d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
bdd4630c9a
freebsd-skq/sys/netinet/ip_mroute.c

2971 lines
76 KiB
C
Raw Normal View History

Add back missing copyright and license statement. This is identical to the statement in ip_mroute.h, as well as being the same as what OpenBSD has done with this file. It matches the copyright in NetBSD's 1.1 through 1.14 versions of the file as well, which they subsequently added back. It appears to have been lost in the 4.4-lite1 import for FreeBSD 2.0, but where and why I've not investigated further. OpenBSD had the same problem. NetBSD had a copyright notice until Multicast 3.5 was integrated verbatim back in 1995. This appears to be the version that made it into 4.4-lite1. Approved by: re (scottl) MFC after: 3 days
2005-06-23 18:42:58 +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
*
Add back missing copyright and license statement. This is identical to the statement in ip_mroute.h, as well as being the same as what OpenBSD has done with this file. It matches the copyright in NetBSD's 1.1 through 1.14 versions of the file as well, which they subsequently added back. It appears to have been lost in the 4.4-lite1 import for FreeBSD 2.0, but where and why I've not investigated further. OpenBSD had the same problem. NetBSD had a copyright notice until Multicast 3.5 was integrated verbatim back in 1995. This appears to be the version that made it into 4.4-lite1. Approved by: re (scottl) MFC after: 3 days
2005-06-23 18:42:58 +00:00
* Copyright (c) 1989 Stephen Deering
* Copyright (c) 1992, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Stephen Deering of Stanford University.
*
* 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
Add back missing copyright and license statement. This is identical to the statement in ip_mroute.h, as well as being the same as what OpenBSD has done with this file. It matches the copyright in NetBSD's 1.1 through 1.14 versions of the file as well, which they subsequently added back. It appears to have been lost in the 4.4-lite1 import for FreeBSD 2.0, but where and why I've not investigated further. OpenBSD had the same problem. NetBSD had a copyright notice until Multicast 3.5 was integrated verbatim back in 1995. This appears to be the version that made it into 4.4-lite1. Approved by: re (scottl) MFC after: 3 days
2005-06-23 18:42:58 +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.
*
* @(#)ip_mroute.c 8.2 (Berkeley) 11/15/93
*/
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* IP multicast forwarding procedures
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*
* Written by David Waitzman, BBN Labs, August 1988.
* Modified by Steve Deering, Stanford, February 1989.
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Modified by Mark J. Steiglitz, Stanford, May, 1991
* Modified by Van Jacobson, LBL, January 1993
* Modified by Ajit Thyagarajan, PARC, August 1993
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* Modified by Bill Fenner, PARC, April 1995
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* Modified by Ahmed Helmy, SGI, June 1996
* Modified by George Edmond Eddy (Rusty), ISI, February 1998
* Modified by Pavlin Radoslavov, USC/ISI, May 1998, August 1999, October 2000
* Modified by Hitoshi Asaeda, WIDE, August 2000
* Modified by Pavlin Radoslavov, ICSI, October 2002
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* MROUTING Revision: 3.5
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* and PIM-SMv2 and PIM-DM support, advanced API support,
* bandwidth metering and signaling
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/*
* TODO: Prefix functions with ipmf_.
* TODO: Maintain a refcount on if_allmulti() in ifnet or in the protocol
* domain attachment (if_afdata) so we can track consumers of that service.
* TODO: Deprecate routing socket path for SIOCGETSGCNT and SIOCGETVIFCNT,
* move it to socket options.
* TODO: Cleanup LSRR removal further.
* TODO: Push RSVP stubs into raw_ip.c.
* TODO: Use bitstring.h for vif set.
* TODO: Fix mrt6_ioctl dangling ref when dynamically loaded.
* TODO: Sync ip6_mroute.c with this file.
*/
Add FBSDID to all files in netinet so that people can more easily include file version information in bug reports. Approved by: re (kensmith)
2007-10-07 20:44:24 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
#include "opt_inet.h"
Add more options into the conf/options and i386/conf/options.i386 files and the #include hooks so that 'make depend' is more useful. This covers most of the options I regularly use (but not all) and some other easy ones.
1996-03-02 18:24:13 +00:00
#include "opt_mrouting.h"
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
#define _PIM_VT 1
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/kernel.h>
Fix brainos introduced during mechanical KTR change. Pointy hat to: bms
2009-03-20 13:13:50 +00:00
#include <sys/stddef.h>
Remove sys/eventhandler.h from net/route.h Reviewed by: ae
2016-01-09 09:34:39 +00:00
#include <sys/eventhandler.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/lock.h>
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
#include <sys/ktr.h>
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
#include <sys/malloc.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/mbuf.h>
add missing #include <sys/module.h>
2004-05-30 20:27:19 +00:00
#include <sys/module.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/protosw.h>
#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>
Don't include <sys/ioctl.h> in the kernel. Stage 2: include <sys/sockio.h> instead of <sys/ioctl.h> in network files.
1997-03-24 11:33:46 +00:00
#include <sys/sockio.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/sx.h>
#include <sys/sysctl.h>
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
#include <sys/syslog.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/systm.h>
#include <sys/time.h>
Migrate structs arpstat, icmpstat, mrtstat, pimstat and udpstat to PCPU counters.
2013-07-09 09:50:15 +00:00
#include <sys/counter.h>
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
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>
Update netisr handling; Each SWI now registers its queue, and all queue drain routines are done by swi_net, which allows for better queue control at some future point. Packets may also be directly dispatched to a netisr instead of queued, this may be of interest at some installations, but currently defaults to off. Reviewed by: hsu, silby, jayanth, sam Sponsored by: DARPA, NAI Labs
2003-03-04 23:19:55 +00:00
#include <net/netisr.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <net/route.h>
Merge the remainder of kern_vimage.c and vimage.h into vnet.c and vnet.h, we now use jails (rather than vimages) as the abstraction for virtualization management, and what remained was specific to virtual network stacks. Minor cleanups are done in the process, and comments updated to reflect these changes. Reviewed by: bz Approved by: re (vimage blanket)
2009-08-01 19:26:27 +00:00
#include <net/vnet.h>
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <netinet/in.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 <netinet/igmp.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <netinet/in_systm.h>
#include <netinet/in_var.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 <netinet/ip.h>
Somewhat modernize ip_mroute.c: - Use sysctl to export stats - Use ip_encap.c's encapsulation support - Update lkm to kld (is 6 years a record for a broken module?) - Remove some unused cruft
2001-07-25 20:15:49 +00:00
#include <netinet/ip_encap.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <netinet/ip_mroute.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 <netinet/ip_var.h>
Consolidate all IP Options handling functions into ip_options.[ch] and include ip_options.h into all files making use of IP Options functions. From ip_input.c rev 1.306: ip_dooptions(struct mbuf *m, int pass) save_rte(m, option, dst) ip_srcroute(m0) ip_stripoptions(m, mopt) From ip_output.c rev 1.249: ip_insertoptions(m, opt, phlen) ip_optcopy(ip, jp) ip_pcbopts(struct inpcb *inp, int optname, struct mbuf *m) No functional changes in this commit. Discussed with: rwatson Sponsored by: TCP/IP Optimization Fundraise 2005
2005-11-18 20:12:40 +00:00
#include <netinet/ip_options.h>
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
#include <netinet/pim.h>
#include <netinet/pim_var.h>
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
#include <netinet/udp.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 machine/in_cksum.h to unbreak options MROUTING.
2000-05-08 23:56:30 +00:00
#include <machine/in_cksum.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
#ifndef KTR_IPMF
Use KTR_INET for MROUTING CTRs.
2009-04-29 10:17:08 +00:00
#define KTR_IPMF KTR_INET
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
#endif
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
#define VIFI_INVALID ((vifi_t) -1)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
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(uint32_t, last_tv_sec); /* last time we processed this */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_last_tv_sec VNET(last_tv_sec)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static MALLOC_DEFINE(M_MRTABLE, "mroutetbl", "multicast forwarding cache");
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
/*
* Locking. We use two locks: one for the virtual interface table and
* one for the forwarding table. These locks may be nested in which case
* the VIF lock must always be taken first. Note that each lock is used
* to cover not only the specific data structure but also related data
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* structures.
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
*/
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
static struct mtx mrouter_mtx;
#define MROUTER_LOCK() mtx_lock(&mrouter_mtx)
#define MROUTER_UNLOCK() mtx_unlock(&mrouter_mtx)
Remove the now-unused NET_{LOCK,UNLOCK,ASSERT}_GIANT() macros, which previously conditionally acquired Giant based on debug.mpsafenet. As that has now been removed, they are no longer required. Removing them significantly simplifies error-handling in the socket layer, eliminated quite a bit of unwinding of locking in error cases. While here clean up the now unneeded opt_net.h, which previously was used for the NET_WITH_GIANT kernel option. Clean up some related gotos for consistency. Reviewed by: bz, csjp Tested by: kris Approved by: re (kensmith)
2007-08-06 14:26:03 +00:00
#define MROUTER_LOCK_ASSERT() mtx_assert(&mrouter_mtx, MA_OWNED)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define MROUTER_LOCK_INIT() \
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
mtx_init(&mrouter_mtx, "IPv4 multicast forwarding", NULL, MTX_DEF)
#define MROUTER_LOCK_DESTROY() mtx_destroy(&mrouter_mtx)
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
static int ip_mrouter_cnt; /* # of vnets with active mrouters */
static int ip_mrouter_unloading; /* Allow no more V_ip_mrouter sockets */
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_PCPUSTAT_DEFINE_STATIC(struct mrtstat, mrtstat);
Migrate structs arpstat, icmpstat, mrtstat, pimstat and udpstat to PCPU counters.
2013-07-09 09:50:15 +00:00
VNET_PCPUSTAT_SYSINIT(mrtstat);
VNET_PCPUSTAT_SYSUNINIT(mrtstat);
SYSCTL_VNET_PCPUSTAT(_net_inet_ip, OID_AUTO, mrtstat, struct mrtstat,
mrtstat, "IPv4 Multicast Forwarding Statistics (struct mrtstat, "
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
"netinet/ip_mroute.h)");
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(u_long, mfchash);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_mfchash VNET(mfchash)
#define MFCHASH(a, g) \
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
((((a).s_addr >> 20) ^ ((a).s_addr >> 10) ^ (a).s_addr ^ \
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
((g).s_addr >> 20) ^ ((g).s_addr >> 10) ^ (g).s_addr) & V_mfchash)
#define MFCHASHSIZE 256
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
static u_long mfchashsize; /* Hash size */
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(u_char *, nexpire); /* 0..mfchashsize-1 */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_nexpire VNET(nexpire)
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(LIST_HEAD(mfchashhdr, mfc)*, mfchashtbl);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_mfchashtbl VNET(mfchashtbl)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
static struct mtx mfc_mtx;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define MFC_LOCK() mtx_lock(&mfc_mtx)
#define MFC_UNLOCK() mtx_unlock(&mfc_mtx)
Remove the now-unused NET_{LOCK,UNLOCK,ASSERT}_GIANT() macros, which previously conditionally acquired Giant based on debug.mpsafenet. As that has now been removed, they are no longer required. Removing them significantly simplifies error-handling in the socket layer, eliminated quite a bit of unwinding of locking in error cases. While here clean up the now unneeded opt_net.h, which previously was used for the NET_WITH_GIANT kernel option. Clean up some related gotos for consistency. Reviewed by: bz, csjp Tested by: kris Approved by: re (kensmith)
2007-08-06 14:26:03 +00:00
#define MFC_LOCK_ASSERT() mtx_assert(&mfc_mtx, MA_OWNED)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define MFC_LOCK_INIT() \
mtx_init(&mfc_mtx, "IPv4 multicast forwarding cache", NULL, MTX_DEF)
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
#define MFC_LOCK_DESTROY() mtx_destroy(&mfc_mtx)
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(vifi_t, numvifs);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_numvifs VNET(numvifs)
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
VNET_DEFINE_STATIC(struct vif *, viftable);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_viftable VNET(viftable)
* makes mfc[MFCTBLSIZ] and vif[MAXVIFS] tables accessible via sysctl: - sysctlbyname("net.inet.ip.mfctable", ...) - sysctlbyname("net.inet.ip.viftable", ...) This change is needed so netstat can use sysctlbyname() to read the data from those tables. Otherwise, in some cases "netstat -g" may fail to report the multicast forwarding information (e.g., if we run a multicast router on PicoBSD). * Bug fix: when sending IGMPMSG_WRONGVIF upcall to the multicast routing daemon, set properly "im->im_vif" to the receiving incoming interface of the packet that triggered that upcall rather than to the expected incoming interface of that packet. * Bug fix: add missing increment of counter "mrtstat.mrts_upcalls" * Few formatting nits (e.g., replace extra spaces with TABs) Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-05 17:01:33 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
static struct mtx vif_mtx;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define VIF_LOCK() mtx_lock(&vif_mtx)
#define VIF_UNLOCK() mtx_unlock(&vif_mtx)
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
#define VIF_LOCK_ASSERT() mtx_assert(&vif_mtx, MA_OWNED)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define VIF_LOCK_INIT() \
mtx_init(&vif_mtx, "IPv4 multicast interfaces", NULL, MTX_DEF)
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
#define VIF_LOCK_DESTROY() mtx_destroy(&vif_mtx)
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
static eventhandler_tag if_detach_event_tag = NULL;
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(struct callout, expire_upcalls_ch);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_expire_upcalls_ch VNET(expire_upcalls_ch)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
#define EXPIRE_TIMEOUT (hz / 4) /* 4x / second */
#define UPCALL_EXPIRE 6 /* number of timeouts */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Bandwidth meter variables and constants
*/
static MALLOC_DEFINE(M_BWMETER, "bwmeter", "multicast upcall bw meters");
/*
* Pending timeouts are stored in a hash table, the key being the
* expiration time. Periodically, the entries are analysed and processed.
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define BW_METER_BUCKETS 1024
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
VNET_DEFINE_STATIC(struct bw_meter **, bw_meter_timers);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_bw_meter_timers VNET(bw_meter_timers)
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(struct callout, bw_meter_ch);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_bw_meter_ch VNET(bw_meter_ch)
#define BW_METER_PERIOD (hz) /* periodical handling of bw meters */
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Pending upcalls are stored in a vector which is flushed when
* full, or periodically
*/
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
VNET_DEFINE_STATIC(struct bw_upcall *, bw_upcalls);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_bw_upcalls VNET(bw_upcalls)
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(u_int, bw_upcalls_n); /* # of pending upcalls */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_bw_upcalls_n VNET(bw_upcalls_n)
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(struct callout, bw_upcalls_ch);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_bw_upcalls_ch VNET(bw_upcalls_ch)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
#define BW_UPCALLS_PERIOD (hz) /* periodical flush of bw upcalls */
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_PCPUSTAT_DEFINE_STATIC(struct pimstat, pimstat);
Migrate structs arpstat, icmpstat, mrtstat, pimstat and udpstat to PCPU counters.
2013-07-09 09:50:15 +00:00
VNET_PCPUSTAT_SYSINIT(pimstat);
VNET_PCPUSTAT_SYSUNINIT(pimstat);
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
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_NODE(_net_inet, IPPROTO_PIM, pim, CTLFLAG_RW | CTLFLAG_MPSAFE, 0,
"PIM");
Migrate structs arpstat, icmpstat, mrtstat, pimstat and udpstat to PCPU counters.
2013-07-09 09:50:15 +00:00
SYSCTL_VNET_PCPUSTAT(_net_inet_pim, PIMCTL_STATS, stats, struct pimstat,
pimstat, "PIM Statistics (struct pimstat, netinet/pim_var.h)");
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
If the rendezvous point for a group is not specified, do not send IGMPMSG_WHOLEPKT notifications to the userland PIM routing daemon, as an optimization to mitigate the effects of high multicast forwarding load. This is an experimental change, therefore it must be explicitly enabled by setting the sysctl/tunable net.inet.pim.squelch_wholepkt to a non-zero value. The tunable may be set from the loader or from within the kernel environment when loading ip_mroute.ko as a module. Submitted by: edrt <edrt at citiz.net> See also: http://mailman.icsi.berkeley.edu/pipermail/xorp-users/2005-June/000639.html
2007-02-10 14:48:42 +00:00
static u_long pim_squelch_wholepkt = 0;
SYSCTL_ULONG(_net_inet_pim, OID_AUTO, squelch_wholepkt, CTLFLAG_RW,
&pim_squelch_wholepkt, 0,
"Disable IGMP_WHOLEPKT notifications if rendezvous point is unspecified");
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
static const struct encaptab *pim_encap_cookie;
static int pim_encapcheck(const struct mbuf *, int, int, void *);
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
static int pim_input(struct mbuf *, int, int, void *);
static const struct encap_config ipv4_encap_cfg = {
.proto = IPPROTO_PIM,
.min_length = sizeof(struct ip) + PIM_MINLEN,
.exact_match = 8,
.check = pim_encapcheck,
.input = pim_input
};
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Note: the PIM Register encapsulation adds the following in front of a
* data packet:
*
* struct pim_encap_hdr {
* struct ip ip;
* struct pim_encap_pimhdr pim;
* }
*
*/
struct pim_encap_pimhdr {
struct pim pim;
uint32_t flags;
};
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define PIM_ENCAP_TTL 64
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
static struct ip pim_encap_iphdr = {
#if BYTE_ORDER == LITTLE_ENDIAN
sizeof(struct ip) >> 2,
IPVERSION,
#else
IPVERSION,
sizeof(struct ip) >> 2,
#endif
0, /* tos */
sizeof(struct ip), /* total length */
0, /* id */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
0, /* frag offset */
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
PIM_ENCAP_TTL,
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
IPPROTO_PIM,
0, /* checksum */
};
static struct pim_encap_pimhdr pim_encap_pimhdr = {
{
PIM_MAKE_VT(PIM_VERSION, PIM_REGISTER), /* PIM vers and message type */
0, /* reserved */
0, /* checksum */
},
0 /* flags */
};
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(vifi_t, reg_vif_num) = VIFI_INVALID;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_reg_vif_num VNET(reg_vif_num)
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(struct ifnet, multicast_register_if);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_multicast_register_if VNET(multicast_register_if)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Private variables.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static u_long X_ip_mcast_src(int);
static int X_ip_mforward(struct ip *, struct ifnet *, struct mbuf *,
struct ip_moptions *);
Remove __P.
