 +00:00
info.rti_info[RTAX_IFP] = NULL;
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
if (w->w_req && w->w_tmem) {
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
if (w->w_op == NET_RT_IFLISTL)
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
error = sysctl_iflist_ifml(ifp, &ifd, &info, w,
len);
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
else
Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock, as ifnet implementations typically can take the if_addr_lock after their own locks during configuration. Refactor a sysctl handler that was violating this to read if_counter data in a temporary buffer before the if_addr_lock is taken, and then copying the data in its final location later, when the if_addr_lock is held. PR: 194109 Reported by: Jean-Sebastien Pedron MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D8498 Reviewed by: sbruno
2016-11-12 19:03:23 +00:00
error = sysctl_iflist_ifm(ifp, &ifd, &info, w,
len);
Cosmetics: to silence gcc -wall.
1994-10-08 22:38:27 +00:00
if (error)
Change sysctl_iflist() so it has a single point of return. This will assist any future locking efforts.
2001-09-29 05:08:04 +00:00
goto done;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
while ((ifa = CK_STAILQ_NEXT(ifa, ifa_link)) != NULL) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (af && af != ifa->ifa_addr->sa_family)
continue;
Standardize the various prison_foo_ip[46] functions and prison_if to return zero on success and an error code otherwise. The possible errors are EADDRNOTAVAIL if an address being checked for doesn't match the prison, and EAFNOSUPPORT if the prison doesn't have any addresses in that address family. For most callers of these functions, use the returned error code instead of e.g. a hard-coded EADDRNOTAVAIL or EINVAL. Always include a jailed() check in these functions, where a non-jailed cred always returns success (and makes no changes). Remove the explicit jailed() checks that preceded many of the function calls. Approved by: bz (mentor)
2009-02-05 14:06:09 +00:00
if (prison_if(w->w_req->td->td_ucred,
ifa->ifa_addr) != 0)
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
continue;
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_IFA] = ifa->ifa_addr;
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
info.rti_info[RTAX_NETMASK] = rtsock_fix_netmask(
ifa->ifa_addr, ifa->ifa_netmask, &ss);
cleanups prior to adding locking (and in some cases to eliminate locking): o move route_cb to be private to rtsock.c o replace global static route_proto by locals o eliminate global #define shorthands for info references o remove some register decls o ansi-fy function decls o move items to be close in scope to their usage o add rt_dispatch function for dispatching the actual message o cleanup tangled logic for doing all-but-me msg send Support by: FreeBSD Foundation
2003-10-03 18:15:54 +00:00
info.rti_info[RTAX_BRD] = ifa->ifa_dstaddr;
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
error = rtsock_msg_buffer(RTM_NEWADDR, &info, w, &len);
if (error != 0)
goto done;
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
if (w->w_req && w->w_tmem) {
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
if (w->w_op == NET_RT_IFLISTL)
error = sysctl_iflist_ifaml(ifa, &info,
w, len);
else
error = sysctl_iflist_ifam(ifa, &info,
w, len);
Cosmetics: to silence gcc -wall.
1994-10-08 22:38:27 +00:00
if (error)
Change sysctl_iflist() so it has a single point of return. This will assist any future locking efforts.
2001-09-29 05:08:04 +00:00
goto done;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
}
Fix incorrect netmasks being passed via rtsock. Since radix has been ignoring sa_family in passed sockaddrs, no one ever has bothered filling valid sa_family in netmasks. Additionally, radix adjusts sa_len field in every netmask not to compare zero bytes at all. This leads us to rt_mask with sa_family of AF_UNSPEC (-1) and arbitrary sa_len field (0 for default route, for example). However, rtsock have been passing that rt_mask intact for ages, requiring all rtsock consumers to make ther own local hacks. We even have unfixed on in base: do `route -n monitor` in one window and issue `route -n get addr` for some directly-connected address. You will probably see the following: got message of size 304 on Thu May 8 15:06:06 2014 RTM_GET: Report Metrics: len 304, pid: 30493, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 (255) ffff ffff ff em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ after the change: got message of size 312 on Thu May 8 15:44:07 2014 RTM_GET: Report Metrics: len 312, pid: 2895, seq 1, errno 0, flags:<UP,DONE,PINNED> locks: inits: sockaddrs: <DST,GATEWAY,NETMASK,IFP,IFA> 10.0.0.0 link#1 255.255.255.0 em0:8.0.27.c5.29.d4 10.0.0.92 _________________^^^^^^^^^^^^^^^^^^ Sponsored by: Yandex LLC MFC after: 1 month
2014-05-08 11:56:06 +00:00
info.rti_info[RTAX_IFA] = NULL;
info.rti_info[RTAX_NETMASK] = NULL;
info.rti_info[RTAX_BRD] = NULL;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Change sysctl_iflist() so it has a single point of return. This will assist any future locking efforts.
2001-09-29 05:08:04 +00:00
done:
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Take the cred from curthread rather than curproc as curproc would need locking but the credential from curthread (usually) never changes. Discussed with: jhb MFC after: 2 weeks
2009-01-09 16:22:32 +00:00
static int
Consistently use ifaddr_byindex() to access the link-level address of an interface. No functional change. On passing, comment a likely bug in net/rtsock.c:sysctl_ifmalist() which, if confirmed, would deserve to be fixed and MFC'ed
2004-04-16 08:14:34 +00:00
sysctl_ifmalist(int af, struct walkarg *w)
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
{
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
struct rt_addrinfo info;
Consistently use ifaddr_byindex() to access the link-level address of an interface. No functional change. On passing, comment a likely bug in net/rtsock.c:sysctl_ifmalist() which, if confirmed, would deserve to be fixed and MFC'ed
2004-04-16 08:14:34 +00:00
struct ifaddr *ifa;
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
struct ifmultiaddr *ifma;
struct ifnet *ifp;
int error, len;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
NET_EPOCH_ASSERT();
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
error = 0;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
bzero((caddr_t)&info, sizeof(info));
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
UDP: further performance improvements on tx Cumulative throughput while running 64 netperf -H $DUT -t UDP_STREAM -- -m 1 on a 2x8x2 SKL went from 1.1Mpps to 2.5Mpps Single stream throughput increases from 910kpps to 1.18Mpps Baseline: https://people.freebsd.org/~mmacy/2018.05.11/udpsender2.svg - Protect read access to global ifnet list with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender3.svg - Protect short lived ifaddr references with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender4.svg - Convert if_afdata read lock path to epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender5.svg A fix for the inpcbhash contention is pending sufficient time on a canary at LLNW. Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15409
2018-05-23 21:02:14 +00:00
CK_STAILQ_FOREACH(ifp, &V_ifnet, if_link) {
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
if (w->w_arg && w->w_arg != ifp->if_index)
continue;
- Store pointer to the link-level address right in "struct ifnet" rather than in ifindex_table[]; all (except one) accesses are through ifp anyway. IF_LLADDR() works faster, and all (except one) ifaddr_byindex() users were converted to use ifp->if_addr. - Stop storing a (pointer to) Ethernet address in "struct arpcom", and drop the IFP2ENADDR() macro; all users have been converted to use IF_LLADDR() instead.
