cvs commit: doc/en_US.ISO8859-1/books/porters-handbook book.sgml

2008-12-15 Thread Qing Li
qingli 2008-12-15 08:18:57 UTC FreeBSD doc repository (src committer) Modified files: en_US.ISO8859-1/books/porters-handbook book.sgml Log: describe the reason behind the increment in version number Revision ChangesPath 1.963 +5 -0

cvs commit: src/sys/net radix_mpath.c radix_mpath.h

2008-05-30 Thread Qing Li
qingli 2008-05-30 09:34:35 UTC FreeBSD src repository Modified files: sys/net radix_mpath.c radix_mpath.h Log: When RADIX_MPATH is enabled, the route selection is not rotating through the multipath entries. The hash value was a signed integer and was always

cvs commit: src/sys/net radix_mpath.c

2008-04-23 Thread Qing Li
qingli 2008-04-24 05:04:52 UTC FreeBSD src repository Modified files: sys/net radix_mpath.c Log: In function rtalloc_mpath(), do not try to release the lock if the ro_rt pointer is NULL. Reported by:(pluknet at gmail dot com) Revision Changes

RE: multipath stuff

2008-04-17 Thread Qing Li
Hi Bruce, First of all thanks for doing this. It would have been nice to have had some advance warning though. Hopefully it wasn't too much of a grief ... I'm surprised XORP hasn't been mentioned here, so I'll mention it now... For what it's worth this should not

Re: cvs commit: src/sys/conf files options src/sys/net radix.c radix.h route.c route.h rtsock.c src/sys/netinet in_proto.c ip_output.c src/sys/netinet6 in6_proto.c in6_src.c nd6_nbr.c

2008-04-17 Thread Qing Li
Andre, Bruce, Claudio: What's the timeline you have in mind for these enhancements in the routing infrastructure? 8.0 release? Are we going to divide this work up among the four of us ? What's the regression environment and regression tests you use for verifying the kernel routing code? --

RE: cvs commit: src/sys/conf files options src/sys/net radix.c radix.h route.c route.h rtsock.c src/sys/netinet in_proto.c ip_output.c src/sys/netinet6 in6_proto.c in6_src.c nd6_nbr.c

2008-04-14 Thread Qing Li
Hi Andre, Do you do hash-based next-hop balancing (flow-aware) or packet based round robin? Should it be made an option to switch between them (globally) like on Cisco routers? This is hash-based next-hop balancing. is disallowed. For example,

cvs commit: src/sys/net radix_mpath.c radix_mpath.h

2008-04-13 Thread Qing Li
qingli 2008-04-13 06:12:13 UTC FreeBSD src repository Added files: sys/net radix_mpath.c radix_mpath.h Log: These files handle the radix tree for the ECMP routes. The original code from KAME did not take care of address aliases or multiple ip addresses that

cvs commit: src/sys/net radix_mpath.c

2008-04-13 Thread Qing Li
qingli 2008-04-13 23:04:46 UTC FreeBSD src repository Modified files: sys/net radix_mpath.c Log: Make this file compile on IPv6 kernels. Revision ChangesPath 1.3 +2 -0 src/sys/net/radix_mpath.c ___

cvs commit: src/sys/conf files options src/sys/net radix.c radix.h route.c route.h rtsock.c src/sys/netinet in_proto.c ip_output.c src/sys/netinet6 in6_proto.c in6_src.c nd6_nbr.c

2008-04-12 Thread Qing Li
qingli 2008-04-13 05:45:14 UTC FreeBSD src repository Modified files: sys/conf files options sys/net radix.c radix.h route.c route.h rtsock.c sys/netinet in_proto.c ip_output.c sys/netinet6 in6_proto.c in6_src.c nd6_nbr.c

cvs commit: src/sys/dev/fxp if_fxp.c

2008-03-22 Thread Qing Li
qingli 2008-03-22 18:13:39 UTC FreeBSD src repository Modified files: sys/dev/fxp if_fxp.c Log: Reuse the mbuf that was just retrieved from the receive ring if mbuf exhaustion is encountered. There was a fix made previously for this problem but the solution

cvs commit: src/sys/netinet tcp_subr.c

2007-12-14 Thread Qing Li
qingli 2007-12-14 19:23:32 UTC FreeBSD src repository Modified files:(Branch: RELENG_6) sys/netinet tcp_subr.c Log: MFC:-r1.296 use the sequence number comparison macro to compare projected_offset against isn_offset to account for

cvs commit: src/sys/netinet tcp_subr.c

2007-08-15 Thread Qing Li
qingli 2007-08-16 01:35:55 UTC FreeBSD src repository Modified files: sys/netinet tcp_subr.c Log: Use the sequence number comparison macro to compare projected_offset against isn_offset to account for wrap around. Reviewed by:gnn, kmacy, silby Submitted

cvs commit: src/sys/netinet tcp_input.c

2007-03-16 Thread Qing Li
qingli 2007-03-16 22:04:25 UTC FreeBSD src repository Modified files:(Branch: RELENG_6) sys/netinet tcp_input.c Log: MFC:r1.316, accept valid RST packet when delayed ack is in effect. RevisionChangesPath 1.281.2.12 +7 -5

cvs commit: src/sys/netinet tcp_input.c

2007-03-07 Thread Qing Li
qingli 2007-03-07 23:21:59 UTC FreeBSD src repository Modified files: sys/netinet tcp_input.c Log: This patch is provided to fix a couple of deployment issues observed in the field. In one situation, one end of the TCP connection sends a back-to-back RST packet,

cvs commit: src/sys/net if_vlan.c

2006-08-01 Thread Qing Li
qingli 2006-08-01 17:28:10 UTC FreeBSD src repository Modified files: sys/net if_vlan.c Log: In vlan_input(), if the network interface does not perform h/w based vlan tag processing, the code will use bcopy() to remove the vlan tag field but the code copies 2