2002-03-19 21:25:46 +00:00
static int X_ip_mrouter_done(void);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static int X_ip_mrouter_get(struct socket *, struct sockopt *);
static int X_ip_mrouter_set(struct socket *, struct sockopt *);
static int X_legal_vif_num(int);
Switch cmd argument to u_long. This matches what if_ethersubr.c does and allows the code to compile cleanly on amd64 with clang. Reviewed by: rwatson Approved by: ed (mentor)
2009-06-21 10:29:31 +00:00
static int X_mrt_ioctl(u_long, caddr_t, int);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static int add_bw_upcall(struct bw_upcall *);
static int add_mfc(struct mfcctl2 *);
static int add_vif(struct vifctl *);
static void bw_meter_prepare_upcall(struct bw_meter *, struct timeval *);
static void bw_meter_process(void);
static void bw_meter_receive_packet(struct bw_meter *, int,
struct timeval *);
static void bw_upcalls_send(void);
static int del_bw_upcall(struct bw_upcall *);
static int del_mfc(struct mfcctl2 *);
static int del_vif(vifi_t);
static int del_vif_locked(vifi_t);
static void expire_bw_meter_process(void *);
static void expire_bw_upcalls_send(void *);
static void expire_mfc(struct mfc *);
static void expire_upcalls(void *);
static void free_bw_list(struct bw_meter *);
static int get_sg_cnt(struct sioc_sg_req *);
static int get_vif_cnt(struct sioc_vif_req *);
static void if_detached_event(void *, struct ifnet *);
static int ip_mdq(struct mbuf *, struct ifnet *, struct mfc *, vifi_t);
static int ip_mrouter_init(struct socket *, int);
static __inline struct mfc *
mfc_find(struct in_addr *, struct in_addr *);
static void phyint_send(struct ip *, struct vif *, struct mbuf *);
static struct mbuf *
pim_register_prepare(struct ip *, struct mbuf *);
static int pim_register_send(struct ip *, struct vif *,
struct mbuf *, struct mfc *);
static int pim_register_send_rp(struct ip *, struct vif *,
struct mbuf *, struct mfc *);
static int pim_register_send_upcall(struct ip *, struct vif *,
struct mbuf *, struct mfc *);
static void schedule_bw_meter(struct bw_meter *, struct timeval *);
static void send_packet(struct vif *, struct mbuf *);
static int set_api_config(uint32_t *);
static int set_assert(int);
static int socket_send(struct socket *, struct mbuf *,
struct sockaddr_in *);
static void unschedule_bw_meter(struct bw_meter *);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Kernel multicast forwarding API capabilities and setup.
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* If more API capabilities are added to the kernel, they should be
* recorded in `mrt_api_support'.
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
#define MRT_API_VERSION 0x0305
static const int mrt_api_version = MRT_API_VERSION;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
static const uint32_t mrt_api_support = (MRT_MFC_FLAGS_DISABLE_WRONGVIF |
MRT_MFC_FLAGS_BORDER_VIF |
MRT_MFC_RP |
MRT_MFC_BW_UPCALL);
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(uint32_t, mrt_api_config);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_mrt_api_config VNET(mrt_api_config)
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(int, pim_assert_enabled);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
#define V_pim_assert_enabled VNET(pim_assert_enabled)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static struct timeval pim_assert_interval = { 3, 0 }; /* Rate limit */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Find a route for a given origin IP address and multicast group address.
* Statistics must be updated by the caller.
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static __inline struct mfc *
mfc_find(struct in_addr *o, struct in_addr *g)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
{
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
struct mfc *rt;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
MFC_LOCK_ASSERT();
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_FOREACH(rt, &V_mfchashtbl[MFCHASH(*o, *g)], mfc_hash) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_hosteq(rt->mfc_origin, *o) &&
in_hosteq(rt->mfc_mcastgrp, *g) &&
TAILQ_EMPTY(&rt->mfc_stall))
break;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (rt);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Handle MRT setsockopt commands to modify the multicast forwarding tables.
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
X_ip_mrouter_set(struct socket *so, struct sockopt *sopt)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
int error, optval;
vifi_t vifi;
struct vifctl vifc;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
struct mfcctl2 mfc;
struct bw_upcall bw_upcall;
uint32_t i;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (so != V_ip_mrouter && sopt->sopt_name != MRT_INIT)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EPERM;
error = 0;
switch (sopt->sopt_name) {
case MRT_INIT:
error = sooptcopyin(sopt, &optval, sizeof optval, sizeof optval);
if (error)
break;
error = ip_mrouter_init(so, optval);
break;
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MRT_DONE:
error = ip_mrouter_done();
break;
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MRT_ADD_VIF:
error = sooptcopyin(sopt, &vifc, sizeof vifc, sizeof vifc);
if (error)
break;
error = add_vif(&vifc);
break;
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MRT_DEL_VIF:
error = sooptcopyin(sopt, &vifi, sizeof vifi, sizeof vifi);
if (error)
break;
error = del_vif(vifi);
break;
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MRT_ADD_MFC:
case MRT_DEL_MFC:
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* select data size depending on API version.
*/
if (sopt->sopt_name == MRT_ADD_MFC &&
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_mrt_api_config & MRT_API_FLAGS_ALL) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
error = sooptcopyin(sopt, &mfc, sizeof(struct mfcctl2),
sizeof(struct mfcctl2));
} else {
error = sooptcopyin(sopt, &mfc, sizeof(struct mfcctl),
sizeof(struct mfcctl));
bzero((caddr_t)&mfc + sizeof(struct mfcctl),
sizeof(mfc) - sizeof(struct mfcctl));
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (error)
break;
if (sopt->sopt_name == MRT_ADD_MFC)
error = add_mfc(&mfc);
else
error = del_mfc(&mfc);
break;
case MRT_ASSERT:
error = sooptcopyin(sopt, &optval, sizeof optval, sizeof optval);
if (error)
break;
set_assert(optval);
break;
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
case MRT_API_CONFIG:
error = sooptcopyin(sopt, &i, sizeof i, sizeof i);
if (!error)
error = set_api_config(&i);
if (!error)
error = sooptcopyout(sopt, &i, sizeof i);
break;
case MRT_ADD_BW_UPCALL:
case MRT_DEL_BW_UPCALL:
error = sooptcopyin(sopt, &bw_upcall, sizeof bw_upcall,
sizeof bw_upcall);
if (error)
break;
if (sopt->sopt_name == MRT_ADD_BW_UPCALL)
error = add_bw_upcall(&bw_upcall);
else
error = del_bw_upcall(&bw_upcall);
break;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
default:
error = EOPNOTSUPP;
break;
}
return error;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
* Handle MRT getsockopt commands
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
X_ip_mrouter_get(struct socket *so, struct sockopt *sopt)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
int error;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
switch (sopt->sopt_name) {
case MRT_VERSION:
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
error = sooptcopyout(sopt, &mrt_api_version, sizeof mrt_api_version);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
break;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MRT_ASSERT:
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
error = sooptcopyout(sopt, &V_pim_assert_enabled,
sizeof V_pim_assert_enabled);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
break;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
case MRT_API_SUPPORT:
error = sooptcopyout(sopt, &mrt_api_support, sizeof mrt_api_support);
break;
case MRT_API_CONFIG:
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
error = sooptcopyout(sopt, &V_mrt_api_config, sizeof V_mrt_api_config);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
break;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
default:
error = EOPNOTSUPP;
break;
}
return error;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Handle ioctl commands to obtain information from the cache
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Switch cmd argument to u_long. This matches what if_ethersubr.c does and allows the code to compile cleanly on amd64 with clang. Reviewed by: rwatson Approved by: ed (mentor)
2009-06-21 10:29:31 +00:00
X_mrt_ioctl(u_long cmd, caddr_t data, int fibnum __unused)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
int error = 0;
When a prison is given the ability to create raw sockets (when the security.jail.allow_raw_sockets sysctl MIB is set to 1) where privileged access to jails is given out, it is possible for prison root to manipulate various network parameters which effect the host environment. This commit plugs a number of security holes associated with the use of raw sockets and prisons. This commit makes the following changes: - Add a comment to rtioctl warning developers that if they add any ioctl commands, they should use super-user checks where necessary, as it is possible for PRISON root to make it this far in execution. - Add super-user checks for the execution of the SIOCGETVIFCNT and SIOCGETSGCNT IP multicast ioctl commands. - Add a super-user check to rip_ctloutput(). If the calling cred is PRISON root, make sure the socket option name is IP_HDRINCL, otherwise deny the request. Although this patch corrects a number of security problems associated with raw sockets and prisons, the warning in jail(8) should still apply, and by default we should keep the default value of security.jail.allow_raw_sockets MIB to 0 (or disabled) until we are certain that we have tracked down all the problems. Looking forward, we will probably want to eliminate the references to curthread. This may be a MFC candidate for RELENG_5. Reviewed by: rwatson Approved by: bmilekic (mentor)
2004-08-21 17:38:57 +00:00
/*
Remove now-unused wrappers for various routing functions.
2016-01-14 08:54:44 +00:00
* Currently the only function calling this ioctl routine is rtioctl_fib().
When a prison is given the ability to create raw sockets (when the security.jail.allow_raw_sockets sysctl MIB is set to 1) where privileged access to jails is given out, it is possible for prison root to manipulate various network parameters which effect the host environment. This commit plugs a number of security holes associated with the use of raw sockets and prisons. This commit makes the following changes: - Add a comment to rtioctl warning developers that if they add any ioctl commands, they should use super-user checks where necessary, as it is possible for PRISON root to make it this far in execution. - Add super-user checks for the execution of the SIOCGETVIFCNT and SIOCGETSGCNT IP multicast ioctl commands. - Add a super-user check to rip_ctloutput(). If the calling cred is PRISON root, make sure the socket option name is IP_HDRINCL, otherwise deny the request. Although this patch corrects a number of security problems associated with raw sockets and prisons, the warning in jail(8) should still apply, and by default we should keep the default value of security.jail.allow_raw_sockets MIB to 0 (or disabled) until we are certain that we have tracked down all the problems. Looking forward, we will probably want to eliminate the references to curthread. This may be a MFC candidate for RELENG_5. Reviewed by: rwatson Approved by: bmilekic (mentor)
2004-08-21 17:38:57 +00:00
* Typically, only root can create the raw socket in order to execute
* this ioctl method, however the request might be coming from a prison
*/
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
error = priv_check(curthread, PRIV_NETINET_MROUTE);
When a prison is given the ability to create raw sockets (when the security.jail.allow_raw_sockets sysctl MIB is set to 1) where privileged access to jails is given out, it is possible for prison root to manipulate various network parameters which effect the host environment. This commit plugs a number of security holes associated with the use of raw sockets and prisons. This commit makes the following changes: - Add a comment to rtioctl warning developers that if they add any ioctl commands, they should use super-user checks where necessary, as it is possible for PRISON root to make it this far in execution. - Add super-user checks for the execution of the SIOCGETVIFCNT and SIOCGETSGCNT IP multicast ioctl commands. - Add a super-user check to rip_ctloutput(). If the calling cred is PRISON root, make sure the socket option name is IP_HDRINCL, otherwise deny the request. Although this patch corrects a number of security problems associated with raw sockets and prisons, the warning in jail(8) should still apply, and by default we should keep the default value of security.jail.allow_raw_sockets MIB to 0 (or disabled) until we are certain that we have tracked down all the problems. Looking forward, we will probably want to eliminate the references to curthread. This may be a MFC candidate for RELENG_5. Reviewed by: rwatson Approved by: bmilekic (mentor)
2004-08-21 17:38:57 +00:00
if (error)
return (error);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
switch (cmd) {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case (SIOCGETVIFCNT):
error = get_vif_cnt((struct sioc_vif_req *)data);
break;
case (SIOCGETSGCNT):
error = get_sg_cnt((struct sioc_sg_req *)data);
break;
default:
error = EINVAL;
break;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
return error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* returns the packet, byte, rpf-failure count for the source group provided
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
get_sg_cnt(struct sioc_sg_req *req)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct mfc *rt;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt = mfc_find(&req->src, &req->grp);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (rt == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
req->pktcnt = req->bytecnt = req->wrong_if = 0xffffffff;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EADDRNOTAVAIL;
}
req->pktcnt = rt->mfc_pkt_cnt;
req->bytecnt = rt->mfc_byte_cnt;
req->wrong_if = rt->mfc_wrong_if;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* returns the input and output packet and byte counts on the vif provided
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
*/
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
get_vif_cnt(struct sioc_vif_req *req)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
vifi_t vifi = req->vifi;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vifi >= V_numvifs) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EINVAL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
req->icount = V_viftable[vifi].v_pkt_in;
req->ocount = V_viftable[vifi].v_pkt_out;
req->ibytes = V_viftable[vifi].v_bytes_in;
req->obytes = V_viftable[vifi].v_bytes_out;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
static void
if_detached_event(void *arg __unused, struct ifnet *ifp)
{
vifi_t vifi;
Use an unsigned long when indexing into mfchashtbl[] and mf6ctable[]. This matches the types used when computing hash indices and the type of the maximum size of mfchashtbl[]. PR: kern/181821 Submitted by: Sven-Thorsten Dietrich <sven@vyatta.com> (IPv4) MFC after: 1 week
2013-09-05 14:16:37 +00:00
u_long i;
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (V_ip_mrouter == NULL) {
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return;
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
VIF_LOCK();
MFC_LOCK();
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
/*
* Tear down multicast forwarder state associated with this ifnet.
* 1. Walk the vif list, matching vifs against this ifnet.
* 2. Walk the multicast forwarding cache (mfc) looking for
* inner matches with this vif's index.
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* 3. Expire any matching multicast forwarding cache entries.
* 4. Free vif state. This should disable ALLMULTI on the interface.
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = 0; vifi < V_numvifs; vifi++) {
if (V_viftable[vifi].v_ifp != ifp)
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
continue;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
for (i = 0; i < mfchashsize; i++) {
struct mfc *rt, *nrt;
Use LIST_FOREACH_SAFE() instead of doing it by hand.
2013-09-05 14:26:37 +00:00
LIST_FOREACH_SAFE(rt, &V_mfchashtbl[i], mfc_hash, nrt) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (rt->mfc_parent == vifi) {
expire_mfc(rt);
}
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
}
}
del_vif_locked(vifi);
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
}
White space cleanup -- remove trailing tab's or spaces from any line. Sponsored by: Netflix Inc.
2020-02-12 13:31:36 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Enable multicast forwarding.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_mrouter_init(struct socket *so, int version)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Step 1.5 of importing the network stack virtualization infrastructure from the vimage project, as per plan established at devsummit 08/08: http://wiki.freebsd.org/Image/Notes200808DevSummit Introduce INIT_VNET_*() initializer macros, VNET_FOREACH() iterator macros, and CURVNET_SET() context setting macros, all currently resolving to NOPs. Prepare for virtualization of selected SYSCTL objects by introducing a family of SYSCTL_V_*() macros, currently resolving to their global counterparts, i.e. SYSCTL_V_INT() == SYSCTL_INT(). Move selected #defines from sys/sys/vimage.h to newly introduced header files specific to virtualized subsystems (sys/net/vnet.h, sys/netinet/vinet.h etc.). All the changes are verified to have zero functional impact at this point in time by doing MD5 comparision between pre- and post-change object files(*). (*) netipsec/keysock.c did not validate depending on compile time options. Implemented by: julian, bz, brooks, zec Reviewed by: julian, bz, brooks, kris, rwatson, ... Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-10-02 15:37:58 +00:00
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR3(KTR_IPMF, "%s: so_type %d, pr_protocol %d", __func__,
so->so_type, so->so_proto->pr_protocol);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (so->so_type != SOCK_RAW || so->so_proto->pr_protocol != IPPROTO_IGMP)
return EOPNOTSUPP;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
if (version != 1)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return ENOPROTOOPT;
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK();
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (ip_mrouter_unloading) {
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
return ENOPROTOOPT;
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_ip_mrouter != NULL) {
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
return EADDRINUSE;
Unlock a mutex which should be unlocked before returning. MFC after: 1 week
2007-02-25 14:22:03 +00:00
}
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_mfchashtbl = hashinit_flags(mfchashsize, M_MRTABLE, &V_mfchash,
HASH_NOWAIT);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
callout_reset(&V_expire_upcalls_ch, EXPIRE_TIMEOUT, expire_upcalls,
curvnet);
callout_reset(&V_bw_upcalls_ch, BW_UPCALLS_PERIOD, expire_bw_upcalls_send,
curvnet);
callout_reset(&V_bw_meter_ch, BW_METER_PERIOD, expire_bw_meter_process,
curvnet);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
V_ip_mrouter = so;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
ip_mrouter_cnt++;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: done", __func__);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Disable multicast forwarding.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
X_ip_mrouter_done(void)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
struct ifnet *ifp;
Use an unsigned long when indexing into mfchashtbl[] and mf6ctable[]. This matches the types used when computing hash indices and the type of the maximum size of mfchashtbl[]. PR: kern/181821 Submitted by: Sven-Thorsten Dietrich <sven@vyatta.com> (IPv4) MFC after: 1 week
2013-09-05 14:16:37 +00:00
u_long i;
Remove unused code and sort variables declarations. PR: kern/181822 MFC after: 1 week
2013-09-05 08:12:36 +00:00
vifi_t vifi;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK();
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (V_ip_mrouter == NULL) {
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
return EINVAL;
}
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
/*
* Detach/disable hooks to the reset of the system.
*/
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
V_ip_mrouter = NULL;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
ip_mrouter_cnt--;
V_mrt_api_config = 0;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
Potential fix for races shutting down callouts when unloading the module. Previously we grabbed the mutex used by the callouts, then stopped the callout with callout_stop, but if the callout was already active and blocked by the mutex then it would continue later and reference the mutex after it was destroyed. Instead stop the callout first then lock. Supported by: FreeBSD Foundation
2003-10-29 19:15:00 +00:00
VIF_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* For each phyint in use, disable promiscuous reception of all IP
* multicasts.
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = 0; vifi < V_numvifs; vifi++) {
if (!in_nullhost(V_viftable[vifi].v_lcl_addr) &&
!(V_viftable[vifi].v_flags & (VIFF_TUNNEL | VIFF_REGISTER))) {
ifp = V_viftable[vifi].v_ifp;
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_allmulti(ifp, 0);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
}
Rather than zeroing MAXVIFS times size of pointer [r362289] (still better than sizeof pointer before [r354857]), we need to zero MAXVIFS times the size of the struct. All good things come in threes; I hope this is it on this one. PR: 246629, 206583 Reported by: kib MFC after: ASAP
2020-06-21 22:09:30 +00:00
bzero((caddr_t)V_viftable, sizeof(*V_viftable) * MAXVIFS);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_numvifs = 0;
V_pim_assert_enabled = 0;
White space cleanup -- remove trailing tab's or spaces from any line. Sponsored by: Netflix Inc.
2020-02-12 13:31:36 +00:00
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
VIF_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
callout_stop(&V_expire_upcalls_ch);
callout_stop(&V_bw_upcalls_ch);
callout_stop(&V_bw_meter_ch);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
Potential fix for races shutting down callouts when unloading the module. Previously we grabbed the mutex used by the callouts, then stopped the callout with callout_stop, but if the callout was already active and blocked by the mutex then it would continue later and reference the mutex after it was destroyed. Instead stop the callout first then lock. Supported by: FreeBSD Foundation
2003-10-29 19:15:00 +00:00
MFC_LOCK();
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/*
* Free all multicast forwarding cache entries.
* Do not use hashdestroy(), as we must perform other cleanup.
*/
for (i = 0; i < mfchashsize; i++) {
struct mfc *rt, *nrt;
Use LIST_FOREACH_SAFE() instead of doing it by hand.
