svn commit: r218039 - head/sys/netinet

Author: rrs Date: Fri Jan 28 21:05:21 2011 New Revision: 218039 URL: http://svn.freebsd.org/changeset/base/218039 Log: Keep track of the real last RTT on each net. This will be used for Data Center congestion control, we won't want to engage it in the ECN code unless we KNOW that the RTT

svn commit: r217635 - head/sys/netinet

Author: rrs Date: Thu Jan 20 12:40:09 2011 New Revision: 217635 URL: http://svn.freebsd.org/changeset/base/217635 Log: Fix it so we align with new socket API draft for state's in destination (i.e. ACTIVE/INACTIVE/UNCONFIRMED) MFC after:1 week Modified:

svn commit: r217592 - head/sys/netinet

Author: rrs Date: Wed Jan 19 19:07:16 2011 New Revision: 217592 URL: http://svn.freebsd.org/changeset/base/217592 Log: Fix a bug where Multicast packets sent from a udp endpoint may end up echoing back to the sender even with OUT joining the multi-cast group. Reviewed by: gnn, bms,

svn commit: r215817 - head/sys/netinet

== --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/sys/netinet/sctp_dtrace_declare.h Thu Nov 25 13:39:55 2010 (r215817) @@ -0,0 +1,73 @@ +/*- + * Copyright (c) 2010, by Randall Stewart Michael Tuexen, + * All rights reserved. + * + * Redistribution and use

svn commit: r215110 - head/sys/netinet

Author: rrs Date: Thu Nov 11 05:40:39 2010 New Revision: 215110 URL: http://svn.freebsd.org/changeset/base/215110 Log: Fix so that a multicast packet can be sent even if there is no route out to that mcast address. The code in in_pcb inadvertantly would error (no route) even though the

svn commit: r212225 - head/sys/netinet

Author: rrs Date: Sun Sep 5 13:41:45 2010 New Revision: 212225 URL: http://svn.freebsd.org/changeset/base/212225 Log: Fix some CLANG warnings. One clang warning is left due to the fact that its bogus.. nam-sa_family will not change from AF_INET6 to AF_INET (but clang thinks it does ;-D)

svn commit: r210600 - head/sys/netinet

Author: rrs Date: Thu Jul 29 12:01:59 2010 New Revision: 210600 URL: http://svn.freebsd.org/changeset/base/210600 Log: Fix the comment block that has the nice table to really have the nice table :-) MFC after:1 month Modified: head/sys/netinet/sctp_output.c Modified:

svn commit: r210493 - head/sys/netinet

Author: rrs Date: Mon Jul 26 09:20:55 2010 New Revision: 210493 URL: http://svn.freebsd.org/changeset/base/210493 Log: When counting the number of chunks in the retransmission queue to validate the retran count, we need to include the chunks in the control send queue too. Otherwise the

svn commit: r210494 - head/sys/netinet

Author: rrs Date: Mon Jul 26 09:22:52 2010 New Revision: 210494 URL: http://svn.freebsd.org/changeset/base/210494 Log: Make sure that we report chunks if a socket still exists that were not sent. In either case carefully remove the data if it does not get taken by the reporting routines.

svn commit: r209760 - head/lib/libc/net

Author: rrs Date: Wed Jul 7 11:19:06 2010 New Revision: 209760 URL: http://svn.freebsd.org/changeset/base/209760 Log: If a user calls sctp_sendx() with a NULL sinfo we will crash. Instead we should provide our own temp structure to use internally. MFC after:1 month Modified:

svn commit: r209663 - head/sys/netinet

Author: rrs Date: Sat Jul 3 14:03:31 2010 New Revision: 209663 URL: http://svn.freebsd.org/changeset/base/209663 Log: This fixes a crash in SCTP. It was possible to have a large number of packets queued to a crashing process. In a specific case you may get 2 ABORT's back (from say two

svn commit: r209644 - head/sys/netinet

Author: rrs Date: Fri Jul 2 09:53:26 2010 New Revision: 209644 URL: http://svn.freebsd.org/changeset/base/209644 Log: Fix a bug that WILL cause a panic. Basically a read-lock is being called to check the vtag-timewait cache. Then in two cases (where a vtag is bad i.e. in the time-wait

svn commit: r209029 - head/sys/netinet

Author: rrs Date: Fri Jun 11 03:54:00 2010 New Revision: 209029 URL: http://svn.freebsd.org/changeset/base/209029 Log: 3 Fixes - a) There was a case where a ICMP message could cause us to return leaving a stuck lock on an stcb. b) The iterator needed some tweaks to fix its lock