2005-11-11 16:04:59 +00:00
ifa = ifp->if_addr;
misc cleanup in sysctl_ifmalist(): + remove a partly incorrect comment that i introduced in the last commit; + deal with the correct part of the above comment by cleaning up the updates of 'info' -- rti_addrs needd not to be updated, rti_info[RTAX_IFP] can be set once outside the loop. While at it, correct a few misspelling of NULL as 0, but there are way too many in this file, and i did not want to clutter the important part of this commit.
2004-04-17 15:09:36 +00:00
info.rti_info[RTAX_IFP] = ifa ? ifa->ifa_addr : NULL;
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
if (af && af != ifma->ifma_addr->sa_family)
continue;
Standardize the various prison_foo_ip[46] functions and prison_if to return zero on success and an error code otherwise. The possible errors are EADDRNOTAVAIL if an address being checked for doesn't match the prison, and EAFNOSUPPORT if the prison doesn't have any addresses in that address family. For most callers of these functions, use the returned error code instead of e.g. a hard-coded EADDRNOTAVAIL or EINVAL. Always include a jailed() check in these functions, where a non-jailed cred always returns success (and makes no changes). Remove the explicit jailed() checks that preceded many of the function calls. Approved by: bz (mentor)
2009-02-05 14:06:09 +00:00
if (prison_if(w->w_req->td->td_ucred,
ifma->ifma_addr) != 0)
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
continue;
info.rti_info[RTAX_IFA] = ifma->ifma_addr;
misc cleanup in sysctl_ifmalist(): + remove a partly incorrect comment that i introduced in the last commit; + deal with the correct part of the above comment by cleaning up the updates of 'info' -- rti_addrs needd not to be updated, rti_info[RTAX_IFP] can be set once outside the loop. While at it, correct a few misspelling of NULL as 0, but there are way too many in this file, and i did not want to clutter the important part of this commit.
2004-04-17 15:09:36 +00:00
info.rti_info[RTAX_GATEWAY] =
(ifma->ifma_addr->sa_family != AF_LINK) ?
ifma->ifma_lladdr : NULL;
Fix sysctl_ifmalist() broken in r265019. Reported by: Olivier Cochard-Labbé MFC with: r265019
2014-05-03 17:57:06 +00:00
error = rtsock_msg_buffer(RTM_NEWMADDR, &info, w, &len);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
if (error != 0)
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
break;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
if (w->w_req && w->w_tmem) {
Consistently use ifaddr_byindex() to access the link-level address of an interface. No functional change. On passing, comment a likely bug in net/rtsock.c:sysctl_ifmalist() which, if confirmed, would deserve to be fixed and MFC'ed
2004-04-16 08:14:34 +00:00
struct ifma_msghdr *ifmam;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
ifmam = (struct ifma_msghdr *)w->w_tmem;
ifmam->ifmam_index = ifma->ifma_ifp->if_index;
ifmam->ifmam_flags = 0;
ifmam->ifmam_addrs = info.rti_addrs;
Plug routing sysctl leaks. Various structures exported by sysctl_rtsock() contain padding fields which were not being zeroed. Reported by: Thomas Barabosch, Fraunhofer FKIE Reviewed by: ae MFC after: 3 days Security: kernel memory disclosure Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D18333
2018-11-26 13:42:18 +00:00
ifmam->_ifmam_spare1 = 0;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
error = SYSCTL_OUT(w->w_req, w->w_tmem, len);
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
if (error != 0)
break;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
}
}
Plug a lock leak in sysctl_ifmalist(). Fix style in the local variable declarations. PR: 214542 MFC after: 1 week
2016-11-15 19:23:48 +00:00
if (error != 0)
break;
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
}
return (error);
}
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
static int
Previous commit changing SYSCTL_HANDLER_ARGS violated KNF. Pointed out by: bde
2000-07-04 11:25:35 +00:00
sysctl_rtsock(SYSCTL_HANDLER_ARGS)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Switch RIB and RADIX_NODE_HEAD lock from rwlock(9) to rmlock(9). Using of rwlock with multiqueue NICs for IP forwarding on high pps produces high lock contention and inefficient. Rmlock fits better for such workloads. Reviewed by: melifaro, olivier Obtained from: Yandex LLC Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15789
2018-06-16 08:26:23 +00:00
RIB_RLOCK_TRACKER;
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
struct epoch_tracker et;
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
int *name = (int *)arg1;
u_int namelen = arg2;
MFP r287070,r287073: split radix implementation and route table structure. There are number of radix consumers in kernel land (pf,ipfw,nfs,route) with different requirements. In fact, first 3 don't have _any_ requirements and first 2 does not use radix locking. On the other hand, routing structure do have these requirements (rnh_gen, multipath, custom to-be-added control plane functions, different locking). Additionally, radix should not known anything about its consumers internals. So, radix code now uses tiny 'struct radix_head' structure along with internal 'struct radix_mask_head' instead of 'struct radix_node_head'. Existing consumers still uses the same 'struct radix_node_head' with slight modifications: they need to pass pointer to (embedded) 'struct radix_head' to all radix callbacks. Routing code now uses new 'struct rib_head' with different locking macro: RADIX_NODE_HEAD prefix was renamed to RIB_ (which stands for routing information base). New net/route_var.h header was added to hold routing subsystem internal data. 'struct rib_head' was placed there. 'struct rtentry' will also be moved there soon.