2013-09-05 14:26:37 +00:00
LIST_FOREACH_SAFE(rt, &V_mfchashtbl[i], mfc_hash, nrt) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
expire_mfc(rt);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
free(V_mfchashtbl, M_MRTABLE);
V_mfchashtbl = NULL;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
bzero(V_nexpire, sizeof(V_nexpire[0]) * mfchashsize);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_bw_upcalls_n = 0;
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
bzero(V_bw_meter_timers, BW_METER_BUCKETS * sizeof(*V_bw_meter_timers));
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_reg_vif_num = VIFI_INVALID;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_UNLOCK();
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: done", __func__);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
* Set PIM assert processing global
*/
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
set_assert(int i)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
if ((i != 1) && (i != 0))
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return EINVAL;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_pim_assert_enabled = i;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return 0;
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Configure API capabilities
*/
int
set_api_config(uint32_t *apival)
{
Use an unsigned long when indexing into mfchashtbl[] and mf6ctable[]. This matches the types used when computing hash indices and the type of the maximum size of mfchashtbl[]. PR: kern/181821 Submitted by: Sven-Thorsten Dietrich <sven@vyatta.com> (IPv4) MFC after: 1 week
2013-09-05 14:16:37 +00:00
u_long i;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* We can set the API capabilities only if it is the first operation
* after MRT_INIT. I.e.:
* - there are no vifs installed
* - pim_assert is not enabled
* - the MFC table is empty
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_numvifs > 0) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
*apival = 0;
return EPERM;
}
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_pim_assert_enabled) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
*apival = 0;
return EPERM;
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
MFC_LOCK();
for (i = 0; i < mfchashsize; i++) {
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (LIST_FIRST(&V_mfchashtbl[i]) != NULL) {
Don't leak lock when returning. PR: kern/177888 Submitted by: Sven-Thorsten Dietrich <sven vyatta com> MFC after: 1 week
2013-04-16 19:25:41 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
*apival = 0;
return EPERM;
}
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
MFC_UNLOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_mrt_api_config = *apival & mrt_api_support;
*apival = V_mrt_api_config;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Add a vif to the vif table
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
add_vif(struct vifctl *vifcp)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
struct vif *vifp = V_viftable + vifcp->vifc_vifi;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct sockaddr_in sin = {sizeof sin, AF_INET};
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
struct ifaddr *ifa;
struct ifnet *ifp;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
int error;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK();
if (vifcp->vifc_vifi >= MAXVIFS) {
VIF_UNLOCK();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EINVAL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Nuke the token bucket filter code. Attempting to request rate limiting by the token bucket filter will result in EINVAL being returned. If you want to rate-limit traffic in future, use ALTQ or dummynet; this isn't a general purpose QoS engine. Preserve the now unused fields in struct vif so as to avoid having to recompile netstat(1) and other tools. Reviewed by: Pavlin Radslavov, Bill Fenner
2007-02-08 22:58:01 +00:00
/* rate limiting is no longer supported by this code */
if (vifcp->vifc_rate_limit != 0) {
log(LOG_ERR, "rate limiting is no longer supported\n");
VIF_UNLOCK();
return EINVAL;
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (!in_nullhost(vifp->v_lcl_addr)) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EADDRINUSE;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_nullhost(vifcp->vifc_lcl_addr)) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EADDRNOTAVAIL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* Find the interface with an address in AF_INET family */
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (vifcp->vifc_flags & VIFF_REGISTER) {
/*
* XXX: Because VIFF_REGISTER does not really need a valid
* local interface (e.g. it could be 127.0.0.2), we don't
* check its address.
*/
ifp = NULL;
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
} else {
The call into ifa_ifwithaddr() needs to be epoch protected; ortherwise we'll panic on an assertion. While here, leave a comment that the ifp was never protected and stable (as glebius pointed out) and this needs to be fixed properly. Discovered while working on: PR 246629 Reviewed by: glebius MFC after: 4 days Sponsored by: Rubicon Communications, LLC (d/b/a "Netgate")
2020-06-17 20:58:37 +00:00
struct epoch_tracker et;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
sin.sin_addr = vifcp->vifc_lcl_addr;
The call into ifa_ifwithaddr() needs to be epoch protected; ortherwise we'll panic on an assertion. While here, leave a comment that the ifp was never protected and stable (as glebius pointed out) and this needs to be fixed properly. Discovered while working on: PR 246629 Reviewed by: glebius MFC after: 4 days Sponsored by: Rubicon Communications, LLC (d/b/a "Netgate")
2020-06-17 20:58:37 +00:00
NET_EPOCH_ENTER(et);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
ifa = ifa_ifwithaddr((struct sockaddr *)&sin);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
if (ifa == NULL) {
The call into ifa_ifwithaddr() needs to be epoch protected; ortherwise we'll panic on an assertion. While here, leave a comment that the ifp was never protected and stable (as glebius pointed out) and this needs to be fixed properly. Discovered while working on: PR 246629 Reviewed by: glebius MFC after: 4 days Sponsored by: Rubicon Communications, LLC (d/b/a "Netgate")
2020-06-17 20:58:37 +00:00
NET_EPOCH_EXIT(et);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EADDRNOTAVAIL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
ifp = ifa->ifa_ifp;
The call into ifa_ifwithaddr() needs to be epoch protected; ortherwise we'll panic on an assertion. While here, leave a comment that the ifp was never protected and stable (as glebius pointed out) and this needs to be fixed properly. Discovered while working on: PR 246629 Reviewed by: glebius MFC after: 4 days Sponsored by: Rubicon Communications, LLC (d/b/a "Netgate")
2020-06-17 20:58:37 +00:00
/* XXX FIXME we need to take a ref on ifp and cleanup properly! */
NET_EPOCH_EXIT(et);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Remove support for IPIP tunnels in IPv4 multicast forwarding. XORP has never used them; with mrouted, their functionality may be replaced by explicitly configuring gif(4) instances and specifying them with the 'phyint' keyword. Bump __FreeBSD_version to 700030, and update UPDATING. A doc update is forthcoming. Discussed on: net Reviewed by: fenner MFC after: 3 months
2007-02-07 16:04:13 +00:00
if ((vifcp->vifc_flags & VIFF_TUNNEL) != 0) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: tunnels are no longer supported", __func__);
Remove support for IPIP tunnels in IPv4 multicast forwarding. XORP has never used them; with mrouted, their functionality may be replaced by explicitly configuring gif(4) instances and specifying them with the 'phyint' keyword. Bump __FreeBSD_version to 700030, and update UPDATING. A doc update is forthcoming. Discussed on: net Reviewed by: fenner MFC after: 3 months
2007-02-07 16:04:13 +00:00
VIF_UNLOCK();
return EOPNOTSUPP;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
} else if (vifcp->vifc_flags & VIFF_REGISTER) {
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
ifp = &V_multicast_register_if;
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR2(KTR_IPMF, "%s: add register vif for ifp %p", __func__, ifp);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_reg_vif_num == VIFI_INVALID) {
if_initname(&V_multicast_register_if, "register_vif", 0);
V_multicast_register_if.if_flags = IFF_LOOPBACK;
V_reg_vif_num = vifcp->vifc_vifi;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
} else { /* Make sure the interface supports multicast */
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
if ((ifp->if_flags & IFF_MULTICAST) == 0) {
VIF_UNLOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return EOPNOTSUPP;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* Enable promiscuous reception of all IP multicasts from the if */
Properly notice error returns from if_allmulti().
1997-02-21 18:34:58 +00:00
error = if_allmulti(ifp, 1);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
if (error) {
VIF_UNLOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return error;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
vifp->v_flags = vifcp->vifc_flags;
vifp->v_threshold = vifcp->vifc_threshold;
vifp->v_lcl_addr = vifcp->vifc_lcl_addr;
vifp->v_rmt_addr = vifcp->vifc_rmt_addr;
vifp->v_ifp = ifp;
/* initialize per vif pkt counters */
vifp->v_pkt_in = 0;
vifp->v_pkt_out = 0;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
vifp->v_bytes_in = 0;
vifp->v_bytes_out = 0;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* Adjust numvifs up if the vifi is higher than numvifs */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_numvifs <= vifcp->vifc_vifi)
V_numvifs = vifcp->vifc_vifi + 1;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR4(KTR_IPMF, "%s: add vif %d laddr 0x%08x thresh %x", __func__,
Add some ntohl() love to r315277 inet_ntoa() and inet_ntoa_r() take the address in network byte-order. When I removed those calls, I should have replaced them with ntohl() to make the hex addresses slightly less unreadable. Here they are. See r315277 regarding classic blunders. vangyzen: you're deep in "no good deed" territory, it seems --badger Reported by: ian MFC after: 3 days MFC when: I finally get it right Sponsored by: Dell EMC
2017-03-14 20:57:54 +00:00
(int)vifcp->vifc_vifi, ntohl(vifcp->vifc_lcl_addr.s_addr),
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
(int)vifcp->vifc_threshold);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Delete a vif from the vif table
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
static int
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
del_vif_locked(vifi_t vifi)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct vif *vifp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
VIF_LOCK_ASSERT();
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vifi >= V_numvifs) {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EINVAL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
vifp = &V_viftable[vifi];
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_nullhost(vifp->v_lcl_addr)) {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return EADDRNOTAVAIL;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (!(vifp->v_flags & (VIFF_TUNNEL | VIFF_REGISTER)))
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if_allmulti(vifp->v_ifp, 0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (vifp->v_flags & VIFF_REGISTER)
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_reg_vif_num = VIFI_INVALID;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
bzero((caddr_t)vifp, sizeof (*vifp));
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR2(KTR_IPMF, "%s: delete vif %d", __func__, (int)vifi);
Yow! Completely change the way socket options are handled, eliminating another specialized mbuf type in the process. Also clean up some of the cruft surrounding IPFW, multicast routing, RSVP, and other ill-explored corners.
1998-08-23 03:07:17 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* Adjust numvifs down */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = V_numvifs; vifi > 0; vifi--)
if (!in_nullhost(V_viftable[vifi-1].v_lcl_addr))
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
break;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_numvifs = vifi;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
return 0;
}
static int
del_vif(vifi_t vifi)
{
int cc;
VIF_LOCK();
cc = del_vif_locked(vifi);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Fix the IPv4 multicast routing detach path. On interface detach whilst the MROUTER is running, the system would panic as described in the PR. The fix in the PR is a good start, however, the other state associated with the multicast forwarding cache has to be freed in order to avoid leaking memory and other possible panics. More care and attention is needed in this area. PR: kern/82882 MFC after: 1 week
2006-09-28 12:21:08 +00:00
return cc;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
/*
* update an mfc entry without resetting counters and S,G addresses.
*/
static void
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
update_mfc_params(struct mfc *rt, struct mfcctl2 *mfccp)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
{
int i;
rt->mfc_parent = mfccp->mfcc_parent;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (i = 0; i < V_numvifs; i++) {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
rt->mfc_ttls[i] = mfccp->mfcc_ttls[i];
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
rt->mfc_flags[i] = mfccp->mfcc_flags[i] & V_mrt_api_config &
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
MRT_MFC_FLAGS_ALL;
}
/* set the RP address */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_mrt_api_config & MRT_MFC_RP)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
rt->mfc_rp = mfccp->mfcc_rp;
else
rt->mfc_rp.s_addr = INADDR_ANY;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
}
/*
* fully initialize an mfc entry from the parameter.
*/
static void
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
init_mfc_params(struct mfc *rt, struct mfcctl2 *mfccp)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
{
rt->mfc_origin = mfccp->mfcc_origin;
rt->mfc_mcastgrp = mfccp->mfcc_mcastgrp;
update_mfc_params(rt, mfccp);
/* initialize pkt counters per src-grp */
rt->mfc_pkt_cnt = 0;
rt->mfc_byte_cnt = 0;
rt->mfc_wrong_if = 0;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
timevalclear(&rt->mfc_last_assert);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static void
expire_mfc(struct mfc *rt)
{
struct rtdetq *rte, *nrte;
In expire_mfc(), add an assert on the multicast forwarding cache mutex. PR: 138666
2009-09-13 01:00:24 +00:00
MFC_LOCK_ASSERT();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
free_bw_list(rt->mfc_bw_meter);
TAILQ_FOREACH_SAFE(rte, &rt->mfc_stall, rte_link, nrte) {
m_freem(rte->m);
TAILQ_REMOVE(&rt->mfc_stall, rte, rte_link);
free(rte, M_MRTABLE);
}
LIST_REMOVE(rt, mfc_hash);
free(rt, M_MRTABLE);
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Add an mfc entry
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
static int
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
add_mfc(struct mfcctl2 *mfccp)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
struct mfc *rt;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
struct rtdetq *rte, *nrte;
u_long hash = 0;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
u_short nstl;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK();
MFC_LOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt = mfc_find(&mfccp->mfcc_origin, &mfccp->mfcc_mcastgrp);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* If an entry already exists, just update the fields */
if (rt) {
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR4(KTR_IPMF, "%s: update mfc orig 0x%08x group %lx parent %x",
Add some ntohl() love to r315277 inet_ntoa() and inet_ntoa_r() take the address in network byte-order. When I removed those calls, I should have replaced them with ntohl() to make the hex addresses slightly less unreadable. Here they are. See r315277 regarding classic blunders. vangyzen: you're deep in "no good deed" territory, it seems --badger Reported by: ian MFC after: 3 days MFC when: I finally get it right Sponsored by: Dell EMC
2017-03-14 20:57:54 +00:00
__func__, ntohl(mfccp->mfcc_origin.s_addr),
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
(u_long)ntohl(mfccp->mfcc_mcastgrp.s_addr),
mfccp->mfcc_parent);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
update_mfc_params(rt, mfccp);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (0);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Find the entry for which the upcall was made and update
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
nstl = 0;
hash = MFCHASH(mfccp->mfcc_origin, mfccp->mfcc_mcastgrp);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_FOREACH(rt, &V_mfchashtbl[hash], mfc_hash) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_hosteq(rt->mfc_origin, mfccp->mfcc_origin) &&
in_hosteq(rt->mfc_mcastgrp, mfccp->mfcc_mcastgrp) &&
!TAILQ_EMPTY(&rt->mfc_stall)) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR5(KTR_IPMF,
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
"%s: add mfc orig 0x%08x group %lx parent %x qh %p",
Add some ntohl() love to r315277 inet_ntoa() and inet_ntoa_r() take the address in network byte-order. When I removed those calls, I should have replaced them with ntohl() to make the hex addresses slightly less unreadable. Here they are. See r315277 regarding classic blunders. vangyzen: you're deep in "no good deed" territory, it seems --badger Reported by: ian MFC after: 3 days MFC when: I finally get it right Sponsored by: Dell EMC
2017-03-14 20:57:54 +00:00
__func__, ntohl(mfccp->mfcc_origin.s_addr),
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
(u_long)ntohl(mfccp->mfcc_mcastgrp.s_addr),
mfccp->mfcc_parent,
TAILQ_FIRST(&rt->mfc_stall));
if (nstl++)
CTR1(KTR_IPMF, "%s: multiple matches", __func__);
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
init_mfc_params(rt, mfccp);
rt->mfc_expire = 0; /* Don't clean this guy up */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_nexpire[hash]--;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/* Free queued packets, but attempt to forward them first. */
TAILQ_FOREACH_SAFE(rte, &rt->mfc_stall, rte_link, nrte) {
if (rte->ifp != NULL)
ip_mdq(rte->m, rte->ifp, rt, -1);
m_freem(rte->m);
TAILQ_REMOVE(&rt->mfc_stall, rte, rte_link);
rt->mfc_nstall--;
free(rte, M_MRTABLE);
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
}
/*
* It is possible that an entry is being inserted without an upcall
*/
if (nstl == 0) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: adding mfc w/o upcall", __func__);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_FOREACH(rt, &V_mfchashtbl[hash], mfc_hash) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_hosteq(rt->mfc_origin, mfccp->mfcc_origin) &&
in_hosteq(rt->mfc_mcastgrp, mfccp->mfcc_mcastgrp)) {
init_mfc_params(rt, mfccp);
if (rt->mfc_expire)
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_nexpire[hash]--;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt->mfc_expire = 0;
break; /* XXX */
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (rt == NULL) { /* no upcall, so make a new entry */
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
rt = (struct mfc *)malloc(sizeof(*rt), M_MRTABLE, M_NOWAIT);
if (rt == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (ENOBUFS);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
init_mfc_params(rt, mfccp);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
TAILQ_INIT(&rt->mfc_stall);
rt->mfc_nstall = 0;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt->mfc_expire = 0;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
rt->mfc_bw_meter = NULL;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
/* insert new entry at head of hash chain */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_INSERT_HEAD(&V_mfchashtbl[hash], rt, mfc_hash);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Delete an mfc entry
*/
static int
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
del_mfc(struct mfcctl2 *mfccp)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
struct in_addr origin;
struct in_addr mcastgrp;
struct mfc *rt;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
origin = mfccp->mfcc_origin;
mcastgrp = mfccp->mfcc_mcastgrp;
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR3(KTR_IPMF, "%s: delete mfc orig 0x%08x group %lx", __func__,
Add some ntohl() love to r315277 inet_ntoa() and inet_ntoa_r() take the address in network byte-order. When I removed those calls, I should have replaced them with ntohl() to make the hex addresses slightly less unreadable. Here they are. See r315277 regarding classic blunders. vangyzen: you're deep in "no good deed" territory, it seems --badger Reported by: ian MFC after: 3 days MFC when: I finally get it right Sponsored by: Dell EMC
2017-03-14 20:57:54 +00:00
ntohl(origin.s_addr), (u_long)ntohl(mcastgrp.s_addr));
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt = mfc_find(&origin, &mcastgrp);
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
if (rt == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return EADDRNOTAVAIL;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* free the bw_meter entries
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
free_bw_list(rt->mfc_bw_meter);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
rt->mfc_bw_meter = NULL;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
LIST_REMOVE(rt, mfc_hash);
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
free(rt, M_MRTABLE);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Send a message to the routing daemon on the multicast routing socket.
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
*/
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
socket_send(struct socket *s, struct mbuf *mm, struct sockaddr_in *src)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (s) {
Reduce the number of unnecessary unlock-relocks on socket buffer mutexes associated with performing a wakeup on the socket buffer: - When performing an sbappend*() followed by a so[rw]wakeup(), explicitly acquire the socket buffer lock and use the _locked() variants of both calls. Note that the _locked() sowakeup() versions unlock the mutex on return. This is done in uipc_send(), divert_packet(), mroute socket_send(), raw_append(), tcp_reass(), tcp_input(), and udp_append(). - When the socket buffer lock is dropped before a sowakeup(), remove the explicit unlock and use the _locked() sowakeup() variant. This is done in soisdisconnecting(), soisdisconnected() when setting the can't send/ receive flags and dropping data, and in uipc_rcvd() which adjusting back-pressure on the sockets. For UNIX domain sockets running mpsafe with a contention-intensive SMP mysql benchmark, this results in a 1.6% query rate improvement due to reduce mutex costs.
2004-06-26 19:10:39 +00:00
SOCKBUF_LOCK(&s->so_rcv);
if (sbappendaddr_locked(&s->so_rcv, (struct sockaddr *)src, mm,
NULL) != 0) {
sorwakeup_locked(s);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return 0;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Reduce the number of unnecessary unlock-relocks on socket buffer mutexes associated with performing a wakeup on the socket buffer: - When performing an sbappend*() followed by a so[rw]wakeup(), explicitly acquire the socket buffer lock and use the _locked() variants of both calls. Note that the _locked() sowakeup() versions unlock the mutex on return. This is done in uipc_send(), divert_packet(), mroute socket_send(), raw_append(), tcp_reass(), tcp_input(), and udp_append(). - When the socket buffer lock is dropped before a sowakeup(), remove the explicit unlock and use the _locked() sowakeup() variant. This is done in soisdisconnecting(), soisdisconnected() when setting the can't send/ receive flags and dropping data, and in uipc_rcvd() which adjusting back-pressure on the sockets. For UNIX domain sockets running mpsafe with a contention-intensive SMP mysql benchmark, this results in a 1.6% query rate improvement due to reduce mutex costs.
2004-06-26 19:10:39 +00:00
SOCKBUF_UNLOCK(&s->so_rcv);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
}
m_freem(mm);
return -1;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* IP multicast forwarding function. This function assumes that the packet
* pointed to by "ip" has arrived on (or is about to be sent to) the interface
* pointed to by "ifp", and the packet is to be relayed to other networks
* that have members of the packet's destination IP multicast group.