svn commit: r208952 - head/sys/netinet

Author: rrs Date: Wed Jun 9 16:39:18 2010 New Revision: 208952 URL: http://svn.freebsd.org/changeset/base/208952 Log: BUG:Turns out we need to use both bit maps to calculate the cum-ack (we were not doing it for the NR-Sack case). With this fix NR-sack should now work correctly. MFC

svn commit: r208953 - head/sys/netinet

Author: rrs Date: Wed Jun 9 16:42:42 2010 New Revision: 208953 URL: http://svn.freebsd.org/changeset/base/208953 Log: Fix serveral bugs all having to do with freeing an sctp_inpcb: 1) Make sure not to remove the flag on the PCB until after the close() caller is back in control with

svn commit: r208970 - head/sys/netinet

Author: rrs Date: Wed Jun 9 22:05:29 2010 New Revision: 208970 URL: http://svn.freebsd.org/changeset/base/208970 Log: Found by Michael. In cases where we run out of memory (no more inp space) we don't propely NULL the INP on return. Obtained from:tuexen MFC after:3 Days

Re: svn commit: r208876 - head/sys/netinet

Cool... I will move these over to that assert... And the invariant is ok since this function only is there when you compile invariant in. R On Jun 7, 2010, at 1:07 PM, Pawel Jakub Dawidek wrote: On Sun, Jun 06, 2010 at 04:11:17PM +, Randall Stewart wrote: Author: rrs Date: Sun Jun 6 16

svn commit: r208902 - head/sys/netinet

Author: rrs Date: Tue Jun 8 03:39:31 2010 New Revision: 208902 URL: http://svn.freebsd.org/changeset/base/208902 Log: 2 Bugs: 1) Only use both mapping arrays when NR sack is off. This way we can hold off moving the cumack (not the best but workable) when NR-sack is on. 2)

svn commit: r208876 - head/sys/netinet

Author: rrs Date: Sun Jun 6 16:11:16 2010 New Revision: 208876 URL: http://svn.freebsd.org/changeset/base/208876 Log: 1) Further enhance the INVARIANT lock validation (no locks) are held by checking the create and inp locks as well. 2) Fix a bug in that when a socket is closed an

svn commit: r208852 - head/sys/netinet

Author: rrs Date: Sat Jun 5 21:17:23 2010 New Revision: 208852 URL: http://svn.freebsd.org/changeset/base/208852 Log: This fixes a bug in the close up of a socket that had un-accepted assoc's. Basically the assoc (and inp) would get stuck and never get cleaned up. MFC after:1 week

svn commit: r208853 - head/sys/netinet

Author: rrs Date: Sat Jun 5 21:20:28 2010 New Revision: 208853 URL: http://svn.freebsd.org/changeset/base/208853 Log: This does two changes: 1) Makes it so that the INVARIANT function validate nolocks is available anywhere. 2) Fixes a BUG where a close has been done on a collision

svn commit: r208854 - head/sys/netinet

Author: rrs Date: Sat Jun 5 21:22:58 2010 New Revision: 208854 URL: http://svn.freebsd.org/changeset/base/208854 Log: Use the proper increment macro when increasing the number on sent_queue_retran_cnt. MFC after:1 week Modified: head/sys/netinet/sctp_indata.c Modified:

svn commit: r208856 - head/sys/netinet

Author: rrs Date: Sat Jun 5 21:33:16 2010 New Revision: 208856 URL: http://svn.freebsd.org/changeset/base/208856 Log: Spacing issues MFC after:1 Week Modified: head/sys/netinet/sctp_auth.c head/sys/netinet/sctp_bsd_addr.c Modified: head/sys/netinet/sctp_auth.c

Re: svn commit: r208857 - head/sys/netinet

we would have to get agreement from quite a few folks... but I am game ;-) R On Jun 5, 2010, at 5:44 PM, Luigi Rizzo wrote: On Sat, Jun 05, 2010 at 09:39:53PM +, Randall Stewart wrote: Author: rrs Date: Sat Jun 5 21:39:52 2010 New Revision: 208857 URL: http://svn.freebsd.org/changeset

svn commit: r208863 - head/sys/netinet

Author: rrs Date: Sun Jun 6 02:32:20 2010 New Revision: 208863 URL: http://svn.freebsd.org/changeset/base/208863 Log: Bruce's fix for some return's in error legs. MFC after:1 week Modified: head/sys/netinet/sctp_usrreq.c Modified: head/sys/netinet/sctp_usrreq.c