2016-01-25 06:33:15 +00:00
struct rib_head *rnh = NULL; /* silence compiler. */
Protect interface and address lists using the appropriate mutex. These locks were not aquired because the user buffers were not wired, thus it was possible that that SYSCTL_OUT could sleep, causing a number of different problems such as lock ordering issues and dead locks. -Wire user supplied buffer to ensure SYSCTL_OUT will not sleep. -Pickup ifnet locks to protect the list. -Where applicable pickup address locks. -Pickup radix node head locks. -Remove splnet stubs -Remove various comments about locking here, because they are no longer needed. It is the hope that these changes will make sysctl_rtsock MP safe. MFC after: 3 weeks
2005-09-10 15:12:24 +00:00
int i, lim, error = EINVAL;
Add a leaf node CTL_NET.PF_ROUTE.0.AF.NET_RT_DUMP.0.FIB. This returns routing table with the specified FIB number, not td->td_proc->p_fibnum.
2013-07-12 12:36:12 +00:00
int fib = 0;
remove an almost-duplicate piece of code by setting the loop limits appropriately.
2004-04-12 20:26:01 +00:00
u_char af;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct walkarg w;
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
name ++;
namelen--;
if (req->newptr)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (EPERM);
Introduce nexthop objects and new routing KPI. This is the foundational change for the routing subsytem rearchitecture. More details and goals are available in https://reviews.freebsd.org/D24141 . This patch introduces concept of nexthop objects and new nexthop-based routing KPI. Nexthops are objects, containing all necessary information for performing the packet output decision. Output interface, mtu, flags, gw address goes there. For most of the cases, these objects will serve the same role as the struct rtentry is currently serving. Typically there will be low tens of such objects for the router even with multiple BGP full-views, as these objects will be shared between routing entries. This allows to store more information in the nexthop. New KPI: struct nhop_object *fib4_lookup(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, uint32_t flowid); struct nhop_object *fib6_lookup(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, uint32_t flowid); These 2 function are intended to replace all all flavours of <in_|in6_>rtalloc[1]<_ign><_fib>, mpath functions and the previous fib[46]-generation functions. Upon successful lookup, they return nexthop object which is guaranteed to exist within current NET_EPOCH. If longer lifetime is desired, one can specify NHR_REF as a flag and get a referenced version of the nexthop. Reference semantic closely resembles rtentry one, allowing sed-style conversion. Additionally, another 2 functions are introduced to support uRPF functionality inside variety of our firewalls. Their primary goal is to hide the multipath implementation details inside the routing subsystem, greatly simplifying firewalls implementation: int fib4_lookup_urpf(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); int fib6_lookup_urpf(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); All functions have a separate scopeid argument, paving way to eliminating IPv6 scope embedding and allowing to support IPv4 link-locals in the future. Structure changes: * rtentry gets new 'rt_nhop' pointer, slightly growing the overall size. * rib_head gets new 'rnh_preadd' callback pointer, slightly growing overall sz. Old KPI: During the transition state old and new KPI will coexists. As there are another 4-5 decent-sized conversion patches, it will probably take a couple of weeks. To support both KPIs, fields not required by the new KPI (most of rtentry) has to be kept, resulting in the temporary size increase. Once conversion is finished, rtentry will notably shrink. More details: * architectural overview: https://reviews.freebsd.org/D24141 * list of the next changes: https://reviews.freebsd.org/D24232 Reviewed by: ae,glebius(initial version) Differential Revision: https://reviews.freebsd.org/D24232
2020-04-12 14:30:00 +00:00
if (name[1] == NET_RT_DUMP || name[1] == NET_RT_NHOP) {
Add a leaf node CTL_NET.PF_ROUTE.0.AF.NET_RT_DUMP.0.FIB. This returns routing table with the specified FIB number, not td->td_proc->p_fibnum.
2013-07-12 12:36:12 +00:00
if (namelen == 3)
fib = req->td->td_proc->p_fibnum;
else if (namelen == 4)
Constanly use RT_ALL_FIBS everywhere instead of -1. MFC after: 2 weeks
2014-01-08 23:09:02 +00:00
fib = (name[3] == RT_ALL_FIBS) ?
Add a leaf node CTL_NET.PF_ROUTE.0.AF.NET_RT_DUMP.0.FIB. This returns routing table with the specified FIB number, not td->td_proc->p_fibnum.
2013-07-12 12:36:12 +00:00
req->td->td_proc->p_fibnum : name[3];
else
return ((namelen < 3) ? EISDIR : ENOTDIR);
if (fib < 0 || fib >= rt_numfibs)
return (EINVAL);
} else if (namelen != 3)
Have sysctl() return the correct errno(2) as documented in the sysctl(3) manpage. Submitted by: ru Obtained from: BSD/OS
2002-01-25 00:50:49 +00:00
return ((namelen < 3) ? EISDIR : ENOTDIR);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
af = name[0];
Range-check the address family parameter passed in to the sysctl handler. Submitted by: ru
2002-12-25 10:51:20 +00:00
if (af > AF_MAX)
return (EINVAL);
+ replace Bcmp/Bzero with 'the real thing' as in the rest of the file. + remember to check and fix or explain a strange cast in route_output()
2004-04-18 11:47:04 +00:00
bzero(&w, sizeof(w));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
w.w_op = name[1];
w.w_arg = name[2];
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
w.w_req = req;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Protect interface and address lists using the appropriate mutex. These locks were not aquired because the user buffers were not wired, thus it was possible that that SYSCTL_OUT could sleep, causing a number of different problems such as lock ordering issues and dead locks. -Wire user supplied buffer to ensure SYSCTL_OUT will not sleep. -Pickup ifnet locks to protect the list. -Where applicable pickup address locks. -Pickup radix node head locks. -Remove splnet stubs -Remove various comments about locking here, because they are no longer needed. It is the hope that these changes will make sysctl_rtsock MP safe. MFC after: 3 weeks
2005-09-10 15:12:24 +00:00
error = sysctl_wire_old_buffer(req, 0);
if (error)
return (error);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
/*
* Allocate reply buffer in advance.