*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* The packet is returned unscathed to the caller, unless it is
* erroneous, in which case a non-zero return value tells the caller to
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* discard it.
*/
#define TUNNEL_LEN 12 /* # bytes of IP option for tunnel encapsulation */
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
X_ip_mforward(struct ip *ip, struct ifnet *ifp, struct mbuf *m,
struct ip_moptions *imo)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct mfc *rt;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
int error;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
vifi_t vifi;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR3(KTR_IPMF, "ip_mforward: delete mfc orig 0x%08x group %lx ifp %p",
ntohl(ip->ip_src.s_addr), (u_long)ntohl(ip->ip_dst.s_addr), ifp);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Somewhat modernize ip_mroute.c: - Use sysctl to export stats - Use ip_encap.c's encapsulation support - Update lkm to kld (is 6 years a record for a broken module?) - Remove some unused cruft
2001-07-25 20:15:49 +00:00
if (ip->ip_hl < (sizeof(struct ip) + TUNNEL_LEN) >> 2 ||
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
((u_char *)(ip + 1))[1] != IPOPT_LSRR ) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* Packet arrived via a physical interface or
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* an encapsulated tunnel or a register_vif.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
} else {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* Packet arrived through a source-route tunnel.
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* Source-route tunnels are no longer supported.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
return (1);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK();
MFC_LOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (imo && ((vifi = imo->imo_multicast_vif) < V_numvifs)) {
Use MAXTTL. Obtained from: NetBSD
2007-02-10 23:15:28 +00:00
if (ip->ip_ttl < MAXTTL)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip->ip_ttl++; /* compensate for -1 in *_send routines */
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
error = ip_mdq(m, ifp, NULL, vifi);
MFC_UNLOCK();
VIF_UNLOCK();
return error;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Don't forward a packet with time-to-live of zero or one,
* or a packet destined to a local-only group.
*/
Diff reduction with NetBSD; use IN_LOCAL_GROUP() to check if an address is within the locally scoped multicast range 224.0.0.0/24.
2007-03-15 08:44:22 +00:00
if (ip->ip_ttl <= 1 || IN_LOCAL_GROUP(ntohl(ip->ip_dst.s_addr))) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return 0;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Determine forwarding vifs from the forwarding cache table
*/
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_mfc_lookups);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt = mfc_find(&ip->ip_src, &ip->ip_dst);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* Entry exists, so forward if necessary */
if (rt != NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
error = ip_mdq(m, ifp, rt, -1);
MFC_UNLOCK();
VIF_UNLOCK();
return error;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
} else {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
* If we don't have a route for packet's origin,
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
* Make a copy of the packet & send message to routing daemon
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct mbuf *mb0;
struct rtdetq *rte;
u_long hash;
Cleaned up uninitialized 'rt' warning properly Make a copy of the header of a packet that gets queued due to lack of forwarding cache entry, so that nobody else can step on it. Thanks to Mike Karels <karels@bsdi.com> for pointing this one out.
1996-03-11 17:11:23 +00:00
int hlen = ip->ip_hl << 2;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_mfc_misses);
MRTSTAT_INC(mrts_no_route);
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR2(KTR_IPMF, "ip_mforward: no mfc for (0x%08x,%lx)",
ntohl(ip->ip_src.s_addr), (u_long)ntohl(ip->ip_dst.s_addr));
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
* Allocate mbufs early so that we don't do extra work if we are
Cleaned up uninitialized 'rt' warning properly Make a copy of the header of a packet that gets queued due to lack of forwarding cache entry, so that nobody else can step on it. Thanks to Mike Karels <karels@bsdi.com> for pointing this one out.
1996-03-11 17:11:23 +00:00
* just going to fail anyway. Make sure to pullup the header so
* that other people can't step on it.
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rte = (struct rtdetq *)malloc((sizeof *rte), M_MRTABLE,
M_NOWAIT|M_ZERO);
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
if (rte == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return ENOBUFS;
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Mechanically substitute flags from historic mbuf allocator with malloc(9) flags within sys. Exceptions: - sys/contrib not touched - sys/mbuf.h edited manually
2012-12-05 08:04:20 +00:00
mb0 = m_copypacket(m, M_NOWAIT);
When deciding whether to call m_pullup() even though there is adequate data in an mbuf, use M_WRITABLE() instead of a direct test of M_EXT; the latter both unnecessarily exposes mbuf-allocator internals in the protocol stack and is also insufficient to catch all cases of non-writability. (NB: m_pullup() does not actually guarantee that a writable mbuf is returned, so further refinement of all of these code paths continues to be required.) Reviewed by: bz MFC after: 3 days Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D900
2014-10-12 15:49:52 +00:00
if (mb0 && (!M_WRITABLE(mb0) || mb0->m_len < hlen))
Cleaned up uninitialized 'rt' warning properly Make a copy of the header of a packet that gets queued due to lack of forwarding cache entry, so that nobody else can step on it. Thanks to Mike Karels <karels@bsdi.com> for pointing this one out.
1996-03-11 17:11:23 +00:00
mb0 = m_pullup(mb0, hlen);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
if (mb0 == NULL) {
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
free(rte, M_MRTABLE);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return ENOBUFS;
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
/* is there an upcall waiting for this flow ? */
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
hash = MFCHASH(ip->ip_src, ip->ip_dst);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_FOREACH(rt, &V_mfchashtbl[hash], mfc_hash) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (in_hosteq(ip->ip_src, rt->mfc_origin) &&
in_hosteq(ip->ip_dst, rt->mfc_mcastgrp) &&
!TAILQ_EMPTY(&rt->mfc_stall))
break;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
if (rt == NULL) {
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
int i;
struct igmpmsg *im;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct sockaddr_in k_igmpsrc = { sizeof k_igmpsrc, AF_INET };
struct mbuf *mm;
/*
* Locate the vifi for the incoming interface for this packet.
* If none found, drop packet.
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = 0; vifi < V_numvifs &&
V_viftable[vifi].v_ifp != ifp; vifi++)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vifi >= V_numvifs) /* vif not found, drop packet */
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
goto non_fatal;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* no upcall, so make a new entry */
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
rt = (struct mfc *)malloc(sizeof(*rt), M_MRTABLE, M_NOWAIT);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (rt == NULL)
goto fail;
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/* Make a copy of the header to send to the user level process */
Remove the 4.3BSD compatible macro m_copy(), use m_copym() instead. Reviewed by: gnn Differential Revision: https://reviews.freebsd.org/D7878
2016-09-15 07:41:48 +00:00
mm = m_copym(mb0, 0, hlen, M_NOWAIT);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (mm == NULL)
goto fail1;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
/*
* Send message to routing daemon to install
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* a route into the kernel table
*/
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
im = mtod(mm, struct igmpmsg *);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
im->im_msgtype = IGMPMSG_NOCACHE;
im->im_mbz = 0;
im->im_vif = vifi;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upcalls);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
k_igmpsrc.sin_addr = ip->ip_src;
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (socket_send(V_ip_mrouter, mm, &k_igmpsrc) < 0) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR0(KTR_IPMF, "ip_mforward: socket queue full");
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upq_sockfull);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
fail1:
free(rt, M_MRTABLE);
fail:
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
free(rte, M_MRTABLE);
Fix a resource allocation bug where multicast forwarding would leak mbufs in certain cases when allocation of another mbuf has already failed. Submitted by: Bill Fenner <fenner@parc.xerox.com>
1995-06-19 17:22:01 +00:00
m_freem(mb0);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return ENOBUFS;
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* insert new entry at head of hash chain */
rt->mfc_origin.s_addr = ip->ip_src.s_addr;
rt->mfc_mcastgrp.s_addr = ip->ip_dst.s_addr;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
rt->mfc_expire = UPCALL_EXPIRE;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_nexpire[hash]++;
for (i = 0; i < V_numvifs; i++) {
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
rt->mfc_ttls[i] = 0;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
rt->mfc_flags[i] = 0;
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
rt->mfc_parent = -1;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/* clear the RP address */
rt->mfc_rp.s_addr = INADDR_ANY;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
rt->mfc_bw_meter = NULL;
Make sure the multicast forwarding cache entry's stall queue is properly initialized before trying to insert an entry into it. PR: kern/142052 Reviewed by: bms MFC after: now
2009-12-30 08:52:13 +00:00
/* initialize pkt counters per src-grp */
rt->mfc_pkt_cnt = 0;
rt->mfc_byte_cnt = 0;
rt->mfc_wrong_if = 0;
timevalclear(&rt->mfc_last_assert);
TAILQ_INIT(&rt->mfc_stall);
rt->mfc_nstall = 0;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/* link into table */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_INSERT_HEAD(&V_mfchashtbl[hash], rt, mfc_hash);
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
TAILQ_INSERT_HEAD(&rt->mfc_stall, rte, rte_link);
rt->mfc_nstall++;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
} else {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/* determine if queue has overflowed */
if (rt->mfc_nstall > MAX_UPQ) {
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upq_ovflw);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
non_fatal:
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
free(rte, M_MRTABLE);
Fix a resource allocation bug where multicast forwarding would leak mbufs in certain cases when allocation of another mbuf has already failed. Submitted by: Bill Fenner <fenner@parc.xerox.com>
1995-06-19 17:22:01 +00:00
m_freem(mb0);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (0);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
TAILQ_INSERT_TAIL(&rt->mfc_stall, rte, rte_link);
rt->mfc_nstall++;
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
rte->m = mb0;
rte->ifp = ifp;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
VIF_UNLOCK();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Clean up the cache entry if upcall is not serviced
*/
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
static void
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
expire_upcalls(void *arg)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Use an unsigned long when indexing into mfchashtbl[] and mf6ctable[]. This matches the types used when computing hash indices and the type of the maximum size of mfchashtbl[]. PR: kern/181821 Submitted by: Sven-Thorsten Dietrich <sven@vyatta.com> (IPv4) MFC after: 1 week
2013-09-05 14:16:37 +00:00
u_long i;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
CURVNET_SET((struct vnet *) arg);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
for (i = 0; i < mfchashsize; i++) {
struct mfc *rt, *nrt;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_nexpire[i] == 0)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
continue;
Use dynamic memory allocation instead of mbuf's for multicast routing state. Note: this requires a recompilation of netstat (but netstat has been broken since rev 1.52 of ip_mroute.c anyway) Obtained from: Significantly based on Steve McCanne's <mccanne@cs.berkeley.edu> work for BSD/OS
1999-01-18 02:06:59 +00:00
Use LIST_FOREACH_SAFE() instead of doing it by hand.
2013-09-05 14:26:37 +00:00
LIST_FOREACH_SAFE(rt, &V_mfchashtbl[i], mfc_hash, nrt) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (TAILQ_EMPTY(&rt->mfc_stall))
continue;
if (rt->mfc_expire == 0 || --rt->mfc_expire > 0)
continue;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* free the bw_meter entries
*/
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
while (rt->mfc_bw_meter != NULL) {
struct bw_meter *x = rt->mfc_bw_meter;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
rt->mfc_bw_meter = x->bm_mfc_next;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
free(x, M_BWMETER);
}
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_cache_cleanups);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR3(KTR_IPMF, "%s: expire (%lx, %lx)", __func__,
(u_long)ntohl(rt->mfc_origin.s_addr),
(u_long)ntohl(rt->mfc_mcastgrp.s_addr));
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
expire_mfc(rt);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
callout_reset(&V_expire_upcalls_ch, EXPIRE_TIMEOUT, expire_upcalls,
curvnet);
CURVNET_RESTORE();
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
/*
* Packet forwarding routine once entry in the cache is made
*/
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_mdq(struct mbuf *m, struct ifnet *ifp, struct mfc *rt, vifi_t xmt_vif)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct ip *ip = mtod(m, struct ip *);
vifi_t vifi;
Switch the entire IPv4 stack to keep the IP packet header in network byte order. Any host byte order processing is done in local variables and host byte order values are never[1] written to a packet. After this change a packet processed by the stack isn't modified at all[2] except for TTL. After this change a network stack hacker doesn't need to scratch his head trying to figure out what is the byte order at the given place in the stack. [1] One exception still remains. The raw sockets convert host byte order before pass a packet to an application. Probably this would remain for ages for compatibility. [2] The ip_input() still subtructs header len from ip->ip_len, but this is planned to be fixed soon. Reviewed by: luigi, Maxim Dounin <mdounin mdounin.ru> Tested by: ray, Olivier Cochard-Labbe <olivier cochard.me>
2012-10-22 21:09:03 +00:00
int plen = ntohs(ip->ip_len);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK_ASSERT();
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
* If xmt_vif is not -1, send on only the requested vif.
*
* (since vifi_t is u_short, -1 becomes MAXUSHORT, which > numvifs.)
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (xmt_vif < V_numvifs) {
if (V_viftable[xmt_vif].v_flags & VIFF_REGISTER)
pim_register_send(ip, V_viftable + xmt_vif, m, rt);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
else
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
phyint_send(ip, V_viftable + xmt_vif, m);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return 1;
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* Don't forward if it didn't arrive from the parent vif for its origin.
*/
vifi = rt->mfc_parent;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if ((vifi >= V_numvifs) || (V_viftable[vifi].v_ifp != ifp)) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR4(KTR_IPMF, "%s: rx on wrong ifp %p (vifi %d, v_ifp %p)",
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
__func__, ifp, (int)vifi, V_viftable[vifi].v_ifp);
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_wrong_if);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
++rt->mfc_wrong_if;
/*
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* If we are doing PIM assert processing, send a message
* to the routing daemon.
*
* XXX: A PIM-SM router needs the WRONGVIF detection so it
* can complete the SPT switch, regardless of the type
* of the iif (broadcast media, GRE tunnel, etc).
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_pim_assert_enabled && (vifi < V_numvifs) &&
V_viftable[vifi].v_ifp) {
if (ifp == &V_multicast_register_if)
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_registers_wrongiif);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* makes mfc[MFCTBLSIZ] and vif[MAXVIFS] tables accessible via sysctl: - sysctlbyname("net.inet.ip.mfctable", ...) - sysctlbyname("net.inet.ip.viftable", ...) This change is needed so netstat can use sysctlbyname() to read the data from those tables. Otherwise, in some cases "netstat -g" may fail to report the multicast forwarding information (e.g., if we run a multicast router on PicoBSD). * Bug fix: when sending IGMPMSG_WRONGVIF upcall to the multicast routing daemon, set properly "im->im_vif" to the receiving incoming interface of the packet that triggered that upcall rather than to the expected incoming interface of that packet. * Bug fix: add missing increment of counter "mrtstat.mrts_upcalls" * Few formatting nits (e.g., replace extra spaces with TABs) Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-05 17:01:33 +00:00
/* Get vifi for the incoming packet */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = 0; vifi < V_numvifs && V_viftable[vifi].v_ifp != ifp;
vifi++)
* makes mfc[MFCTBLSIZ] and vif[MAXVIFS] tables accessible via sysctl: - sysctlbyname("net.inet.ip.mfctable", ...) - sysctlbyname("net.inet.ip.viftable", ...) This change is needed so netstat can use sysctlbyname() to read the data from those tables. Otherwise, in some cases "netstat -g" may fail to report the multicast forwarding information (e.g., if we run a multicast router on PicoBSD). * Bug fix: when sending IGMPMSG_WRONGVIF upcall to the multicast routing daemon, set properly "im->im_vif" to the receiving incoming interface of the packet that triggered that upcall rather than to the expected incoming interface of that packet. * Bug fix: add missing increment of counter "mrtstat.mrts_upcalls" * Few formatting nits (e.g., replace extra spaces with TABs) Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-05 17:01:33 +00:00
;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vifi >= V_numvifs)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0; /* The iif is not found: ignore the packet. */
if (rt->mfc_flags[vifi] & MRT_MFC_FLAGS_DISABLE_WRONGVIF)
return 0; /* WRONGVIF disabled: ignore the packet */
* makes mfc[MFCTBLSIZ] and vif[MAXVIFS] tables accessible via sysctl: - sysctlbyname("net.inet.ip.mfctable", ...) - sysctlbyname("net.inet.ip.viftable", ...) This change is needed so netstat can use sysctlbyname() to read the data from those tables. Otherwise, in some cases "netstat -g" may fail to report the multicast forwarding information (e.g., if we run a multicast router on PicoBSD). * Bug fix: when sending IGMPMSG_WRONGVIF upcall to the multicast routing daemon, set properly "im->im_vif" to the receiving incoming interface of the packet that triggered that upcall rather than to the expected incoming interface of that packet. * Bug fix: add missing increment of counter "mrtstat.mrts_upcalls" * Few formatting nits (e.g., replace extra spaces with TABs) Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-05 17:01:33 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (ratecheck(&rt->mfc_last_assert, &pim_assert_interval)) {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct sockaddr_in k_igmpsrc = { sizeof k_igmpsrc, AF_INET };
struct igmpmsg *im;
int hlen = ip->ip_hl << 2;
Remove the 4.3BSD compatible macro m_copy(), use m_copym() instead. Reviewed by: gnn Differential Revision: https://reviews.freebsd.org/D7878
2016-09-15 07:41:48 +00:00
struct mbuf *mm = m_copym(m, 0, hlen, M_NOWAIT);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
When deciding whether to call m_pullup() even though there is adequate data in an mbuf, use M_WRITABLE() instead of a direct test of M_EXT; the latter both unnecessarily exposes mbuf-allocator internals in the protocol stack and is also insufficient to catch all cases of non-writability. (NB: m_pullup() does not actually guarantee that a writable mbuf is returned, so further refinement of all of these code paths continues to be required.) Reviewed by: bz MFC after: 3 days Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D900
2014-10-12 15:49:52 +00:00
if (mm && (!M_WRITABLE(mm) || mm->m_len < hlen))
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
mm = m_pullup(mm, hlen);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if (mm == NULL)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return ENOBUFS;
im = mtod(mm, struct igmpmsg *);
im->im_msgtype = IGMPMSG_WRONGVIF;
im->im_mbz = 0;
im->im_vif = vifi;
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upcalls);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* makes mfc[MFCTBLSIZ] and vif[MAXVIFS] tables accessible via sysctl: - sysctlbyname("net.inet.ip.mfctable", ...) - sysctlbyname("net.inet.ip.viftable", ...) This change is needed so netstat can use sysctlbyname() to read the data from those tables. Otherwise, in some cases "netstat -g" may fail to report the multicast forwarding information (e.g., if we run a multicast router on PicoBSD). * Bug fix: when sending IGMPMSG_WRONGVIF upcall to the multicast routing daemon, set properly "im->im_vif" to the receiving incoming interface of the packet that triggered that upcall rather than to the expected incoming interface of that packet. * Bug fix: add missing increment of counter "mrtstat.mrts_upcalls" * Few formatting nits (e.g., replace extra spaces with TABs) Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-05 17:01:33 +00:00
k_igmpsrc.sin_addr = im->im_src;
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (socket_send(V_ip_mrouter, mm, &k_igmpsrc) < 0) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: socket queue full", __func__);
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upq_sockfull);
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
return ENOBUFS;
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
}
return 0;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/* If I sourced this packet, it counts as output, else it was input. */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (in_hosteq(ip->ip_src, V_viftable[vifi].v_lcl_addr)) {
V_viftable[vifi].v_pkt_out++;
V_viftable[vifi].v_bytes_out += plen;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
} else {
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_viftable[vifi].v_pkt_in++;
V_viftable[vifi].v_bytes_in += plen;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
rt->mfc_pkt_cnt++;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
rt->mfc_byte_cnt += plen;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
/*
* For each vif, decide if a copy of the packet should be forwarded.