svn commit: r208864 - head/sys/netinet

Author: rrs Date: Sun Jun 6 02:33:46 2010 New Revision: 208864 URL: http://svn.freebsd.org/changeset/base/208864 Log: Hopefully this fixes a LOR by making so we only hold the iterator lock during updates to the iterators work. MFC after:1 week Modified:

svn commit: r207963 - head/sys/netinet

Author: rrs Date: Wed May 12 13:45:46 2010 New Revision: 207963 URL: http://svn.freebsd.org/changeset/base/207963 Log: This fixes PR-SCTP issues: - Slide the map at the proper place. - Mark the bits in the nr_array ONLY if there is no marking. - When generating a FWD-TSN we allow

svn commit: r207983 - head/sys/netinet

Author: rrs Date: Wed May 12 18:00:15 2010 New Revision: 207983 URL: http://svn.freebsd.org/changeset/base/207983 Log: More PR-SCTP bugs: - Make sure that when you kick the streams you add correctly using a 16 bit unsigned. - Make sure when sending out you allow FWD-TSN to skip

svn commit: r207985 - head/sys/netinet

Author: rrs Date: Wed May 12 18:33:25 2010 New Revision: 207985 URL: http://svn.freebsd.org/changeset/base/207985 Log: Fix an old long time bug in generating a fwd-tsn. This would appear when greater than the size of mbuf TSN's would need to be skipped. MFC after:3 days Modified:

svn commit: r207924 - head/sys/netinet

Author: rrs Date: Tue May 11 17:02:29 2010 New Revision: 207924 URL: http://svn.freebsd.org/changeset/base/207924 Log: This fixes a bug with the one-2-one model socket when a user sets up a socket to a server sends data and closes the socket before the server has called accept(). It used

svn commit: r205627 - head/sys/netinet

Author: rrs Date: Wed Mar 24 19:45:36 2010 New Revision: 205627 URL: http://svn.freebsd.org/changeset/base/205627 Log: Fix for NR-Sack code. The code was NOT working properly when enabled. Basically most of the operations were incorrect causing bad sacks when you enabled nr-sack. The fixes

svn commit: r205628 - head/sys/netinet

Author: rrs Date: Wed Mar 24 20:01:14 2010 New Revision: 205628 URL: http://svn.freebsd.org/changeset/base/205628 Log: lagging file I forgot to commit with my nr-sack fixes... opps Reviewed by: tue...@freebsd.org Modified: head/sys/netinet/sctputil.h Modified:

svn commit: r205629 - in head/sys: conf netinet

Author: rrs Date: Wed Mar 24 20:02:40 2010 New Revision: 205629 URL: http://svn.freebsd.org/changeset/base/205629 Log: Adds the option of keeping per-cpu statistics in SCTP. This may be useful since it gets rid of atomics but I want it to remain an option until I can do further testing on

svn commit: r205502 - head/sys/netinet

Author: rrs Date: Tue Mar 23 01:36:50 2010 New Revision: 205502 URL: http://svn.freebsd.org/changeset/base/205502 Log: Fixes a bug where SACKs in the face of mapping_array expansion would break. Basically once we expanded the array we no longer had both mapping arrays in sync which the

Re: svn commit: r205104 - in head/sys: dev/xen/netback netinet netinet6

... it would mean a few more de-ref's and looking to see if its a v4 or v6 packet and then doing the proper offset... not to bad but awkward ;-0 R On Mar 13, 2010, at 9:58 PM, Robert Watson wrote: On Fri, 12 Mar 2010, Randall Stewart wrote: The proper fix for the delayed SCTP checksum

svn commit: r205075 - head/sys/netinet6

Author: rrs Date: Fri Mar 12 08:10:30 2010 New Revision: 205075 URL: http://svn.freebsd.org/changeset/base/205075 Log: With the recent change of the sctp checksum to support offload, no delayed checksum was added to the ip6 output code. This causes cards that do not support SCTP checksum

svn commit: r204577 - in head/sys/mips: include rmi

Author: rrs Date: Tue Mar 2 12:11:00 2010 New Revision: 204577 URL: http://svn.freebsd.org/changeset/base/204577 Log: - Move rmi_pci_bus_space to header and avoid extern - remove unused and commented code (MIPS_BUS_SPACE_PCI, pic_usb_ack) - use rmi_pci_bus_space for USB too (needs

svn commit: r204135 - head/sys/mips/rmi

Author: rrs Date: Sat Feb 20 17:12:07 2010 New Revision: 204135 URL: http://svn.freebsd.org/changeset/base/204135 Log: Opps forgot to add this: - add bus_space_rmi_pci.c for PCI bus space Obtained from:C. Jayachandran - c.jayachand...@gmail.com Added:

<    1   2   3