* All rtsock messages has maximum length of u_short.
*/
w.w_tmemsize = 65536;
w.w_tmem = malloc(w.w_tmemsize, M_TEMP, M_WAITOK);
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
NET_EPOCH_ENTER(et);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
switch (w.w_op) {
case NET_RT_DUMP:
case NET_RT_FLAGS:
remove an almost-duplicate piece of code by setting the loop limits appropriately.
2004-04-12 20:26:01 +00:00
if (af == 0) { /* dump all tables */
i = 1;
lim = AF_MAX;
} else /* dump only one table */
i = lim = af;
This checkin addresses a couple of issues: 1. The "route" command allows route insertion through the interface-direct option "-iface". During if_attach(), an sockaddr_dl{} entry is created for the interface and is part of the interface address list. This sockaddr_dl{} entry describes the interface in detail. The "route" command selects this entry as the "gateway" object when the "-iface" option is present. The "arp" and "ndp" commands also interact with the kernel through the routing socket when adding and removing static L2 entries. The static L2 information is also provided through the "gateway" object with an AF_LINK family type, similar to what is provided by the "route" command. In order to differentiate between these two types of operations, a RTF_LLDATA flag is introduced. This flag is set by the "arp" and "ndp" commands when issuing the add and delete commands. This flag is also set in each L2 entry returned by the kernel. The "arp" and "ndp" command follows a convention where a RTM_GET is issued first followed by a RTM_ADD/DELETE. This RTM_GET request fills in the fields for a "rtm" object, which is reinjected into the kernel by a subsequent RTM_ADD/DELETE command. The entry returend from RTM_GET is a prefix route, so the RTF_LLDATA flag must be specified when issuing the RTM_ADD/DELETE messages. 2. Enforce the convention that NET_RT_FLAGS with a 0 w_arg is the specification for retrieving L2 information. Also optimized the code logic. Reviewed by: julian
2008-12-26 19:45:24 +00:00
/*
* take care of llinfo entries, the caller must
* specify an AF
*/
Revive the RTF_LLINFO flag in route.h. The kernel code is guarded by the new kernel option COMPAT_ROUTE_FLAGS for binary backward compatibility. The RTF_LLDATA flag maps to the same value as RTF_LLINFO. RTF_LLDATA is used by the arp and ndp utilities. The RTF_LLDATA flag is always returned to the userland regardless whether the COMPAT_ROUTE_FLAGS is defined.
2009-01-12 11:24:32 +00:00
if (w.w_op == NET_RT_FLAGS &&
The RTF_LLINFO was revived unconditionally, but within the kernel the check on the sysctl argument value being RTF_LLINFO is conditioned on the COMPAT_ROUTE_FLAGS kernel option. This mismatch caused the L2 table retrieval failure, and the arp/ndp -an command displays empty L2 tables. Reviewed by: pjd
2009-01-16 09:01:45 +00:00
(w.w_arg == 0 || w.w_arg & RTF_LLINFO)) {
This checkin addresses a couple of issues: 1. The "route" command allows route insertion through the interface-direct option "-iface". During if_attach(), an sockaddr_dl{} entry is created for the interface and is part of the interface address list. This sockaddr_dl{} entry describes the interface in detail. The "route" command selects this entry as the "gateway" object when the "-iface" option is present. The "arp" and "ndp" commands also interact with the kernel through the routing socket when adding and removing static L2 entries. The static L2 information is also provided through the "gateway" object with an AF_LINK family type, similar to what is provided by the "route" command. In order to differentiate between these two types of operations, a RTF_LLDATA flag is introduced. This flag is set by the "arp" and "ndp" commands when issuing the add and delete commands. This flag is also set in each L2 entry returned by the kernel. The "arp" and "ndp" command follows a convention where a RTM_GET is issued first followed by a RTM_ADD/DELETE. This RTM_GET request fills in the fields for a "rtm" object, which is reinjected into the kernel by a subsequent RTM_ADD/DELETE command. The entry returend from RTM_GET is a prefix route, so the RTF_LLDATA flag must be specified when issuing the RTM_ADD/DELETE messages. 2. Enforce the convention that NET_RT_FLAGS with a 0 w_arg is the specification for retrieving L2 information. Also optimized the code logic. Reviewed by: julian
2008-12-26 19:45:24 +00:00
if (af != 0)
error = lltable_sysctl_dumparp(af, w.w_req);
else
error = EINVAL;
break;
}
/*
* take care of routing entries
*/
Put multiple instructions into a block when iterating; unbreaks NET_RT_DUMP, which otherwise only returned information of AF_MAX. This was broken in r193232 (save your time - my bug, my fix). PR: kern/137700 Reported by: Larry Baird (lab gta.com) Tested by: Larry Baird (lab gta.com) Reviewed by: zec, lstewart, qing Approved by: re (kib)
2009-08-13 09:29:52 +00:00
for (error = 0; error == 0 && i <= lim; i++) {
Add a leaf node CTL_NET.PF_ROUTE.0.AF.NET_RT_DUMP.0.FIB. This returns routing table with the specified FIB number, not td->td_proc->p_fibnum.