* Forward if:
* - the ttl exceeds the vif's threshold
* - there are group members downstream on interface
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (vifi = 0; vifi < V_numvifs; vifi++)
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
if ((rt->mfc_ttls[vifi] > 0) && (ip->ip_ttl > rt->mfc_ttls[vifi])) {
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_viftable[vifi].v_pkt_out++;
V_viftable[vifi].v_bytes_out += plen;
if (V_viftable[vifi].v_flags & VIFF_REGISTER)
pim_register_send(ip, V_viftable + vifi, m, rt);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
else
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
phyint_send(ip, V_viftable + vifi, m);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Perform upcall-related bw measuring.
*/
if (rt->mfc_bw_meter != NULL) {
struct bw_meter *x;
struct timeval now;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
microtime(&now);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
for (x = rt->mfc_bw_meter; x != NULL; x = x->bm_mfc_next)
bw_meter_receive_packet(x, plen, &now);
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
return 0;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* Check if a vif number is legal/ok. This is used by in_mcast.c.
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static int
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
X_legal_vif_num(int vif)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
int ret;
ret = 0;
if (vif < 0)
return (ret);
VIF_LOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vif < V_numvifs)
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
ret = 1;
VIF_UNLOCK();
return (ret);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
/*
* Return the local address used by this vif
*/
New style sysctl & staticize alot of stuff.
1995-11-14 20:34:56 +00:00
static u_long
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
X_ip_mcast_src(int vifi)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
in_addr_t addr;
addr = INADDR_ANY;
if (vifi < 0)
return (addr);
VIF_LOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (vifi < V_numvifs)
addr = V_viftable[vifi].v_lcl_addr.s_addr;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
VIF_UNLOCK();
return (addr);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
static void
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
phyint_send(struct ip *ip, struct vif *vifp, struct mbuf *m)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct mbuf *mb_copy;
int hlen = ip->ip_hl << 2;
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK_ASSERT();
Fix some endianness and packet header bugs found in BSDi's port of this code. (From mbone mailing-list.)
1994-10-13 22:12:42 +00:00
/*
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
* Make a new reference to the packet; make sure that
* the IP header is actually copied, not just referenced,
* so that ip_output() only scribbles on the copy.
Fix some endianness and packet header bugs found in BSDi's port of this code. (From mbone mailing-list.)
1994-10-13 22:12:42 +00:00
*/
Mechanically substitute flags from historic mbuf allocator with malloc(9) flags within sys. Exceptions: - sys/contrib not touched - sys/mbuf.h edited manually
2012-12-05 08:04:20 +00:00
mb_copy = m_copypacket(m, M_NOWAIT);
When deciding whether to call m_pullup() even though there is adequate data in an mbuf, use M_WRITABLE() instead of a direct test of M_EXT; the latter both unnecessarily exposes mbuf-allocator internals in the protocol stack and is also insufficient to catch all cases of non-writability. (NB: m_pullup() does not actually guarantee that a writable mbuf is returned, so further refinement of all of these code paths continues to be required.) Reviewed by: bz MFC after: 3 days Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D900
2014-10-12 15:49:52 +00:00
if (mb_copy && (!M_WRITABLE(mb_copy) || mb_copy->m_len < hlen))
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
mb_copy = m_pullup(mb_copy, hlen);
Fix some endianness and packet header bugs found in BSDi's port of this code. (From mbone mailing-list.)
1994-10-13 22:12:42 +00:00
if (mb_copy == NULL)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
return;
Fix some endianness and packet header bugs found in BSDi's port of this code. (From mbone mailing-list.)
1994-10-13 22:12:42 +00:00
Nuke the token bucket filter code. Attempting to request rate limiting by the token bucket filter will result in EINVAL being returned. If you want to rate-limit traffic in future, use ALTQ or dummynet; this isn't a general purpose QoS engine. Preserve the now unused fields in struct vif so as to avoid having to recompile netstat(1) and other tools. Reviewed by: Pavlin Radslavov, Bill Fenner
2007-02-08 22:58:01 +00:00
send_packet(vifp, mb_copy);
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
static void
Nuke the token bucket filter code. Attempting to request rate limiting by the token bucket filter will result in EINVAL being returned. If you want to rate-limit traffic in future, use ALTQ or dummynet; this isn't a general purpose QoS engine. Preserve the now unused fields in struct vif so as to avoid having to recompile netstat(1) and other tools. Reviewed by: Pavlin Radslavov, Bill Fenner
2007-02-08 22:58:01 +00:00
send_packet(struct vif *vifp, struct mbuf *m)
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
{
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
struct ip_moptions imo;
netinet silence warnings
2018-05-19 05:56:21 +00:00
int error __unused;
Nuke the token bucket filter code. Attempting to request rate limiting by the token bucket filter will result in EINVAL being returned. If you want to rate-limit traffic in future, use ALTQ or dummynet; this isn't a general purpose QoS engine. Preserve the now unused fields in struct vif so as to avoid having to recompile netstat(1) and other tools. Reviewed by: Pavlin Radslavov, Bill Fenner
2007-02-08 22:58:01 +00:00
VIF_LOCK_ASSERT();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
Fix some problems with multicast forwarding: Garrett, Here are some patches for the rate limiting code. It should be faster, and in particular it doesn't leak malloc'd memory any more when rate_limit'ing a phyint. It now uses an mbuf chain at each vif, instead of the static queue array. This means that the MAXQSIZE is now variable per vif (although there is no interface to change it other than a debugger); this is an area for more experimentation. Bill Submitted by: Bill Fenner <fenner@parc.xerox.com>
1995-08-23 18:20:17 +00:00
imo.imo_multicast_ifp = vifp->v_ifp;
imo.imo_multicast_ttl = mtod(m, struct ip *)->ip_ttl - 1;
imo.imo_multicast_loop = 1;
imo.imo_multicast_vif = -1;
Convert all IPv4 and IPv6 multicast memberships into using a STAILQ instead of a linear array. The multicast memberships for the inpcb structure are protected by a non-sleepable lock, INP_WLOCK(), which needs to be dropped when calling the underlying possibly sleeping if_ioctl() method. When using a linear array to keep track of multicast memberships, the computed memory location of the multicast filter may suddenly change, due to concurrent insertion or removal of elements in the linear array. This in turn leads to various invalid memory access issues and kernel panics. To avoid this problem, put all multicast memberships on a STAILQ based list. Then the memory location of the IPv4 and IPv6 multicast filters become fixed during their lifetime and use after free and memory leak issues are easier to track, for example by: vmstat -m | grep multi All list manipulation has been factored into inline functions including some macros, to easily allow for a future hash-list implementation, if needed. This patch has been tested by pho@ . Differential Revision: https://reviews.freebsd.org/D20080 Reviewed by: markj @ MFC after: 1 week Sponsored by: Mellanox Technologies
2019-06-25 11:54:41 +00:00
STAILQ_INIT(&imo.imo_head);
Fix some problems with multicast forwarding: Garrett, Here are some patches for the rate limiting code. It should be faster, and in particular it doesn't leak malloc'd memory any more when rate_limit'ing a phyint. It now uses an mbuf chain at each vif, instead of the static queue array. This means that the MAXQSIZE is now variable per vif (although there is no interface to change it other than a debugger); this is an area for more experimentation. Bill Submitted by: Bill Fenner <fenner@parc.xerox.com>
1995-08-23 18:20:17 +00:00
Always call ip_output() with a valid route pointer. For igmp, also get the multicast option structure off the stack rather than malloc.
1996-04-18 15:41:11 +00:00
/*
* Re-entrancy should not be a problem here, because
* the packets that we send out and are looped back at us
* should get rejected because they appear to come from
* the loopback interface, thus preventing looping.
*/
Remove route caching from IP multicast routing code. There is no reason to do that, and also, cached route never got unreferenced, which meant a reference leak. Reviewed by: bms
2012-07-02 19:44:18 +00:00
error = ip_output(m, NULL, NULL, IP_FORWARDING, &imo, NULL);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR3(KTR_IPMF, "%s: vif %td err %d", __func__,
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
(ptrdiff_t)(vifp - V_viftable), error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Initial get-the-easy-case-working upgrade of the multicast code to something more recent than the ancient 1.2 release contained in 4.4. This code has the following advantages as compared to previous versions (culled from the README file for the SunOS release): - True multicast delivery - Configurable rate-limiting of forwarded multicast traffic on each physical interface or tunnel, using a token-bucket limiter. - Simplistic classification of packets for prioritized dropping. - Administrative scoping of multicast address ranges. - Faster detection of hosts leaving groups. - Support for multicast traceroute (code not yet available). - Support for RSVP, the Resource Reservation Protocol. What still needs to be done: - The multicast forwarder needs testing. - The multicast routing daemon needs to be ported. - Network interface drivers need to have the `#ifdef MULTICAST' goop ripped out of them. - The IGMP code should probably be bogon-tested. Some notes about the porting process: In some cases, the Berkeley people decided to incorporate functionality from later releases of the multicast code, but then had to do things differently. As a result, if you look at Deering's patches, and then look at our code, it is not always obvious whether the patch even applies. Let the reader beware. I ran ip_mroute.c through several passes of `unifdef' to get rid of useless grot, and to permanently enable the RSVP support, which we will include as standard. Ported by: Garrett Wollman Submitted by: Steve Deering and Ajit Thyagarajan (among others)
1994-09-06 22:42:31 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
/*
* Stubs for old RSVP socket shim implementation.
*/
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
static int
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
X_ip_rsvp_vif(struct socket *so __unused, struct sockopt *sopt __unused)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (EOPNOTSUPP);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
static void
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
X_ip_rsvp_force_done(struct socket *so __unused)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
}
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
static int
X_rsvp_input(struct mbuf **mp, int *offp, int proto)
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
{
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
struct mbuf *m;
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
m = *mp;
*mp = NULL;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
if (!V_rsvp_on)
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
Kernel side of 3.5 multicast routing code, based on work by Bill Fenner and other work done here. The LKM support is probably broken, but it still compiles and will be fixed later.
1995-06-13 17:51:16 +00:00
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Code for bandwidth monitors
*/
/*
* Define common interface for timeval-related methods
*/
#define BW_TIMEVALCMP(tvp, uvp, cmp) timevalcmp((tvp), (uvp), cmp)
#define BW_TIMEVALDECR(vvp, uvp) timevalsub((vvp), (uvp))
#define BW_TIMEVALADD(vvp, uvp) timevaladd((vvp), (uvp))
static uint32_t
compute_bw_meter_flags(struct bw_upcall *req)
{
uint32_t flags = 0;
if (req->bu_flags & BW_UPCALL_UNIT_PACKETS)
flags |= BW_METER_UNIT_PACKETS;
if (req->bu_flags & BW_UPCALL_UNIT_BYTES)
flags |= BW_METER_UNIT_BYTES;
if (req->bu_flags & BW_UPCALL_GEQ)
flags |= BW_METER_GEQ;
if (req->bu_flags & BW_UPCALL_LEQ)
flags |= BW_METER_LEQ;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return flags;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Add a bw_meter entry
*/
static int
add_bw_upcall(struct bw_upcall *req)
{
struct mfc *mfc;
struct timeval delta = { BW_UPCALL_THRESHOLD_INTERVAL_MIN_SEC,
BW_UPCALL_THRESHOLD_INTERVAL_MIN_USEC };
struct timeval now;
struct bw_meter *x;
uint32_t flags;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (!(V_mrt_api_config & MRT_MFC_BW_UPCALL))
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EOPNOTSUPP;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Test if the flags are valid */
if (!(req->bu_flags & (BW_UPCALL_UNIT_PACKETS | BW_UPCALL_UNIT_BYTES)))
return EINVAL;
if (!(req->bu_flags & (BW_UPCALL_GEQ | BW_UPCALL_LEQ)))
return EINVAL;
if ((req->bu_flags & (BW_UPCALL_GEQ | BW_UPCALL_LEQ))
== (BW_UPCALL_GEQ | BW_UPCALL_LEQ))
return EINVAL;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Test if the threshold time interval is valid */
if (BW_TIMEVALCMP(&req->bu_threshold.b_time, &delta, <))
return EINVAL;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
flags = compute_bw_meter_flags(req);
/*
* Find if we have already same bw_meter entry
*/
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
mfc = mfc_find(&req->bu_src, &req->bu_dst);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mfc == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EADDRNOTAVAIL;
}
for (x = mfc->mfc_bw_meter; x != NULL; x = x->bm_mfc_next) {
if ((BW_TIMEVALCMP(&x->bm_threshold.b_time,
&req->bu_threshold.b_time, ==)) &&
(x->bm_threshold.b_packets == req->bu_threshold.b_packets) &&
(x->bm_threshold.b_bytes == req->bu_threshold.b_bytes) &&
(x->bm_flags & BW_METER_USER_FLAGS) == flags) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0; /* XXX Already installed */
}
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Allocate the new bw_meter entry */
x = (struct bw_meter *)malloc(sizeof(*x), M_BWMETER, M_NOWAIT);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
if (x == NULL) {
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return ENOBUFS;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Set the new bw_meter entry */
x->bm_threshold.b_time = req->bu_threshold.b_time;
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
microtime(&now);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
x->bm_start_time = now;
x->bm_threshold.b_packets = req->bu_threshold.b_packets;
x->bm_threshold.b_bytes = req->bu_threshold.b_bytes;
x->bm_measured.b_packets = 0;
x->bm_measured.b_bytes = 0;
x->bm_flags = flags;
x->bm_time_next = NULL;
x->bm_time_hash = BW_METER_BUCKETS;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Add the new bw_meter entry to the front of entries for this MFC */
x->bm_mfc = mfc;
x->bm_mfc_next = mfc->mfc_bw_meter;
mfc->mfc_bw_meter = x;
schedule_bw_meter(x, &now);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
}
static void
free_bw_list(struct bw_meter *list)
{
while (list != NULL) {
struct bw_meter *x = list;
list = list->bm_mfc_next;
unschedule_bw_meter(x);
free(x, M_BWMETER);
}
}
/*
* Delete one or multiple bw_meter entries
*/
static int
del_bw_upcall(struct bw_upcall *req)
{
struct mfc *mfc;
struct bw_meter *x;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (!(V_mrt_api_config & MRT_MFC_BW_UPCALL))
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EOPNOTSUPP;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Find the corresponding MFC entry */
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
mfc = mfc_find(&req->bu_src, &req->bu_dst);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mfc == NULL) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EADDRNOTAVAIL;
} else if (req->bu_flags & BW_UPCALL_DELETE_ALL) {
/*
* Delete all bw_meter entries for this mfc
*/
struct bw_meter *list;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
list = mfc->mfc_bw_meter;
mfc->mfc_bw_meter = NULL;
free_bw_list(list);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
} else { /* Delete a single bw_meter entry */
struct bw_meter *prev;
uint32_t flags = 0;
flags = compute_bw_meter_flags(req);
/* Find the bw_meter entry to delete */
for (prev = NULL, x = mfc->mfc_bw_meter; x != NULL;
Fix bug with tracking the previous element in a list. Found by: edrt@citiz.net Submitted by: pavlin@icir.org
2004-08-03 02:01:44 +00:00
prev = x, x = x->bm_mfc_next) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if ((BW_TIMEVALCMP(&x->bm_threshold.b_time,
&req->bu_threshold.b_time, ==)) &&
(x->bm_threshold.b_packets == req->bu_threshold.b_packets) &&
(x->bm_threshold.b_bytes == req->bu_threshold.b_bytes) &&
(x->bm_flags & BW_METER_USER_FLAGS) == flags)
break;
}
if (x != NULL) { /* Delete entry from the list for this MFC */
if (prev != NULL)
prev->bm_mfc_next = x->bm_mfc_next; /* remove from middle*/
else
x->bm_mfc->mfc_bw_meter = x->bm_mfc_next;/* new head of list */
unschedule_bw_meter(x);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Free the bw_meter entry */
free(x, M_BWMETER);
return 0;
} else {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return EINVAL;
}
}
/* NOTREACHED */
}
/*
* Perform bandwidth measurement processing that may result in an upcall
*/
static void
bw_meter_receive_packet(struct bw_meter *x, int plen, struct timeval *nowp)
{
struct timeval delta;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
delta = *nowp;
BW_TIMEVALDECR(&delta, &x->bm_start_time);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (x->bm_flags & BW_METER_GEQ) {
/*
* Processing for ">=" type of bw_meter entry
*/
if (BW_TIMEVALCMP(&delta, &x->bm_threshold.b_time, >)) {
/* Reset the bw_meter entry */
x->bm_start_time = *nowp;
x->bm_measured.b_packets = 0;
x->bm_measured.b_bytes = 0;
x->bm_flags &= ~BW_METER_UPCALL_DELIVERED;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Record that a packet is received */
x->bm_measured.b_packets++;
x->bm_measured.b_bytes += plen;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Test if we should deliver an upcall
*/
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
if (!(x->bm_flags & BW_METER_UPCALL_DELIVERED)) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (((x->bm_flags & BW_METER_UNIT_PACKETS) &&
(x->bm_measured.b_packets >= x->bm_threshold.b_packets)) ||
((x->bm_flags & BW_METER_UNIT_BYTES) &&
(x->bm_measured.b_bytes >= x->bm_threshold.b_bytes))) {
/* Prepare an upcall for delivery */
bw_meter_prepare_upcall(x, nowp);
x->bm_flags |= BW_METER_UPCALL_DELIVERED;
}
}
} else if (x->bm_flags & BW_METER_LEQ) {
/*
* Processing for "<=" type of bw_meter entry
*/
if (BW_TIMEVALCMP(&delta, &x->bm_threshold.b_time, >)) {
/*
* We are behind time with the multicast forwarding table
* scanning for "<=" type of bw_meter entries, so test now
* if we should deliver an upcall.
*/
if (((x->bm_flags & BW_METER_UNIT_PACKETS) &&
(x->bm_measured.b_packets <= x->bm_threshold.b_packets)) ||
((x->bm_flags & BW_METER_UNIT_BYTES) &&
(x->bm_measured.b_bytes <= x->bm_threshold.b_bytes))) {
/* Prepare an upcall for delivery */
bw_meter_prepare_upcall(x, nowp);
}
/* Reschedule the bw_meter entry */
unschedule_bw_meter(x);
schedule_bw_meter(x, nowp);
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Record that a packet is received */
x->bm_measured.b_packets++;
x->bm_measured.b_bytes += plen;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Test if we should restart the measuring interval
*/
if ((x->bm_flags & BW_METER_UNIT_PACKETS &&
x->bm_measured.b_packets <= x->bm_threshold.b_packets) ||
(x->bm_flags & BW_METER_UNIT_BYTES &&
x->bm_measured.b_bytes <= x->bm_threshold.b_bytes)) {
/* Don't restart the measuring interval */
} else {
/* Do restart the measuring interval */
/*
* XXX: note that we don't unschedule and schedule, because this
* might be too much overhead per packet. Instead, when we process
* all entries for a given timer hash bin, we check whether it is
* really a timeout. If not, we reschedule at that time.