Re: cvs commit: src/sys/net if_vlan.c

2006-08-01 Thread Qing Li
Removing the dead assignment to evl-evl_encap_proto above and reducing the bcopy window by 2 bytes here is a nano-optimization. Which is worse, it smuggles hidden dot1q details in this otherwise encapsulation-neutral block of code. E.g., I could use the former code for both dot1q and ISL

cvs commit: src/sys/net route.c

2006-06-05 Thread Qing Li
qingli 2006-06-05 21:20:21 UTC FreeBSD src repository Modified files: sys/net route.c Log: Assuming the interface has an address of x.x.x.195, a mask of 255.255.255.0, and a default route with gateway x.x.x.1. Now if the address mask is changed to something

cvs commit: src/sys/net route.c

2006-05-16 Thread Qing Li
qingli 2006-05-16 19:11:11 UTC FreeBSD src repository Modified files: sys/net route.c Log: The current routing code allows insertion of indirect routes that have gateways which are unreachable except through the default router. For example, assuming there is a

cvs commit: src/sys/netinet tcp_input.c tcp_output.c

2006-02-23 Thread Qing Li
qingli 2006-02-23 21:14:34 UTC FreeBSD src repository Modified files: sys/netinet tcp_input.c tcp_output.c Log: This patch fixes the problem where the current TCP code can not handle simultaneous open. Both the bug and the patch were verified using the ANVL test

cvs commit: src/sys/net radix.c

2006-02-16 Thread Qing Li
qingli 2006-02-16 17:51:53 UTC FreeBSD src repository Modified files:(Branch: RELENG_5) sys/net radix.c Log: Patch for PR kern/38752 MFC:rev 1.38 Approved by:re (scottl) Revision ChangesPath 1.36.2.2 +9 -1

cvs commit: src/sys/net route.c

2006-02-16 Thread Qing Li
qingli 2006-02-16 17:59:19 UTC FreeBSD src repository Modified files:(Branch: RELENG_5) sys/net route.c Log: Require gateways for routes to be of the same address family as the route itself. MFC:rev 1.109 Approved by:re (scottl)

cvs commit: src/sys/netinet tcp_syncache.c

2006-02-16 Thread Qing Li
qingli 2006-02-16 18:05:03 UTC FreeBSD src repository Modified files:(Branch: RELENG_5) sys/netinet tcp_syncache.c Log: Fixes a crash due to the memory of the newly allocated syncache entry in syncache_lookup() is not cleared and may lead to an arbitrary and

cvs commit: src/sys/net if_gre.c

2006-02-16 Thread Qing Li
qingli 2006-02-16 18:10:14 UTC FreeBSD src repository Modified files:(Branch: RELENG_5) sys/net if_gre.c Log: Remove two unnecessary type casts, of which both had a typo in it anyways. MFC:rev 1.40 Approved by:re (scottl)

cvs commit: src/sys/netinet tcp_syncache.c

2006-02-15 Thread Qing Li
qingli 2006-02-16 01:06:22 UTC FreeBSD src repository Modified files:(Branch: RELENG_6) sys/netinet tcp_syncache.c Log: MFC:rev 1.84 Approved by:re (scottl) Revision ChangesPath 1.74.2.5 +3 -4 src/sys/netinet/tcp_syncache.c

cvs commit: src/sys/net if_gre.c

2006-02-15 Thread Qing Li
qingli 2006-02-16 01:08:40 UTC FreeBSD src repository Modified files:(Branch: RELENG_6) sys/net if_gre.c Log: MFC:rev 1.40 Approved by:re (scottl) Revision ChangesPath 1.32.2.5 +2 -2 src/sys/net/if_gre.c

cvs commit: src/sys/net radix.c

2006-02-15 Thread Qing Li
qingli 2006-02-16 01:10:45 UTC FreeBSD src repository Modified files:(Branch: RELENG_6) sys/net radix.c Log: MFC:rev 1.38 Approved by:re (scottl) Revision ChangesPath 1.37.2.1 +9 -1 src/sys/net/radix.c

cvs commit: src/sys/netinet tcp_syncache.c

2006-02-08 Thread Qing Li
qingli 2006-02-08 23:32:58 UTC FreeBSD src repository Modified files: sys/netinet tcp_syncache.c Log: Redo the previous fix by setting the UMA_ZONE_ZINIT bit in the syncache zone, eliminating the need to call bzero() after each syncache entry allocation.

cvs commit: src/sys/netinet tcp_syncache.c

2006-02-07 Thread Qing Li
qingli 2006-02-07 19:59:46 UTC FreeBSD src repository Modified files: sys/netinet tcp_syncache.c Log: Fixes a crash due to the memory of the newly allocated syncache entry in syncache_lookup() is not cleared and may lead to an arbitrary and bogus rtentry pointer

cvs commit: src/sys/net if_gre.c

2006-02-07 Thread Qing Li
qingli 2006-02-07 20:09:02 UTC FreeBSD src repository Modified files: sys/net if_gre.c Log: Remove two unnecessary type casts, of which both had a typo in it anyways. Approved by: andre MFC after: 3 days Revision ChangesPath 1.40 +2 -2

cvs commit: src/sys/net radix.c

2006-02-07 Thread Qing Li
qingli 2006-02-07 20:25:39 UTC FreeBSD src repository Modified files: sys/net radix.c Log: The code in rn_walktree_from() that checks if we backed up too far did not stop at the right node. Change the backtracking check from smaller-than to smaller-or-equal to