2013-07-12 12:36:12 +00:00
rnh = rt_tables_get_rnh(fib, i);
Convert the two dimensional array to be malloced and introduce an accessor function to get the correct rnh pointer back. Update netstat to get the correct pointer using kvm_read() as well. This not only fixes the ABI problem depending on the kernel option but also permits the tunable to overwrite the kernel option at boot time up to MAXFIBS, enlarging the number of FIBs without having to recompile. So people could just use GENERIC now. Reviewed by: julian, rwatson, zec X-MFC: not possible
2009-06-01 15:49:42 +00:00
if (rnh != NULL) {
MFP r287070,r287073: split radix implementation and route table structure. There are number of radix consumers in kernel land (pf,ipfw,nfs,route) with different requirements. In fact, first 3 don't have _any_ requirements and first 2 does not use radix locking. On the other hand, routing structure do have these requirements (rnh_gen, multipath, custom to-be-added control plane functions, different locking). Additionally, radix should not known anything about its consumers internals. So, radix code now uses tiny 'struct radix_head' structure along with internal 'struct radix_mask_head' instead of 'struct radix_node_head'. Existing consumers still uses the same 'struct radix_node_head' with slight modifications: they need to pass pointer to (embedded) 'struct radix_head' to all radix callbacks. Routing code now uses new 'struct rib_head' with different locking macro: RADIX_NODE_HEAD prefix was renamed to RIB_ (which stands for routing information base). New net/route_var.h header was added to hold routing subsystem internal data. 'struct rib_head' was placed there. 'struct rtentry' will also be moved there soon.
2016-01-25 06:33:15 +00:00
RIB_RLOCK(rnh);
error = rnh->rnh_walktree(&rnh->head,
Protect interface and address lists using the appropriate mutex. These locks were not aquired because the user buffers were not wired, thus it was possible that that SYSCTL_OUT could sleep, causing a number of different problems such as lock ordering issues and dead locks. -Wire user supplied buffer to ensure SYSCTL_OUT will not sleep. -Pickup ifnet locks to protect the list. -Where applicable pickup address locks. -Pickup radix node head locks. -Remove splnet stubs -Remove various comments about locking here, because they are no longer needed. It is the hope that these changes will make sysctl_rtsock MP safe. MFC after: 3 weeks
2005-09-10 15:12:24 +00:00
sysctl_dumpentry, &w);
MFP r287070,r287073: split radix implementation and route table structure. There are number of radix consumers in kernel land (pf,ipfw,nfs,route) with different requirements. In fact, first 3 don't have _any_ requirements and first 2 does not use radix locking. On the other hand, routing structure do have these requirements (rnh_gen, multipath, custom to-be-added control plane functions, different locking). Additionally, radix should not known anything about its consumers internals. So, radix code now uses tiny 'struct radix_head' structure along with internal 'struct radix_mask_head' instead of 'struct radix_node_head'. Existing consumers still uses the same 'struct radix_node_head' with slight modifications: they need to pass pointer to (embedded) 'struct radix_head' to all radix callbacks. Routing code now uses new 'struct rib_head' with different locking macro: RADIX_NODE_HEAD prefix was renamed to RIB_ (which stands for routing information base). New net/route_var.h header was added to hold routing subsystem internal data. 'struct rib_head' was placed there. 'struct rtentry' will also be moved there soon.
2016-01-25 06:33:15 +00:00
RIB_RUNLOCK(rnh);
remove an almost-duplicate piece of code by setting the loop limits appropriately.
2004-04-12 20:26:01 +00:00
} else if (af != 0)
SMP locking for radix nodes.
2002-12-24 03:03:39 +00:00
error = EAFNOSUPPORT;
Put multiple instructions into a block when iterating; unbreaks NET_RT_DUMP, which otherwise only returned information of AF_MAX. This was broken in r193232 (save your time - my bug, my fix). PR: kern/137700 Reported by: Larry Baird (lab gta.com) Tested by: Larry Baird (lab gta.com) Reviewed by: zec, lstewart, qing Approved by: re (kib)
2009-08-13 09:29:52 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
break;
Introduce nexthop objects and new routing KPI. This is the foundational change for the routing subsytem rearchitecture. More details and goals are available in https://reviews.freebsd.org/D24141 . This patch introduces concept of nexthop objects and new nexthop-based routing KPI. Nexthops are objects, containing all necessary information for performing the packet output decision. Output interface, mtu, flags, gw address goes there. For most of the cases, these objects will serve the same role as the struct rtentry is currently serving. Typically there will be low tens of such objects for the router even with multiple BGP full-views, as these objects will be shared between routing entries. This allows to store more information in the nexthop. New KPI: struct nhop_object *fib4_lookup(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, uint32_t flowid); struct nhop_object *fib6_lookup(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, uint32_t flowid); These 2 function are intended to replace all all flavours of <in_|in6_>rtalloc[1]<_ign><_fib>, mpath functions and the previous fib[46]-generation functions. Upon successful lookup, they return nexthop object which is guaranteed to exist within current NET_EPOCH. If longer lifetime is desired, one can specify NHR_REF as a flag and get a referenced version of the nexthop. Reference semantic closely resembles rtentry one, allowing sed-style conversion. Additionally, another 2 functions are introduced to support uRPF functionality inside variety of our firewalls. Their primary goal is to hide the multipath implementation details inside the routing subsystem, greatly simplifying firewalls implementation: int fib4_lookup_urpf(uint32_t fibnum, struct in_addr dst, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); int fib6_lookup_urpf(uint32_t fibnum, const struct in6_addr *dst6, uint32_t scopeid, uint32_t flags, const struct ifnet *src_if); All functions have a separate scopeid argument, paving way to eliminating IPv6 scope embedding and allowing to support IPv4 link-locals in the future. Structure changes: * rtentry gets new 'rt_nhop' pointer, slightly growing the overall size. * rib_head gets new 'rnh_preadd' callback pointer, slightly growing overall sz. Old KPI: During the transition state old and new KPI will coexists. As there are another 4-5 decent-sized conversion patches, it will probably take a couple of weeks. To support both KPIs, fields not required by the new KPI (most of rtentry) has to be kept, resulting in the temporary size increase. Once conversion is finished, rtentry will notably shrink. More details: * architectural overview: https://reviews.freebsd.org/D24141 * list of the next changes: https://reviews.freebsd.org/D24232 Reviewed by: ae,glebius(initial version) Differential Revision: https://reviews.freebsd.org/D24232