*/
x->bm_start_time = *nowp;
x->bm_measured.b_packets = 0;
x->bm_measured.b_bytes = 0;
x->bm_flags &= ~BW_METER_UPCALL_DELIVERED;
}
}
}
/*
* Prepare a bandwidth-related upcall
*/
static void
bw_meter_prepare_upcall(struct bw_meter *x, struct timeval *nowp)
{
struct timeval delta;
struct bw_upcall *u;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
* Compute the measured time interval
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
*/
delta = *nowp;
BW_TIMEVALDECR(&delta, &x->bm_start_time);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* If there are too many pending upcalls, deliver them now
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_bw_upcalls_n >= BW_UPCALLS_MAX)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
bw_upcalls_send();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Set the bw_upcall entry
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
u = &V_bw_upcalls[V_bw_upcalls_n++];
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
u->bu_src = x->bm_mfc->mfc_origin;
u->bu_dst = x->bm_mfc->mfc_mcastgrp;
u->bu_threshold.b_time = x->bm_threshold.b_time;
u->bu_threshold.b_packets = x->bm_threshold.b_packets;
u->bu_threshold.b_bytes = x->bm_threshold.b_bytes;
u->bu_measured.b_time = delta;
u->bu_measured.b_packets = x->bm_measured.b_packets;
u->bu_measured.b_bytes = x->bm_measured.b_bytes;
u->bu_flags = 0;
if (x->bm_flags & BW_METER_UNIT_PACKETS)
u->bu_flags |= BW_UPCALL_UNIT_PACKETS;
if (x->bm_flags & BW_METER_UNIT_BYTES)
u->bu_flags |= BW_UPCALL_UNIT_BYTES;
if (x->bm_flags & BW_METER_GEQ)
u->bu_flags |= BW_UPCALL_GEQ;
if (x->bm_flags & BW_METER_LEQ)
u->bu_flags |= BW_UPCALL_LEQ;
}
/*
* Send the pending bandwidth-related upcalls
*/
static void
bw_upcalls_send(void)
{
struct mbuf *m;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
int len = V_bw_upcalls_n * sizeof(V_bw_upcalls[0]);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
struct sockaddr_in k_igmpsrc = { sizeof k_igmpsrc, AF_INET };
static struct igmpmsg igmpmsg = { 0, /* unused1 */
0, /* unused2 */
IGMPMSG_BW_UPCALL,/* im_msgtype */
0, /* im_mbz */
0, /* im_vif */
0, /* unused3 */
{ 0 }, /* im_src */
{ 0 } }; /* im_dst */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_bw_upcalls_n == 0)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return; /* No pending upcalls */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_bw_upcalls_n = 0;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Allocate a new mbuf, initialize it with the header and
* the payload for the pending calls.
*/
Use m_get/m_gethdr instead of compat macros. Sponsored by: Nginx, Inc.
2013-03-15 12:55:30 +00:00
m = m_gethdr(M_NOWAIT, MT_DATA);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (m == NULL) {
log(LOG_WARNING, "bw_upcalls_send: cannot allocate mbuf\n");
return;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_copyback(m, 0, sizeof(struct igmpmsg), (caddr_t)&igmpmsg);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
m_copyback(m, sizeof(struct igmpmsg), len, (caddr_t)&V_bw_upcalls[0]);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Send the upcalls
* XXX do we need to set the address in k_igmpsrc ?
*/
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upcalls);
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (socket_send(V_ip_mrouter, m, &k_igmpsrc) < 0) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
log(LOG_WARNING, "bw_upcalls_send: ip_mrouter socket queue full\n");
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upq_sockfull);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
}
/*
* Compute the timeout hash value for the bw_meter entries
*/
#define BW_METER_TIMEHASH(bw_meter, hash) \
do { \
struct timeval next_timeval = (bw_meter)->bm_start_time; \
\
BW_TIMEVALADD(&next_timeval, &(bw_meter)->bm_threshold.b_time); \
(hash) = next_timeval.tv_sec; \
if (next_timeval.tv_usec) \
(hash)++; /* XXX: make sure we don't timeout early */ \
(hash) %= BW_METER_BUCKETS; \
} while (0)
/*
* Schedule a timer to process periodically bw_meter entry of type "<="
* by linking the entry in the proper hash bucket.
*/
static void
schedule_bw_meter(struct bw_meter *x, struct timeval *nowp)
{
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
int time_hash;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (!(x->bm_flags & BW_METER_LEQ))
return; /* XXX: we schedule timers only for "<=" entries */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Reset the bw_meter entry
*/
x->bm_start_time = *nowp;
x->bm_measured.b_packets = 0;
x->bm_measured.b_bytes = 0;
x->bm_flags &= ~BW_METER_UPCALL_DELIVERED;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Compute the timeout hash value and insert the entry
*/
BW_METER_TIMEHASH(x, time_hash);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
x->bm_time_next = V_bw_meter_timers[time_hash];
V_bw_meter_timers[time_hash] = x;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
x->bm_time_hash = time_hash;
}
/*
* Unschedule the periodic timer that processes bw_meter entry of type "<="
* by removing the entry from the proper hash bucket.
*/
static void
unschedule_bw_meter(struct bw_meter *x)
{
int time_hash;
struct bw_meter *prev, *tmp;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK_ASSERT();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (!(x->bm_flags & BW_METER_LEQ))
return; /* XXX: we schedule timers only for "<=" entries */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Compute the timeout hash value and delete the entry
*/
time_hash = x->bm_time_hash;
if (time_hash >= BW_METER_BUCKETS)
return; /* Entry was not scheduled */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
for (prev = NULL, tmp = V_bw_meter_timers[time_hash];
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
tmp != NULL; prev = tmp, tmp = tmp->bm_time_next)
if (tmp == x)
break;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (tmp == NULL)
panic("unschedule_bw_meter: bw_meter entry not found");
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (prev != NULL)
prev->bm_time_next = x->bm_time_next;
else
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_bw_meter_timers[time_hash] = x->bm_time_next;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
x->bm_time_next = NULL;
x->bm_time_hash = BW_METER_BUCKETS;
}
/*
* Process all "<=" type of bw_meter that should be processed now,
* and for each entry prepare an upcall if necessary. Each processed
* entry is rescheduled again for the (periodic) processing.
*
* This is run periodically (once per second normally). On each round,
* all the potentially matching entries are in the hash slot that we are
* looking at.
*/
static void
bw_meter_process()
{
uint32_t loops;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
int i;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
struct timeval now, process_endtime;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
microtime(&now);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_last_tv_sec == now.tv_sec)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return; /* nothing to do */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
loops = now.tv_sec - V_last_tv_sec;
V_last_tv_sec = now.tv_sec;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (loops > BW_METER_BUCKETS)
loops = BW_METER_BUCKETS;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Process all bins of bw_meter entries from the one after the last
* processed to the current one. On entry, i points to the last bucket
* visited, so we need to increment i at the beginning of the loop.
*/
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
for (i = (now.tv_sec - loops) % BW_METER_BUCKETS; loops > 0; loops--) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
struct bw_meter *x, *tmp_list;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (++i >= BW_METER_BUCKETS)
i = 0;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
/* Disconnect the list of bw_meter entries from the bin */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
tmp_list = V_bw_meter_timers[i];
V_bw_meter_timers[i] = NULL;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
/* Process the list of bw_meter entries */
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
while (tmp_list != NULL) {
x = tmp_list;
tmp_list = tmp_list->bm_time_next;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Test if the time interval is over */
process_endtime = x->bm_start_time;
BW_TIMEVALADD(&process_endtime, &x->bm_threshold.b_time);
if (BW_TIMEVALCMP(&process_endtime, &now, >)) {
/* Not yet: reschedule, but don't reset */
int time_hash;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
BW_METER_TIMEHASH(x, time_hash);
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
if (time_hash == i && process_endtime.tv_sec == now.tv_sec) {
/*
* XXX: somehow the bin processing is a bit ahead of time.
* Put the entry in the next bin.
*/
if (++time_hash >= BW_METER_BUCKETS)
time_hash = 0;
}
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
x->bm_time_next = V_bw_meter_timers[time_hash];
V_bw_meter_timers[time_hash] = x;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
x->bm_time_hash = time_hash;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
continue;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Test if we should deliver an upcall
*/
if (((x->bm_flags & BW_METER_UNIT_PACKETS) &&
(x->bm_measured.b_packets <= x->bm_threshold.b_packets)) ||
((x->bm_flags & BW_METER_UNIT_BYTES) &&
(x->bm_measured.b_bytes <= x->bm_threshold.b_bytes))) {
/* Prepare an upcall for delivery */
bw_meter_prepare_upcall(x, &now);
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Reschedule for next processing
*/
schedule_bw_meter(x, &now);
}
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Send all upcalls that are pending delivery */
bw_upcalls_send();
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
/*
* A periodic function for sending all upcalls that are pending delivery
*/
static void
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
expire_bw_upcalls_send(void *arg)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
CURVNET_SET((struct vnet *) arg);
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_LOCK();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
bw_upcalls_send();
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
MFC_UNLOCK();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
callout_reset(&V_bw_upcalls_ch, BW_UPCALLS_PERIOD, expire_bw_upcalls_send,
curvnet);
CURVNET_RESTORE();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
/*
* A periodic function for periodic scanning of the multicast forwarding
* table for processing all "<=" bw_meter entries.
*/
static void
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
expire_bw_meter_process(void *arg)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
CURVNET_SET((struct vnet *) arg);
if (V_mrt_api_config & MRT_MFC_BW_UPCALL)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
bw_meter_process();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
callout_reset(&V_bw_meter_ch, BW_METER_PERIOD, expire_bw_meter_process,
curvnet);
CURVNET_RESTORE();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
/*
* End of bandwidth monitoring code
*/
/*
* Send the packet up to the user daemon, or eventually do kernel encapsulation
*
*/
static int
Move universally to ANSI C function declarations, with relatively consistent style(9)-ish layout.
2007-05-10 15:58:48 +00:00
pim_register_send(struct ip *ip, struct vif *vifp, struct mbuf *m,
struct mfc *rt)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
struct mbuf *mb_copy, *mm;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
If the rendezvous point for a group is not specified, do not send IGMPMSG_WHOLEPKT notifications to the userland PIM routing daemon, as an optimization to mitigate the effects of high multicast forwarding load. This is an experimental change, therefore it must be explicitly enabled by setting the sysctl/tunable net.inet.pim.squelch_wholepkt to a non-zero value. The tunable may be set from the loader or from within the kernel environment when loading ip_mroute.ko as a module. Submitted by: edrt <edrt at citiz.net> See also: http://mailman.icsi.berkeley.edu/pipermail/xorp-users/2005-June/000639.html
2007-02-10 14:48:42 +00:00
/*
* Do not send IGMP_WHOLEPKT notifications to userland, if the
* rendezvous point was unspecified, and we were told not to.
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (pim_squelch_wholepkt != 0 && (V_mrt_api_config & MRT_MFC_RP) &&
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
in_nullhost(rt->mfc_rp))
If the rendezvous point for a group is not specified, do not send IGMPMSG_WHOLEPKT notifications to the userland PIM routing daemon, as an optimization to mitigate the effects of high multicast forwarding load. This is an experimental change, therefore it must be explicitly enabled by setting the sysctl/tunable net.inet.pim.squelch_wholepkt to a non-zero value. The tunable may be set from the loader or from within the kernel environment when loading ip_mroute.ko as a module. Submitted by: edrt <edrt at citiz.net> See also: http://mailman.icsi.berkeley.edu/pipermail/xorp-users/2005-June/000639.html
2007-02-10 14:48:42 +00:00
return 0;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
mb_copy = pim_register_prepare(ip, m);
if (mb_copy == NULL)
return ENOBUFS;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Send all the fragments. Note that the mbuf for each fragment
* is freed by the sending machinery.
*/
for (mm = mb_copy; mm; mm = mb_copy) {
mb_copy = mm->m_nextpkt;
mm->m_nextpkt = 0;
mm = m_pullup(mm, sizeof(struct ip));
if (mm != NULL) {
ip = mtod(mm, struct ip *);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if ((V_mrt_api_config & MRT_MFC_RP) && !in_nullhost(rt->mfc_rp)) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
pim_register_send_rp(ip, vifp, mm, rt);
} else {
pim_register_send_upcall(ip, vifp, mm, rt);
}
}
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
}
/*
* Return a copy of the data packet that is ready for PIM Register
* encapsulation.
* XXX: Note that in the returned copy the IP header is a valid one.
*/
static struct mbuf *
pim_register_prepare(struct ip *ip, struct mbuf *m)
{
struct mbuf *mb_copy = NULL;
int mtu;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
/* Take care of delayed checksums */
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (m->m_pkthdr.csum_flags & CSUM_DELAY_DATA) {
in_delayed_cksum(m);
m->m_pkthdr.csum_flags &= ~CSUM_DELAY_DATA;
}
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Copy the old packet & pullup its IP header into the
* new mbuf so we can modify it.
*/
Mechanically substitute flags from historic mbuf allocator with malloc(9) flags within sys. Exceptions: - sys/contrib not touched - sys/mbuf.h edited manually
2012-12-05 08:04:20 +00:00
mb_copy = m_copypacket(m, M_NOWAIT);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mb_copy == NULL)
return NULL;
mb_copy = m_pullup(mb_copy, ip->ip_hl << 2);
if (mb_copy == NULL)
return NULL;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* take care of the TTL */
ip = mtod(mb_copy, struct ip *);
--ip->ip_ttl;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Compute the MTU after the PIM Register encapsulation */
mtu = 0xffff - sizeof(pim_encap_iphdr) - sizeof(pim_encap_pimhdr);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Switch the entire IPv4 stack to keep the IP packet header in network byte order. Any host byte order processing is done in local variables and host byte order values are never[1] written to a packet. After this change a packet processed by the stack isn't modified at all[2] except for TTL. After this change a network stack hacker doesn't need to scratch his head trying to figure out what is the byte order at the given place in the stack. [1] One exception still remains. The raw sockets convert host byte order before pass a packet to an application. Probably this would remain for ages for compatibility. [2] The ip_input() still subtructs header len from ip->ip_len, but this is planned to be fixed soon. Reviewed by: luigi, Maxim Dounin <mdounin mdounin.ru> Tested by: ray, Olivier Cochard-Labbe <olivier cochard.me>
2012-10-22 21:09:03 +00:00
if (ntohs(ip->ip_len) <= mtu) {
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
/* Turn the IP header into a valid one */
ip->ip_sum = 0;
ip->ip_sum = in_cksum(mb_copy, ip->ip_hl << 2);
} else {
/* Fragment the packet */
o Remove last argument to ip_fragment(), and obtain all needed information on checksums directly from mbuf flags. This simplifies code. o Clear CSUM_IP from the mbuf in ip_fragment() if we did checksums in hardware. Some driver may not announce CSUM_IP in theur if_hwassist, although try to do checksums if CSUM_IP set on mbuf. Example is em(4). o While here, consistently use CSUM_IP instead of its alias CSUM_DELAY_IP. After this change CSUM_DELAY_IP vanishes from the stack. Submitted by: Sebastian Kuzminsky <seb lineratesystems.com>
2012-10-26 21:06:33 +00:00
mb_copy->m_pkthdr.csum_flags |= CSUM_IP;
if (ip_fragment(ip, &mb_copy, mtu, 0) != 0) {
* Bug fix in bw_meter_process(): the periodically processed bins of bw_meter entries were processed up to one second ahead. After an unappropriate rescheduling of some of the bw_meter entries, the upcalls weren't delivered. * pim_register_prepare() uses the appropriate sw_csum flag to call ip_fragment() so the IP checksum is computed properly. * Modify pim_register_prepare() to take care of IP packets that don't need fragmentation. * Add-back in_delayed_cksum() to encap_send(), because it seems it should be there. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-19 17:22:51 +00:00
m_freem(mb_copy);
return NULL;
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
return mb_copy;
}
/*
* Send an upcall with the data packet to the user-level process.
*/
static int
pim_register_send_upcall(struct ip *ip, struct vif *vifp,
Move universally to ANSI C function declarations, with relatively consistent style(9)-ish layout.
2007-05-10 15:58:48 +00:00
struct mbuf *mb_copy, struct mfc *rt)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
struct mbuf *mb_first;
int len = ntohs(ip->ip_len);
struct igmpmsg *im;
struct sockaddr_in k_igmpsrc = { sizeof k_igmpsrc, AF_INET };
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK_ASSERT();
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Add a new mbuf with an upcall header
*/
Use m_get/m_gethdr instead of compat macros. Sponsored by: Nginx, Inc.
2013-03-15 12:55:30 +00:00
mb_first = m_gethdr(M_NOWAIT, MT_DATA);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mb_first == NULL) {
m_freem(mb_copy);
return ENOBUFS;
}
mb_first->m_data += max_linkhdr;
mb_first->m_pkthdr.len = len + sizeof(struct igmpmsg);
mb_first->m_len = sizeof(struct igmpmsg);
mb_first->m_next = mb_copy;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Send message to routing daemon */
im = mtod(mb_first, struct igmpmsg *);
im->im_msgtype = IGMPMSG_WHOLEPKT;
im->im_mbz = 0;
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
im->im_vif = vifp - V_viftable;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
im->im_src = ip->ip_src;
im->im_dst = ip->ip_dst;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
k_igmpsrc.sin_addr = ip->ip_src;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upcalls);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
if (socket_send(V_ip_mrouter, mb_first, &k_igmpsrc) < 0) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: socket queue full", __func__);
Update stats in struct mrtstat using two new macros: MRTSTAT_ADD() and MRTSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:00:36 +00:00
MRTSTAT_INC(mrts_upq_sockfull);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return ENOBUFS;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Keep statistics */
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_snd_registers_msgs);
PIMSTAT_ADD(pims_snd_registers_bytes, len);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
}
/*
* Encapsulate the data packet in PIM Register message and send it to the RP.
*/
static int
Move universally to ANSI C function declarations, with relatively consistent style(9)-ish layout.
2007-05-10 15:58:48 +00:00
pim_register_send_rp(struct ip *ip, struct vif *vifp, struct mbuf *mb_copy,
struct mfc *rt)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
struct mbuf *mb_first;
struct ip *ip_outer;
struct pim_encap_pimhdr *pimhdr;
int len = ntohs(ip->ip_len);
vifi_t vifi = rt->mfc_parent;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK_ASSERT();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if ((vifi >= V_numvifs) || in_nullhost(V_viftable[vifi].v_lcl_addr)) {
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(mb_copy);
return EADDRNOTAVAIL; /* The iif vif is invalid */
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Add a new mbuf with the encapsulating header
*/
Use m_get/m_gethdr instead of compat macros. Sponsored by: Nginx, Inc.
2013-03-15 12:55:30 +00:00
mb_first = m_gethdr(M_NOWAIT, MT_DATA);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mb_first == NULL) {
m_freem(mb_copy);
return ENOBUFS;
}
mb_first->m_data += max_linkhdr;
mb_first->m_len = sizeof(pim_encap_iphdr) + sizeof(pim_encap_pimhdr);
mb_first->m_next = mb_copy;
mb_first->m_pkthdr.len = len + mb_first->m_len;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Fill in the encapsulating IP and PIM header
*/
ip_outer = mtod(mb_first, struct ip *);
*ip_outer = pim_encap_iphdr;
Switch the entire IPv4 stack to keep the IP packet header in network byte order. Any host byte order processing is done in local variables and host byte order values are never[1] written to a packet. After this change a packet processed by the stack isn't modified at all[2] except for TTL. After this change a network stack hacker doesn't need to scratch his head trying to figure out what is the byte order at the given place in the stack. [1] One exception still remains. The raw sockets convert host byte order before pass a packet to an application. Probably this would remain for ages for compatibility. [2] The ip_input() still subtructs header len from ip->ip_len, but this is planned to be fixed soon. Reviewed by: luigi, Maxim Dounin <mdounin mdounin.ru> Tested by: ray, Olivier Cochard-Labbe <olivier cochard.me>
2012-10-22 21:09:03 +00:00
ip_outer->ip_len = htons(len + sizeof(pim_encap_iphdr) +
sizeof(pim_encap_pimhdr));
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
ip_outer->ip_src = V_viftable[vifi].v_lcl_addr;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
ip_outer->ip_dst = rt->mfc_rp;
/*
* Copy the inner header TOS to the outer header, and take care of the
* IP_DF bit.