2020-04-12 14:30:00 +00:00
case NET_RT_NHOP:
/* Allow dumping one specific af/fib at a time */
if (namelen < 4) {
error = EINVAL;
break;
}
fib = name[3];
if (fib < 0 || fib > rt_numfibs) {
error = EINVAL;
break;
}
rnh = rt_tables_get_rnh(fib, af);
if (rnh == NULL) {
error = EAFNOSUPPORT;
break;
}
if (w.w_op == NET_RT_NHOP)
error = nhops_dump_sysctl(rnh, w.w_req);
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
case NET_RT_IFLIST:
Introduce a new NET_RT_IFLISTL API to query the address list. It works on extended and extensible structs if_msghdrl and ifa_msghdrl. This will allow us to extend both the msghdrl structs and eventually if_data in the future without breaking the ABI. Bump __FreeBSD_version to allow ports to more easily detect the new API. Reviewed by: glebius, brooks MFC after: 3 days
2012-02-11 06:02:16 +00:00
case NET_RT_IFLISTL:
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = sysctl_iflist(af, &w);
Add a sysctl MIB, NET_RT_IFMALIST, to retrieve multicast group memberships in a protocol-independent way. Submitted by: harti
2003-11-14 18:48:15 +00:00
break;
case NET_RT_IFMALIST:
error = sysctl_ifmalist(af, &w);
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Widen NET_EPOCH coverage. When epoch(9) was introduced to network stack, it was basically dropped in place of existing locking, which was mutexes and rwlocks. For the sake of performance mutex covered areas were as small as possible, so became epoch covered areas. However, epoch doesn't introduce any contention, it just delays memory reclaim. So, there is no point to minimise epoch covered areas in sense of performance. Meanwhile entering/exiting epoch also has non-zero CPU usage, so doing this less often is a win. Not the least is also code maintainability. In the new paradigm we can assume that at any stage of processing a packet, we are inside network epoch. This makes coding both input and output path way easier. On output path we already enter epoch quite early - in the ip_output(), in the ip6_output(). This patch does the same for the input path. All ISR processing, network related callouts, other ways of packet injection to the network stack shall be performed in net_epoch. Any leaf function that walks network configuration now asserts epoch. Tricky part is configuration code paths - ioctls, sysctls. They also call into leaf functions, so some need to be changed. This patch would introduce more epoch recursions (see EPOCH_TRACE) than we had before. They will be cleaned up separately, as several of them aren't trivial. Note, that unlike a lock recursion the epoch recursion is safe and just wastes a bit of resources. Reviewed by: gallatin, hselasky, cy, adrian, kristof Differential Revision: https://reviews.freebsd.org/D19111
2019-10-07 22:40:05 +00:00
NET_EPOCH_EXIT(et);
Improve memory allocation model for rt_msg2() rtsock messages: * memory is now allocated as early as possible, without holding locks. * sysctl users are now guaranteed to get a response (M_WAITOK buffer prealloc). * socket users are more likely to use on-stack buffer for replies. * standard kernel malloc/free functions are now used instead of radix wrappers. rt_msg2() has been renamed to rtsock_msg_buffer(). MFC after: 1 month
2014-04-27 17:41:18 +00:00
free(w.w_tmem, M_TEMP);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
sysctl_rtsock() has all necessary locking and doesn't need Giant to run. While here add description.
2019-11-07 19:06:18 +00:00
static SYSCTL_NODE(_net, PF_ROUTE, routetable, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_rtsock, "Return route tables and interface/address lists");
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Definitions of protocols supported in the ROUTE domain.
*/
Forward declaring static variables as extern is invalid ISO-C. Now that GCC can properly handle forward static declarations, do this properly.
2005-09-07 10:06:14 +00:00
static struct domain routedomain; /* or at least forward */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
All net.* sysctl converted now.
1995-11-16 19:00:27 +00:00
static struct protosw routesw[] = {
Use sparse initializers for "struct domain" and "struct protosw", so they are easier to follow for the human being.
2005-11-09 13:29:16 +00:00
{
.pr_type = SOCK_RAW,
.pr_domain = &routedomain,
.pr_flags = PR_ATOMIC|PR_ADDR,
.pr_output = route_output,
.pr_ctlinput = raw_ctlinput,
.pr_init = raw_init,
.pr_usrreqs = &route_usrreqs
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
};
Use sparse initializers for "struct domain" and "struct protosw", so they are easier to follow for the human being.
2005-11-09 13:29:16 +00:00
static struct domain routedomain = {
.dom_family = PF_ROUTE,
.dom_name = "route",
.dom_protosw = routesw,
kernel: use our nitems() macro when it is available through param.h. No functional change, only trivial cases are done in this sweep, Discussed in: freebsd-current
2016-04-19 23:48:27 +00:00
.dom_protoswNPROTOSW = &routesw[nitems(routesw)]
Use sparse initializers for "struct domain" and "struct protosw", so they are easier to follow for the human being.
2005-11-09 13:29:16 +00:00
};
Updated routing-socket code from Berkeley Obtained from: Keith Bostic by way of Paul Traina
1995-05-10 16:48:52 +00:00
Introduce and use a sysinit-based initialization scheme for virtual network stacks, VNET_SYSINIT: - Add VNET_SYSINIT and VNET_SYSUNINIT macros to declare events that will occur each time a network stack is instantiated and destroyed. In the !VIMAGE case, these are simply mapped into regular SYSINIT/SYSUNINIT. For the VIMAGE case, we instead use SYSINIT's to track their order and properties on registration, using them for each vnet when created/ destroyed, or immediately on module load for already-started vnets. - Remove vnet_modinfo mechanism that existed to serve this purpose previously, as well as its dependency scheme: we now just use the SYSINIT ordering scheme. - Implement VNET_DOMAIN_SET() to allow protocol domains to declare that they want init functions to be called for each virtual network stack rather than just once at boot, compiling down to DOMAIN_SET() in the non-VIMAGE case. - Walk all virtualized kernel subsystems and make use of these instead of modinfo or DOMAIN_SET() for init/uninit events. In some cases, convert modular components from using modevent to using sysinit (where appropriate). In some cases, do minor rejuggling of SYSINIT ordering to make room for or better manage events. Portions submitted by: jhb (VNET_SYSINIT), bz (cleanup) Discussed with: jhb, bz, julian, zec Reviewed by: bz Approved by: re (VIMAGE blanket)
2009-07-23 20:46:49 +00:00
VNET_DOMAIN_SET(route);
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
#ifdef DDB
/*
* Unfortunately, RTF_ values are expressed as raw masks rather than powers of
* 2, so we cannot use them as nice C99 initializer indices below.
*/
static const char * const rtf_flag_strings[] = {
"UP",
"GATEWAY",
"HOST",
"REJECT",
"DYNAMIC",
"MODIFIED",
"DONE",
"UNUSED_0x80",
"UNUSED_0x100",
"XRESOLVE",
"LLDATA",
"STATIC",
"BLACKHOLE",
"UNUSED_0x2000",
"PROTO2",
"PROTO1",
"UNUSED_0x10000",
"UNUSED_0x20000",
"PROTO3",
"FIXEDMTU",
"PINNED",
"LOCAL",
"BROADCAST",
"MULTICAST",
/* Big gap. */
[28] = "STICKY",
[30] = "RNH_LOCKED",
[31] = "GWFLAG_COMPAT",
};
static const char * __pure
rt_flag_name(unsigned idx)
{
if (idx >= nitems(rtf_flag_strings))
return ("INVALID_FLAG");
if (rtf_flag_strings[idx] == NULL)
return ("UNKNOWN");
return (rtf_flag_strings[idx]);
}
static void
rt_dumpaddr_ddb(const char *name, const struct sockaddr *sa)
{
char buf[INET6_ADDRSTRLEN], *res;
res = NULL;
if (sa == NULL)
res = "NULL";
else if (sa->sa_family == AF_INET) {
res = inet_ntop(AF_INET,
&((const struct sockaddr_in *)sa)->sin_addr,
buf, sizeof(buf));
} else if (sa->sa_family == AF_INET6) {
res = inet_ntop(AF_INET6,
&((const struct sockaddr_in6 *)sa)->sin6_addr,
buf, sizeof(buf));
} else if (sa->sa_family == AF_LINK) {
res = "on link";
}
if (res != NULL) {
db_printf("%s <%s> ", name, res);
return;
}
db_printf("%s <af:%d> ", name, sa->sa_family);
}
static int
rt_dumpentry_ddb(struct radix_node *rn, void *arg __unused)
{
struct sockaddr_storage ss;
struct rtentry *rt;
int flags, idx;
/* If RNTORT is important, put it in a header. */
rt = (void *)rn;
rt_dumpaddr_ddb("dst", rt_key(rt));
rt_dumpaddr_ddb("gateway", rt->rt_gateway);
rt_dumpaddr_ddb("netmask", rtsock_fix_netmask(rt_key(rt), rt_mask(rt),
&ss));
if (rt->rt_ifp != NULL && (rt->rt_ifp->if_flags & IFF_DYING) == 0) {
rt_dumpaddr_ddb("ifp", rt->rt_ifp->if_addr->ifa_addr);
rt_dumpaddr_ddb("ifa", rt->rt_ifa->ifa_addr);
}
db_printf("flags ");
flags = rt->rt_flags;
if (flags == 0)
db_printf("none");
while ((idx = ffs(flags)) > 0) {
idx--;
if (flags != rt->rt_flags)
db_printf(",");
Appease Clang false-positive Werrors in r352112 Reported by: bcran
2019-09-10 01:56:47 +00:00
db_printf("%s", rt_flag_name(idx));
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
flags &= ~(1ul << idx);
}
db_printf("\n");
return (0);
}
DB_SHOW_COMMAND(routetable, db_show_routetable_cmd)
{
struct rib_head *rnh;
int error, i, lim;
if (have_addr)
i = lim = addr;
else {
i = 1;
lim = AF_MAX;
}
for (; i <= lim; i++) {
rnh = rt_tables_get_rnh(0, i);
if (rnh == NULL) {
if (have_addr) {
db_printf("%s: AF %d not supported?\n",
__func__, i);
break;
}
continue;
}
if (!have_addr && i > 1)
db_printf("\n");
db_printf("Route table for AF %d%s%s%s:\n", i,
(i == AF_INET || i == AF_INET6) ? " (" : "",
(i == AF_INET) ? "INET" : (i == AF_INET6) ? "INET6" : "",
(i == AF_INET || i == AF_INET6) ? ")" : "");
error = rnh->rnh_walktree(&rnh->head, rt_dumpentry_ddb, NULL);
if (error != 0)
db_printf("%s: walktree(%d): %d\n", __func__, i,
error);
}
}
_DB_FUNC(_show, route, db_show_route_cmd, db_show_table, CS_OWN, NULL)
{
char buf[INET6_ADDRSTRLEN], *bp;
const void *dst_addrp;
struct sockaddr *dstp;
struct rtentry *rt;
union {
struct sockaddr_in dest_sin;
struct sockaddr_in6 dest_sin6;
} u;
uint16_t hextets[8];
unsigned i, tets;
int t, af, exp, tokflags;
/*
* Undecoded address family. No double-colon expansion seen yet.
*/
af = -1;
exp = -1;
/* Assume INET6 to start; we can work back if guess was wrong. */
tokflags = DRT_WSPACE | DRT_HEX | DRT_HEXADECIMAL;
/*
* db_command has lexed 'show route' for us.
*/
t = db_read_token_flags(tokflags);
if (t == tWSPACE)
t = db_read_token_flags(tokflags);
/*
* tEOL: Just 'show route' isn't a valid mode.
* tMINUS: It's either '-h' or some invalid option. Regardless, usage.
*/
if (t == tEOL || t == tMINUS)
goto usage;
db_unread_token(t);
tets = nitems(hextets);
/*
* Each loop iteration, we expect to read one octet (v4) or hextet
* (v6), followed by an appropriate field separator ('.' or ':' or
* '::').
*
* At the start of each loop, we're looking for a number (octet or
* hextet).
*
* INET6 addresses have a special case where they may begin with '::'.
*/
for (i = 0; i < tets; i++) {
t = db_read_token_flags(tokflags);
if (t == tCOLONCOLON) {
/* INET6 with leading '::' or invalid. */
if (i != 0) {
db_printf("Parse error: unexpected extra "
"colons.\n");
goto exit;
}
af = AF_INET6;
exp = i;
hextets[i] = 0;
continue;
} else if (t == tNUMBER) {
/*
* Lexer separates out '-' as tMINUS, but make the
* assumption explicit here.