*/
ip_outer->ip_tos = ip->ip_tos;
Switch the entire IPv4 stack to keep the IP packet header in network byte order. Any host byte order processing is done in local variables and host byte order values are never[1] written to a packet. After this change a packet processed by the stack isn't modified at all[2] except for TTL. After this change a network stack hacker doesn't need to scratch his head trying to figure out what is the byte order at the given place in the stack. [1] One exception still remains. The raw sockets convert host byte order before pass a packet to an application. Probably this would remain for ages for compatibility. [2] The ip_input() still subtructs header len from ip->ip_len, but this is planned to be fixed soon. Reviewed by: luigi, Maxim Dounin <mdounin mdounin.ru> Tested by: ray, Olivier Cochard-Labbe <olivier cochard.me>
2012-10-22 21:09:03 +00:00
if (ip->ip_off & htons(IP_DF))
ip_outer->ip_off |= htons(IP_DF);
o Use new function ip_fillid() in all places throughout the kernel, where we want to create a new IP datagram. o Add support for RFC6864, which allows to set IP ID for atomic IP datagrams to any value, to improve performance. The behaviour is controlled by net.inet.ip.rfc6864 sysctl knob, which is enabled by default. o In case if we generate IP ID, use counter(9) to improve performance. o Gather all code related to IP ID into ip_id.c. Differential Revision: https://reviews.freebsd.org/D2177 Reviewed by: adrian, cy, rpaulo Tested by: Emeric POUPON <emeric.poupon stormshield.eu> Sponsored by: Netflix Sponsored by: Nginx, Inc. Relnotes: yes
2015-04-01 22:26:39 +00:00
ip_fillid(ip_outer);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
pimhdr = (struct pim_encap_pimhdr *)((caddr_t)ip_outer
+ sizeof(pim_encap_iphdr));
*pimhdr = pim_encap_pimhdr;
/* If the iif crosses a border, set the Border-bit */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (rt->mfc_flags[vifi] & MRT_MFC_FLAGS_BORDER_VIF & V_mrt_api_config)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
pimhdr->flags |= htonl(PIM_BORDER_REGISTER);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
mb_first->m_data += sizeof(pim_encap_iphdr);
pimhdr->pim.pim_cksum = in_cksum(mb_first, sizeof(pim_encap_pimhdr));
mb_first->m_data -= sizeof(pim_encap_iphdr);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Nuke the token bucket filter code. Attempting to request rate limiting by the token bucket filter will result in EINVAL being returned. If you want to rate-limit traffic in future, use ALTQ or dummynet; this isn't a general purpose QoS engine. Preserve the now unused fields in struct vif so as to avoid having to recompile netstat(1) and other tools. Reviewed by: Pavlin Radslavov, Bill Fenner
2007-02-08 22:58:01 +00:00
send_packet(vifp, mb_first);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Keep statistics */
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_snd_registers_msgs);
PIMSTAT_ADD(pims_snd_registers_bytes, len);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
return 0;
}
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
/*
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
* pim_encapcheck() is called by the encap4_input() path at runtime to
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
* determine if a packet is for PIM; allowing PIM to be dynamically loaded
* into the kernel.
*/
static int
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
pim_encapcheck(const struct mbuf *m __unused, int off __unused,
int proto __unused, void *arg __unused)
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
{
KASSERT(proto == IPPROTO_PIM, ("not for IPPROTO_PIM"));
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
return (8); /* claim the datagram. */
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
}
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* PIM-SMv2 and PIM-DM messages processing.
* Receives and verifies the PIM control messages, and passes them
* up to the listening socket, using rip_input().
* The only message with special processing is the PIM_REGISTER message
* (used by PIM-SM): the PIM header is stripped off, and the inner packet
* is passed to if_simloop().
*/
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
static int
pim_input(struct mbuf *m, int off, int proto, void *arg __unused)
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
{
struct ip *ip = mtod(m, struct ip *);
struct pim *pim;
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
int iphlen = off;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
int minlen;
Fix fallout from r241923. Calculate length of payload in pim_input() properly. While here, remove extra variable and incorrect condition before m_pullup(). Reported by: Olivier Cochard-Labbé <olivier cochard.me> Sponsored by: Nginx, Inc.
2014-01-22 10:57:42 +00:00
int datalen = ntohs(ip->ip_len) - iphlen;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
int ip_tos;
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Keep statistics */
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_total_msgs);
PIMSTAT_ADD(pims_rcv_total_bytes, datalen);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Validate lengths
*/
if (datalen < PIM_MINLEN) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_tooshort);
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR3(KTR_IPMF, "%s: short packet (%d) from 0x%08x",
__func__, datalen, ntohl(ip->ip_src.s_addr));
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* If the packet is at least as big as a REGISTER, go agead
* and grab the PIM REGISTER header size, to avoid another
* possible m_pullup() later.
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
*
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* PIM_MINLEN == pimhdr + u_int32_t == 4 + 4 = 8
* PIM_REG_MINLEN == pimhdr + reghdr + encap_iphdr == 4 + 4 + 20 = 28
*/
minlen = iphlen + (datalen >= PIM_REG_MINLEN ? PIM_REG_MINLEN : PIM_MINLEN);
/*
* Get the IP and PIM headers in contiguous memory, and
* possibly the PIM REGISTER header.
*/
netinet: for pointers replace 0 with NULL. These are mostly cosmetical, no functional change. Found with devel/coccinelle. Reviewed by: ae. tuexen
2016-04-15 15:46:41 +00:00
if (m->m_len < minlen && (m = m_pullup(m, minlen)) == NULL) {
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: m_pullup() failed", __func__);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* m_pullup() may have given us a new mbuf so reset ip. */
ip = mtod(m, struct ip *);
ip_tos = ip->ip_tos;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* adjust mbuf to point to the PIM header */
m->m_data += iphlen;
m->m_len -= iphlen;
pim = mtod(m, struct pim *);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Validate checksum. If PIM REGISTER, exclude the data packet.
*
* XXX: some older PIMv2 implementations don't make this distinction,
* so for compatibility reason perform the checksum over part of the
* message, and if error, then over the whole message.
*/
if (PIM_VT_T(pim->pim_vt) == PIM_REGISTER && in_cksum(m, PIM_MINLEN) == 0) {
/* do nothing, checksum okay */
} else if (in_cksum(m, datalen)) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_badsum);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: invalid checksum", __func__);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
/* PIM version check */
if (PIM_VT_V(pim->pim_vt) < PIM_VERSION) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_badversion);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR3(KTR_IPMF, "%s: bad version %d expect %d", __func__,
(int)PIM_VT_V(pim->pim_vt), PIM_VERSION);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* restore mbuf back to the outer IP */
m->m_data -= iphlen;
m->m_len += iphlen;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (PIM_VT_T(pim->pim_vt) == PIM_REGISTER) {
/*
* Since this is a REGISTER, we'll make a copy of the register
* headers ip + pim + u_int32 + encap_ip, to be passed up to the
* routing daemon.
*/
struct sockaddr_in dst = { sizeof(dst), AF_INET };
struct mbuf *mcp;
struct ip *encap_ip;
u_int32_t *reghdr;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
struct ifnet *vifp;
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_LOCK();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if ((V_reg_vif_num >= V_numvifs) || (V_reg_vif_num == VIFI_INVALID)) {
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR2(KTR_IPMF, "%s: register vif not set: %d", __func__,
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
(int)V_reg_vif_num);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
/* XXX need refcnt? */
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
vifp = V_viftable[V_reg_vif_num].v_ifp;
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
VIF_UNLOCK();
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Validate length
*/
if (datalen < PIM_REG_MINLEN) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_tooshort);
PIMSTAT_INC(pims_rcv_badregisters);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: register packet size too small", __func__);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
reghdr = (u_int32_t *)(pim + 1);
encap_ip = (struct ip *)(reghdr + 1);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR3(KTR_IPMF, "%s: register: encap ip src 0x%08x len %d",
__func__, ntohl(encap_ip->ip_src.s_addr),
Use inet_ntoa_r() instead of inet_ntoa() throughout the kernel inet_ntoa() cannot be used safely in a multithreaded environment because it uses a static local buffer. Instead, use inet_ntoa_r() with a buffer on the caller's stack. Suggested by: glebius, emaste Reviewed by: gnn MFC after: 2 weeks Sponsored by: Dell EMC Differential Revision: https://reviews.freebsd.org/D9625
2017-02-16 20:47:41 +00:00
ntohs(encap_ip->ip_len));
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* verify the version number of the inner packet */
if (encap_ip->ip_v != IPVERSION) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_badregisters);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR1(KTR_IPMF, "%s: bad encap ip version", __func__);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* verify the inner packet is destined to a mcast group */
if (!IN_MULTICAST(ntohl(encap_ip->ip_dst.s_addr))) {
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_badregisters);
KTR: log IPv4 addresses in hex rather than dotted-quad When I made the changes in r313821, I fell victim to one of the classic blunders, the most famous of which is: never get involved in a land war in Asia. But only slightly less well known is this: Keep your brain turned on and engaged when making a tedious, sweeping, mechanical change. KTR can correctly log the immediate integral values passed to it, as well as constant strings, but not non-constant strings, since they might change by the time ktrdump retrieves them. Reported by: glebius MFC after: 3 days Sponsored by: Dell EMC
2017-03-14 18:27:48 +00:00
CTR2(KTR_IPMF, "%s: bad encap ip dest 0x%08x", __func__,
ntohl(encap_ip->ip_dst.s_addr));
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
To comply with the spec, do not copy the TOS from the outer IP header to the inner IP header of the PIM Register if this is a PIM Null-Register message. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2004-03-08 07:47:27 +00:00
/* If a NULL_REGISTER, pass it to the daemon */
if ((ntohl(*reghdr) & PIM_NULL_REGISTER))
goto pim_input_to_daemon;
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Copy the TOS from the outer IP header to the inner IP header.
*/
if (encap_ip->ip_tos != ip_tos) {
/* Outer TOS -> inner TOS */
encap_ip->ip_tos = ip_tos;
/* Recompute the inner header checksum. Sigh... */
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* adjust mbuf to point to the inner IP header */
m->m_data += (iphlen + PIM_MINLEN);
m->m_len -= (iphlen + PIM_MINLEN);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
encap_ip->ip_sum = 0;
encap_ip->ip_sum = in_cksum(m, encap_ip->ip_hl << 2);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* restore mbuf to point back to the outer IP header */
m->m_data -= (iphlen + PIM_MINLEN);
m->m_len += (iphlen + PIM_MINLEN);
}
To comply with the spec, do not copy the TOS from the outer IP header to the inner IP header of the PIM Register if this is a PIM Null-Register message. Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2004-03-08 07:47:27 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Decapsulate the inner IP packet and loopback to forward it
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
* as a normal multicast packet. Also, make a copy of the
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
* outer_iphdr + pimhdr + reghdr + encap_iphdr
* to pass to the daemon later, so it can take the appropriate
* actions (e.g., send back PIM_REGISTER_STOP).
* XXX: here m->m_data points to the outer IP header.
*/
Remove the 4.3BSD compatible macro m_copy(), use m_copym() instead. Reviewed by: gnn Differential Revision: https://reviews.freebsd.org/D7878
2016-09-15 07:41:48 +00:00
mcp = m_copym(m, 0, iphlen + PIM_REG_MINLEN, M_NOWAIT);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
if (mcp == NULL) {
Remove the 4.3BSD compatible macro m_copy(), use m_copym() instead. Reviewed by: gnn Differential Revision: https://reviews.freebsd.org/D7878
2016-09-15 07:41:48 +00:00
CTR1(KTR_IPMF, "%s: m_copym() failed", __func__);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
m_freem(m);
Merge 'struct ip6protosw' and 'struct protosw' into one. Now we have only one protocol switch structure that is shared between ipv4 and ipv6. Phabric: D476 Reviewed by: jhb
2014-08-08 01:57:15 +00:00
return (IPPROTO_DONE);
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* Keep statistics */
/* XXX: registers_bytes include only the encap. mcast pkt */
Update stats in struct pimstat using two new macros: PIMSTAT_ADD() and PIMSTAT_INC(), rather than directly manipulating the fields of the structure. This will make it easier to change the implementation of these statistics, such as using per-CPU versions of the data structure. MFC after: 3 days
2009-04-12 14:06:26 +00:00
PIMSTAT_INC(pims_rcv_registers_msgs);
PIMSTAT_ADD(pims_rcv_registers_bytes, ntohs(encap_ip->ip_len));
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* forward the inner ip packet; point m_data at the inner ip.
*/
m_adj(m, iphlen + PIM_MINLEN);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
CTR4(KTR_IPMF,
"%s: forward decap'd REGISTER: src %lx dst %lx vif %d",
__func__,
(u_long)ntohl(encap_ip->ip_src.s_addr),
(u_long)ntohl(encap_ip->ip_dst.s_addr),
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
(int)V_reg_vif_num);
Cleanup: Nuke debug.mrtdebug, and replace it with KTR.
2009-03-19 14:14:21 +00:00
Add locking. Special thanks to Pavlin Radoslavov <pavlin@icir.org> for testing and fixing numerous problems. Sponsored by: FreeBSD Foundation Reviewed by: Pavlin Radoslavov <pavlin@icir.org>
2003-09-06 04:53:43 +00:00
/* NB: vifp was collected above; can it change on us? */
if_simloop(vifp, m, dst.sin_family, 0);
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/* prepare the register head to send to the mrouting daemon */
m = mcp;
}
White space cleanup for netinet before branch: - Trailing tab/space cleanup - Remove spurious spaces between or before tabs This change avoids touching files that Andre likely has in his working set for PFIL hooks changes for IPFW/DUMMYNET. Approved by: re (scottl) Submitted by: Xin LI <delphij@frontfree.net>
2004-08-16 18:32:07 +00:00
pim_input_to_daemon:
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
/*
* Pass the PIM message up to the daemon; if it is a Register message,
* pass the 'head' only up to the daemon. This includes the
* outer IP header, PIM header, PIM-Register header and the
* inner IP header.
* XXX: the outer IP header pkt size of a Register is not adjust to
* reflect the fact that the inner multicast data is truncated.
*/
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
return (rip_input(&m, &off, proto));
1. Basic PIM kernel support Disabled by default. To enable it, the new "options PIM" must be added to the kernel configuration file (in addition to MROUTING): options MROUTING # Multicast routing options PIM # Protocol Independent Multicast 2. Add support for advanced multicast API setup/configuration and extensibility. 3. Add support for kernel-level PIM Register encapsulation. Disabled by default. Can be enabled by the advanced multicast API. 4. Implement a mechanism for "multicast bandwidth monitoring and upcalls". Submitted by: Pavlin Radoslavov <pavlin@icir.org>
2003-08-07 18:16:59 +00:00
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
static int
sysctl_mfctable(SYSCTL_HANDLER_ARGS)
{
struct mfc *rt;
int error, i;
if (req->newptr)
return (EPERM);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if (V_mfchashtbl == NULL) /* XXX unlocked */
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (0);
error = sysctl_wire_old_buffer(req, 0);
if (error)
return (error);
MFC_LOCK();
for (i = 0; i < mfchashsize; i++) {
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
LIST_FOREACH(rt, &V_mfchashtbl[i], mfc_hash) {
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
error = SYSCTL_OUT(req, rt, sizeof(struct mfc));
if (error)
goto out_locked;
}
}
out_locked:
MFC_UNLOCK();
return (error);
}
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_inet_ip, OID_AUTO, mfctable,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_mfctable,
"IPv4 Multicast Forwarding Table "
Mark all SYSCTL_NODEs static that have no corresponding SYSCTL_DECLs. The SYSCTL_NODE macro defines a list that stores all child-elements of that node. If there's no SYSCTL_DECL macro anywhere else, there's no reason why it shouldn't be static.