*/
MPASS(db_tok_number >= 0);
if (af == AF_INET && db_tok_number > UINT8_MAX) {
db_printf("Not a valid v4 octet: %ld\n",
Fix build for the platforms where db_expr_t is not long Sponsored by: The FreeBSD Foundation
2019-09-10 08:51:11 +00:00
(long)db_tok_number);
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
goto exit;
}
hextets[i] = db_tok_number;
} else if (t == tEOL) {
/*
* We can only detect the end of an IPv6 address in
* compact representation with EOL.
*/
if (af != AF_INET6 || exp < 0) {
db_printf("Parse failed. Got unexpected EOF "
"when the address is not a compact-"
"representation IPv6 address.\n");
goto exit;
}
break;
} else {
db_printf("Parse failed. Unexpected token %d.\n", t);
goto exit;
}
/* Next, look for a separator, if appropriate. */
if (i == tets - 1)
continue;
t = db_read_token_flags(tokflags);
if (af < 0) {
if (t == tCOLON) {
af = AF_INET6;
continue;
}
if (t == tCOLONCOLON) {
af = AF_INET6;
i++;
hextets[i] = 0;
exp = i;
continue;
}
if (t == tDOT) {
unsigned hn, dn;
af = AF_INET;
/* Need to fixup the first parsed number. */
if (hextets[0] > 0x255 ||
(hextets[0] & 0xf0) > 0x90 ||
(hextets[0] & 0xf) > 9) {
db_printf("Not a valid v4 octet: %x\n",
hextets[0]);
goto exit;
}
hn = hextets[0];
dn = (hn >> 8) * 100 +
((hn >> 4) & 0xf) * 10 +
(hn & 0xf);
hextets[0] = dn;
/* Switch to decimal for remaining octets. */
tokflags &= ~DRT_RADIX_MASK;
tokflags |= DRT_DECIMAL;
tets = 4;
continue;
}
db_printf("Parse error. Unexpected token %d.\n", t);
goto exit;
} else if (af == AF_INET) {
if (t == tDOT)
continue;
db_printf("Expected '.' (%d) between octets but got "
"(%d).\n", tDOT, t);
goto exit;
} else if (af == AF_INET6) {
if (t == tCOLON)
continue;
if (t == tCOLONCOLON) {
if (exp < 0) {
i++;
hextets[i] = 0;
exp = i;
continue;
}
db_printf("Got bogus second '::' in v6 "
"address.\n");
goto exit;
}
if (t == tEOL) {
/*
* Handle in the earlier part of the loop
* because we need to handle trailing :: too.
*/
db_unread_token(t);
continue;
}
db_printf("Expected ':' (%d) or '::' (%d) between "
"hextets but got (%d).\n", tCOLON, tCOLONCOLON, t);
goto exit;
}
}
/* Check for trailing garbage. */
if (i == tets) {
t = db_read_token_flags(tokflags);
if (t != tEOL) {
db_printf("Got unexpected garbage after address "
"(%d).\n", t);
goto exit;
}
}
/*
* Need to expand compact INET6 addresses.
*
* Technically '::' for a single ':0:' is MUST NOT but just in case,
* don't bother expanding that form (exp >= 0 && i == tets case).
*/
if (af == AF_INET6 && exp >= 0 && i < tets) {
if (exp + 1 < i) {
memmove(&hextets[exp + 1 + (nitems(hextets) - i)],
&hextets[exp + 1],
(i - (exp + 1)) * sizeof(hextets[0]));
}
memset(&hextets[exp + 1], 0, (nitems(hextets) - i) *
sizeof(hextets[0]));
}
memset(&u, 0, sizeof(u));
if (af == AF_INET) {
u.dest_sin.sin_family = AF_INET;
u.dest_sin.sin_len = sizeof(u.dest_sin);
u.dest_sin.sin_addr.s_addr = htonl(
((uint32_t)hextets[0] << 24) |
((uint32_t)hextets[1] << 16) |
((uint32_t)hextets[2] << 8) |
(uint32_t)hextets[3]);
dstp = (void *)&u.dest_sin;
dst_addrp = &u.dest_sin.sin_addr;
} else if (af == AF_INET6) {
u.dest_sin6.sin6_family = AF_INET6;
u.dest_sin6.sin6_len = sizeof(u.dest_sin6);
for (i = 0; i < nitems(hextets); i++)
u.dest_sin6.sin6_addr.s6_addr16[i] = htons(hextets[i]);
dstp = (void *)&u.dest_sin6;
dst_addrp = &u.dest_sin6.sin6_addr;
Appease Clang false-positive Werrors in r352112 Reported by: bcran
2019-09-10 01:56:47 +00:00
} else {
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
MPASS(false);
Appease Clang false-positive Werrors in r352112 Reported by: bcran
2019-09-10 01:56:47 +00:00
/* UNREACHABLE */
/* Appease Clang false positive: */
dstp = NULL;
}
ddb(4): Add 'show route <dest>' and 'show routetable [<af>]' These commands show the route resolved for a specified destination, or print out the entire routing table for a given address family (or all families, if none is explicitly provided). Discussed with: emaste Differential Revision: https://reviews.freebsd.org/D21510
2019-09-09 22:54:27 +00:00
bp = inet_ntop(af, dst_addrp, buf, sizeof(buf));
if (bp != NULL)
db_printf("Looking up route to destination '%s'\n", bp);
CURVNET_SET(vnet0);
rt = rtalloc1(dstp, 0, RTF_RNH_LOCKED);
CURVNET_RESTORE();
if (rt == NULL) {
db_printf("Could not get route for that server.\n");
return;
}
rt_dumpentry_ddb((void *)rt, NULL);
RTFREE_LOCKED(rt);
return;
usage:
db_printf("Usage: 'show route <address>'\n"
" Currently accepts only dotted-decimal INET or colon-separated\n"
" hextet INET6 addresses.\n");
exit:
db_skip_to_eol();
}
#endif
Reference in New Issue Copy Permalink