2011-11-07 15:43:11 +00:00
"(struct *mfc[mfchashsize], netinet/ip_mroute.h)");
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
ip_mroute: fix the viftable export sysctl It seems that in r354857 I got more than one thing wrong. Convert the SYSCTL_OPAQUE to a SYSCTL_PROC to properly export the these days allocated and not longer static per-vnet viftable array. This fixes a problem with netstat -g which would show bogus information for the IPv4 Virtual Interface Table. PR: 246626 Reported by: Ozkan KIRIK (ozkan.kirik gmail.com) MFC after: 3 days
2020-10-11 00:01:00 +00:00
static int
sysctl_viflist(SYSCTL_HANDLER_ARGS)
{
int error;
if (req->newptr)
return (EPERM);
if (V_viftable == NULL) /* XXX unlocked */
return (0);
error = sysctl_wire_old_buffer(req, sizeof(*V_viftable) * MAXVIFS);
if (error)
return (error);
VIF_LOCK();
error = SYSCTL_OUT(req, V_viftable, sizeof(*V_viftable) * MAXVIFS);
VIF_UNLOCK();
return (error);
}
SYSCTL_PROC(_net_inet_ip, OID_AUTO, viftable,
CTLTYPE_OPAQUE | CTLFLAG_VNET | CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, 0,
sysctl_viflist, "S,vif[MAXVIFS]",
"IPv4 Multicast Interfaces (struct vif[MAXVIFS], netinet/ip_mroute.h)");
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
static void
vnet_mroute_init(const void *unused __unused)
{
style: Remove remaining deprecated MALLOC/FREE macros Mechanically replace uses of MALLOC/FREE with appropriate invocations of malloc(9) / free(9) (a series of sed expressions). Something like: * MALLOC(a, b, ... -> a = malloc(... * FREE( -> free( * free((caddr_t) -> free( No functional change. For now, punt on modifying contrib ipfilter code, leaving a definition of the macro in its KMALLOC(). Reported by: jhb Reviewed by: cy, imp, markj, rmacklem Sponsored by: Dell EMC Isilon Differential Revision: https://reviews.freebsd.org/D14035
2018-01-25 22:25:13 +00:00
V_nexpire = malloc(mfchashsize, M_MRTABLE, M_WAITOK|M_ZERO);
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
V_viftable = mallocarray(MAXVIFS, sizeof(*V_viftable),
M_MRTABLE, M_WAITOK|M_ZERO);
V_bw_meter_timers = mallocarray(BW_METER_BUCKETS,
sizeof(*V_bw_meter_timers), M_MRTABLE, M_WAITOK|M_ZERO);
V_bw_upcalls = mallocarray(BW_UPCALLS_MAX, sizeof(*V_bw_upcalls),
M_MRTABLE, M_WAITOK|M_ZERO);
CALLOUT_MPSAFE has lost its meaning since r141428, i.e., for more than ten years for head. However, it is continuously misused as the mpsafe argument for callout_init(9). Deprecate the flag and clean up callout_init() calls to make them more consistent. Differential Revision: https://reviews.freebsd.org/D2613 Reviewed by: jhb MFC after: 2 weeks
2015-05-22 17:05:21 +00:00
callout_init(&V_expire_upcalls_ch, 1);
callout_init(&V_bw_upcalls_ch, 1);
callout_init(&V_bw_meter_ch, 1);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
}
Get closer to a VIMAGE network stack teardown from top to bottom rather than removing the network interfaces first. This change is rather larger and convoluted as the ordering requirements cannot be separated. Move the pfil(9) framework to SI_SUB_PROTO_PFIL, move Firewalls and related modules to their own SI_SUB_PROTO_FIREWALL. Move initialization of "physical" interfaces to SI_SUB_DRIVERS, move virtual (cloned) interfaces to SI_SUB_PSEUDO. Move Multicast to SI_SUB_PROTO_MC. Re-work parts of multicast initialisation and teardown, not taking the huge amount of memory into account if used as a module yet. For interface teardown we try to do as many of them as we can on SI_SUB_INIT_IF, but for some this makes no sense, e.g., when tunnelling over a higher layer protocol such as IP. In that case the interface has to go along (or before) the higher layer protocol is shutdown. Kernel hhooks need to go last on teardown as they may be used at various higher layers and we cannot remove them before we cleaned up the higher layers. For interface teardown there are multiple paths: (a) a cloned interface is destroyed (inside a VIMAGE or in the base system), (b) any interface is moved from a virtual network stack to a different network stack ("vmove"), or (c) a virtual network stack is being shut down. All code paths go through if_detach_internal() where we, depending on the vmove flag or the vnet state, make a decision on how much to shut down; in case we are destroying a VNET the individual protocol layers will cleanup their own parts thus we cannot do so again for each interface as we end up with, e.g., double-frees, destroying locks twice or acquiring already destroyed locks. When calling into protocol cleanups we equally have to tell them whether they need to detach upper layer protocols ("ulp") or not (e.g., in6_ifdetach()). Provide or enahnce helper functions to do proper cleanup at a protocol rather than at an interface level. Approved by: re (hrs) Obtained from: projects/vnet Reviewed by: gnn, jhb Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6747
2016-06-21 13:48:49 +00:00
VNET_SYSINIT(vnet_mroute_init, SI_SUB_PROTO_MC, SI_ORDER_ANY, vnet_mroute_init,
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
NULL);
static void
vnet_mroute_uninit(const void *unused __unused)
{
Reduce the vnet_set module size of ip_mroute to allow loading as a module. With VIMAGE kernels modules get special treatment as they need to also keep the original values and make copies for each instance. For that a few pages of vnet modspace are provided and the kernel-linker and the VNET framework know how to deal with things. When the modspace is (almost) full, other modules which would overflow the modspace cannot be loaded and kldload will fail. ip_mroute uses a lot of variable space, mostly be four big arrays: set_vnet 0000000000000510 vnet_entry_multicast_register_if set_vnet 0000000000000700 vnet_entry_viftable set_vnet 0000000000002000 vnet_entry_bw_meter_timers set_vnet 0000000000002800 vnet_entry_bw_upcalls Dynamically malloc the three big ones for each instance we need and free them again on vnet teardown (the 4th is an ifnet). That way they only need module space for a single pointer and allow a lot more modules using virtualized variables to be loaded on a VNET kernel. PR: 206583 Reviewed by: hselasky, kp MFC after: 3 weeks Differential Revision: https://reviews.freebsd.org/D22443
2019-11-19 15:38:55 +00:00
free(V_bw_upcalls, M_MRTABLE);
free(V_bw_meter_timers, M_MRTABLE);
free(V_viftable, M_MRTABLE);
style: Remove remaining deprecated MALLOC/FREE macros Mechanically replace uses of MALLOC/FREE with appropriate invocations of malloc(9) / free(9) (a series of sed expressions). Something like: * MALLOC(a, b, ... -> a = malloc(... * FREE( -> free( * free((caddr_t) -> free( No functional change. For now, punt on modifying contrib ipfilter code, leaving a definition of the macro in its KMALLOC(). Reported by: jhb Reviewed by: cy, imp, markj, rmacklem Sponsored by: Dell EMC Isilon Differential Revision: https://reviews.freebsd.org/D14035
2018-01-25 22:25:13 +00:00
free(V_nexpire, M_MRTABLE);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
V_nexpire = NULL;
}
White space cleanup -- remove trailing tab's or spaces from any line. Sponsored by: Netflix Inc.
2020-02-12 13:31:36 +00:00
VNET_SYSUNINIT(vnet_mroute_uninit, SI_SUB_PROTO_MC, SI_ORDER_MIDDLE,
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
vnet_mroute_uninit, NULL);
Add code to make multicast routing be an LKM.
1994-09-14 21:21:29 +00:00
static int
Somewhat modernize ip_mroute.c: - Use sysctl to export stats - Use ip_encap.c's encapsulation support - Update lkm to kld (is 6 years a record for a broken module?) - Remove some unused cruft
2001-07-25 20:15:49 +00:00
ip_mroute_modevent(module_t mod, int type, void *unused)
Add code to make multicast routing be an LKM.
1994-09-14 21:21:29 +00:00
{
Step 1.5 of importing the network stack virtualization infrastructure from the vimage project, as per plan established at devsummit 08/08: http://wiki.freebsd.org/Image/Notes200808DevSummit Introduce INIT_VNET_*() initializer macros, VNET_FOREACH() iterator macros, and CURVNET_SET() context setting macros, all currently resolving to NOPs. Prepare for virtualization of selected SYSCTL objects by introducing a family of SYSCTL_V_*() macros, currently resolving to their global counterparts, i.e. SYSCTL_V_INT() == SYSCTL_INT(). Move selected #defines from sys/sys/vimage.h to newly introduced header files specific to virtualized subsystems (sys/net/vnet.h, sys/netinet/vinet.h etc.). All the changes are verified to have zero functional impact at this point in time by doing MD5 comparision between pre- and post-change object files(*). (*) netipsec/keysock.c did not validate depending on compile time options. Implemented by: julian, bz, brooks, zec Reviewed by: julian, bz, brooks, kris, rwatson, ... Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-10-02 15:37:58 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
switch (type) {
case MOD_LOAD:
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK_INIT();
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
White space cleanup -- remove trailing tab's or spaces from any line. Sponsored by: Netflix Inc.
2020-02-12 13:31:36 +00:00
if_detach_event_tag = EVENTHANDLER_REGISTER(ifnet_departure_event,
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
if_detached_event, NULL, EVENTHANDLER_PRI_ANY);
if (if_detach_event_tag == NULL) {
Fix incomplete printf. PR: kern/177889 Submitted by: Sven-Thorsten Dietrich <sven vyatta com> MFC after: 1 week
2013-04-16 19:32:12 +00:00
printf("ip_mroute: unable to register "
"ifnet_departure_event handler\n");
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
MROUTER_LOCK_DESTROY();
return (EINVAL);
}
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
MFC_LOCK_INIT();
VIF_LOCK_INIT();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
mfchashsize = MFCHASHSIZE;
if (TUNABLE_ULONG_FETCH("net.inet.ip.mfchashsize", &mfchashsize) &&
!powerof2(mfchashsize)) {
printf("WARNING: %s not a power of 2; using default\n",
"net.inet.ip.mfchashsize");
mfchashsize = MFCHASHSIZE;
}
pim_squelch_wholepkt = 0;
If the rendezvous point for a group is not specified, do not send IGMPMSG_WHOLEPKT notifications to the userland PIM routing daemon, as an optimization to mitigate the effects of high multicast forwarding load. This is an experimental change, therefore it must be explicitly enabled by setting the sysctl/tunable net.inet.pim.squelch_wholepkt to a non-zero value. The tunable may be set from the loader or from within the kernel environment when loading ip_mroute.ko as a module. Submitted by: edrt <edrt at citiz.net> See also: http://mailman.icsi.berkeley.edu/pipermail/xorp-users/2005-June/000639.html
2007-02-10 14:48:42 +00:00
TUNABLE_ULONG_FETCH("net.inet.pim.squelch_wholepkt",
&pim_squelch_wholepkt);
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
pim_encap_cookie = ip_encap_attach(&ipv4_encap_cfg, NULL, M_WAITOK);
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
if (pim_encap_cookie == NULL) {
printf("ip_mroute: unable to attach pim encap\n");
VIF_LOCK_DESTROY();
MFC_LOCK_DESTROY();
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK_DESTROY();
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
return (EINVAL);
}
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_mcast_src = X_ip_mcast_src;
ip_mforward = X_ip_mforward;
ip_mrouter_done = X_ip_mrouter_done;
ip_mrouter_get = X_ip_mrouter_get;
ip_mrouter_set = X_ip_mrouter_set;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_rsvp_force_done = X_ip_rsvp_force_done;
ip_rsvp_vif = X_ip_rsvp_vif;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
legal_vif_num = X_legal_vif_num;
mrt_ioctl = X_mrt_ioctl;
rsvp_input_p = X_rsvp_input;
break;
Add code to make multicast routing be an LKM.
1994-09-14 21:21:29 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
case MOD_UNLOAD:
o restructure initialization code so data structures are setup when loaded as a module o cleanup data structures on module unload when no application has been started (i.e. kldload, kldunload w/o mrtd) o remove extraneous unlocks immediately prior to destroying them Supported by: FreeBSD Foundation
2003-10-24 00:09:18 +00:00
/*
* Typically module unload happens after the user-level
* process has shutdown the kernel services (the check
* below insures someone can't just yank the module out
* from under a running process). But if the module is
* just loaded and then unloaded w/o starting up a user
* process we still need to cleanup.
*/
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
MROUTER_LOCK();
if (ip_mrouter_cnt != 0) {
MROUTER_UNLOCK();
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
return (EINVAL);
Virtualize the IPv4 multicast routing code. Submitted by: iprebeg Reviewed by: bms, bz, Pavlin Radoslavov MFC after: 30 days
2010-06-02 15:44:43 +00:00
}
ip_mrouter_unloading = 1;
MROUTER_UNLOCK();
EVENTHANDLER_DEREGISTER(ifnet_departure_event, if_detach_event_tag);
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
if (pim_encap_cookie) {
Rework IP encapsulation handling code. Currently it has several disadvantages: - it uses single mutex to protect internal structures. It is used by data- and control- path, thus there are no parallelism at all. - it uses single list to keep encap handlers for both INET and INET6 families. - struct encaptab keeps unneeded information (src, dst, masks, protosw), that isn't used by code in the source tree. - matches are prioritized and when many tunneling interfaces are registered, encapcheck handler of each interface is invoked for each packet. The search takes O(n) for n interfaces. All this work is done with exclusive lock held. What this patch includes: - the datapath is converted to be lockless using epoch(9) KPI. - struct encaptab now linked using CK_LIST. - all unused fields removed from struct encaptab. Several new fields addedr: min_length is the minimum packet length, that encapsulation handler expects to see; exact_match is maximum number of bits, that can return an encapsulation handler, when it wants to consume a packet. - IPv6 and IPv4 handlers are stored in separate lists; - added new "encap_lookup_t" method, that will be used later. It is targeted to speedup lookup of needed interface, when gif(4)/gre(4) have many interfaces. - the need to use protosw structure is eliminated. The only pr_input method was used from this structure, so I don't see the need to keep using it. - encap_input_t method changed to avoid using mbuf tags to store softc pointer. Now it is passed directly trough encap_input_t method. encap_getarg() funtions is removed. - all sockaddr structures and code that uses them removed. We don't have any code in the tree that uses them. All consumers use encap_attach_func() method, that relies on invoking of encapcheck() to determine the needed handler. - introduced struct encap_config, it contains parameters of encap handler that is going to be registered by encap_attach() function. - encap handlers are stored in lists ordered by exact_match value, thus handlers that need more bits to match will be checked first, and if encapcheck method returns exact_match value, the search will be stopped. - all current consumers changed to use new KPI. Reviewed by: mmacy Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D15617
2018-06-05 20:51:01 +00:00
ip_encap_detach(pim_encap_cookie);
Build PIM by default as part of the IPv4 multicast forwarding path. Make PIM dynamically loadable by using encap_attach_func(). PIM may now be loaded into a GENERIC kernel. Tested with: ports/net/pimdd && tcpreplay && wireshark Reviewed by: Pavlin Radoslavov
2007-02-10 13:59:13 +00:00
pim_encap_cookie = NULL;
}
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_mcast_src = NULL;
ip_mforward = NULL;
ip_mrouter_done = NULL;
ip_mrouter_get = NULL;
ip_mrouter_set = NULL;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
ip_rsvp_force_done = NULL;
ip_rsvp_vif = NULL;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
legal_vif_num = NULL;
mrt_ioctl = NULL;
rsvp_input_p = NULL;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
o move mutex init/destroy logic to the module load/unload hooks; otherwise they are initialized twice when the code is statically configured in the kernel because the module load method gets invoked before the user application calls ip_mrouter_init o add a mutex to synchronize the module init/done operations; this sort of was done using the value of ip_mroute but X_ip_mrouter_done sets it to NULL very early on which can lead to a race against ip_mrouter_init--using the additional mutex means this is safe now o don't call ip_mrouter_reset from ip_mrouter_init; this now happens once at module load and X_ip_mrouter_done does the appropriate cleanup work to insure the data structures are in a consistent state so that a subsequent init operation inherits good state Reviewed by: juli
2003-12-20 18:32:48 +00:00
VIF_LOCK_DESTROY();
MFC_LOCK_DESTROY();
Style: Move declaration of subsystem mutex to where other mutexes are in this file, and use macros for dealing with it.
2007-02-28 20:02:24 +00:00
MROUTER_LOCK_DESTROY();
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
break;
Make IPv6 multicast forwarding dynamically loadable from a GENERIC kernel. It is built in the same module as IPv4 multicast forwarding, i.e. ip_mroute.ko, if and only if IPv6 support is enabled for loadable modules. Export IPv6 forwarding structs to userland netstat(1) via sysctl(9).
2007-02-24 11:38:47 +00:00
Do a pass over all modules in the kernel and make them return EOPNOTSUPP for unknown events. A number of modules return EINVAL in this instance, and I have left those alone for now and instead taught MOD_QUIESCE to accept this as "didn't do anything".
2004-07-15 08:26:07 +00:00
default:
return EOPNOTSUPP;
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
}
return 0;
Add code to make multicast routing be an LKM.
1994-09-14 21:21:29 +00:00
}
Somewhat modernize ip_mroute.c: - Use sysctl to export stats - Use ip_encap.c's encapsulation support - Update lkm to kld (is 6 years a record for a broken module?) - Remove some unused cruft
2001-07-25 20:15:49 +00:00
static moduledata_t ip_mroutemod = {
Massive cleanup of the ip_mroute code. No functional changes, but: + the mrouting module now should behave the same as the compiled-in version (it did not before, some of the rsvp code was not loaded properly); + netinet/ip_mroute.c is now truly optional; + removed some redundant/unused code; + changed many instances of '0' to NULL and INADDR_ANY as appropriate; + removed several static variables to make the code more SMP-friendly; + fixed some minor bugs in the mrouting code (mostly, incorrect return values from functions). This commit is also a prerequisite to the addition of support for PIM, which i would like to put in before DP2 (it does not change any of the existing APIs, anyways). Note, in the process we found out that some device drivers fail to properly handle changes in IFF_ALLMULTI, leading to interesting behaviour when a multicast router is started. This bug is not corrected by this commit, and will be fixed with a separate commit. Detailed changes: -------------------- netinet/ip_mroute.c all the above. conf/files make ip_mroute.c optional net/route.c fix mrt_ioctl hook netinet/ip_input.c fix ip_mforward hook, move rsvp_input() here together with other rsvp code, and a couple of indentation fixes. netinet/ip_output.c fix ip_mforward and ip_mcast_src hooks netinet/ip_var.h rsvp function hooks netinet/raw_ip.c hooks for mrouting and rsvp functions, plus interface cleanup. netinet/ip_mroute.h remove an unused and optional field from a struct Most of the code is from Pavlin Radoslavov and the XORP project Reviewed by: sam MFC after: 1 week
2002-11-15 22:53:53 +00:00
"ip_mroute",
ip_mroute_modevent,
Revert previous commit... Pointyhat to: kevlo (myself)
2012-10-10 08:36:38 +00:00
0
Somewhat modernize ip_mroute.c: - Use sysctl to export stats - Use ip_encap.c's encapsulation support - Update lkm to kld (is 6 years a record for a broken module?) - Remove some unused cruft
2001-07-25 20:15:49 +00:00
};
Introduce a number of changes to the MROUTING code. This is purely a forwarding plane cleanup; no control plane code is involved. Summary: * Split IPv4 and IPv6 MROUTING support. The static compile-time kernel option remains the same, however, the modules may now be built for IPv4 and IPv6 separately as ip_mroute_mod and ip6_mroute_mod. * Clean up the IPv4 multicast forwarding code to use BSD queue and hash table constructs. Don't build our own timer abstractions when ratecheck() and timevalclear() etc will do. * Expose the multicast forwarding cache (MFC) and virtual interface table (VIF) as sysctls, to reduce netstat's dependence on libkvm for this information for running kernels. * bandwidth meters however still require libkvm. * Make the MFC hash table size a boot/load-time tunable ULONG, net.inet.ip.mfchashsize (defaults to 256). * Remove unused members from struct vif and struct mfc. * Kill RSVP support, as no current RSVP implementation uses it. These stubs could be moved to raw_ip.c. * Don't share locks or initialization between IPv4 and IPv6. * Don't use a static struct route_in6 in ip6_mroute.c. The v6 code is still using a cached struct route_in6, this is moved to mif6 for the time being. * More cleanup remains to be merged from ip_mroute.c to ip6_mroute.c. v4 path tested using ports/net/mcast-tools. v6 changes are mostly mechanical locking and *have not* been tested. As these changes partially break some kernel ABIs, they will not be MFCed. There is a lot more work to be done here. Reviewed by: Pavlin Radoslavov
2009-03-19 01:43:03 +00:00
Get closer to a VIMAGE network stack teardown from top to bottom rather than removing the network interfaces first. This change is rather larger and convoluted as the ordering requirements cannot be separated. Move the pfil(9) framework to SI_SUB_PROTO_PFIL, move Firewalls and related modules to their own SI_SUB_PROTO_FIREWALL. Move initialization of "physical" interfaces to SI_SUB_DRIVERS, move virtual (cloned) interfaces to SI_SUB_PSEUDO. Move Multicast to SI_SUB_PROTO_MC. Re-work parts of multicast initialisation and teardown, not taking the huge amount of memory into account if used as a module yet. For interface teardown we try to do as many of them as we can on SI_SUB_INIT_IF, but for some this makes no sense, e.g., when tunnelling over a higher layer protocol such as IP. In that case the interface has to go along (or before) the higher layer protocol is shutdown. Kernel hhooks need to go last on teardown as they may be used at various higher layers and we cannot remove them before we cleaned up the higher layers. For interface teardown there are multiple paths: (a) a cloned interface is destroyed (inside a VIMAGE or in the base system), (b) any interface is moved from a virtual network stack to a different network stack ("vmove"), or (c) a virtual network stack is being shut down. All code paths go through if_detach_internal() where we, depending on the vmove flag or the vnet state, make a decision on how much to shut down; in case we are destroying a VNET the individual protocol layers will cleanup their own parts thus we cannot do so again for each interface as we end up with, e.g., double-frees, destroying locks twice or acquiring already destroyed locks. When calling into protocol cleanups we equally have to tell them whether they need to detach upper layer protocols ("ulp") or not (e.g., in6_ifdetach()). Provide or enahnce helper functions to do proper cleanup at a protocol rather than at an interface level. Approved by: re (hrs) Obtained from: projects/vnet Reviewed by: gnn, jhb Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6747
2016-06-21 13:48:49 +00:00
DECLARE_MODULE(ip_mroute, ip_mroutemod, SI_SUB_PROTO_MC, SI_ORDER_MIDDLE);
Reference in New Issue Copy Permalink