linux.git/net/ipv4/ip_output.c, branch v2.6.20 Linux kernel source tree [TCP]: Restore SKB socket owner setting in tcp_transmit_skb(). 2007-01-26T09:04:55+00:00 David S. Miller davem@sunset.davemloft.net 2007-01-26T09:04:55+00:00 e89862f4c5b3c4ac9afcd8cb1365d2f1e16ddc3b Revert 931731123a103cfb3f70ac4b7abfc71d94ba1f03 We can't elide the skb_set_owner_w() here because things like certain netfilter targets (such as owner MATCH) need a socket to be set on the SKB for correct operation. Thanks to Jan Engelhardt and other netfilter list members for pointing this out. Signed-off-by: David S. Miller <davem@davemloft.net> 
Revert 931731123a103cfb3f70ac4b7abfc71d94ba1f03

We can't elide the skb_set_owner_w() here because things like certain
netfilter targets (such as owner MATCH) need a socket to be set on the
SKB for correct operation.

Thanks to Jan Engelhardt and other netfilter list members for
pointing this out.

Signed-off-by: David S. Miller <davem@davemloft.net>
[NET]: Convert hh_lock to seqlock. 2006-12-09T01:19:20+00:00 Stephen Hemminger shemminger@osdl.org 2006-12-07T23:08:17+00:00 3644f0cee77494190452de132e82245107939284 The hard header cache is in the main output path, so using seqlock instead of reader/writer lock should reduce overhead. Signed-off-by: Stephen Hemminger <shemminger@osdl.org> Signed-off-by: David S. Miller <davem@davemloft.net> 
The hard header cache is in the main output path, so using
seqlock instead of reader/writer lock should reduce overhead.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
[PATCH] severing skbuff.h -> highmem.h 2006-12-04T07:00:29+00:00 Al Viro viro@zeniv.linux.org.uk 2006-10-19T20:08:53+00:00 a1f8e7f7fb9d7e2cbcb53170edca7c0ac4680697 Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[NET]: Annotate callers of csum_partial_copy_...() and csum_and_copy...() in net/* 2006-12-03T05:23:33+00:00 Al Viro viro@zeniv.linux.org.uk 2006-11-15T05:36:34+00:00 5084205faf45384fff25c4cf77dd5c96279283ad Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
[NET]: Annotate csum_partial() callers in net/* 2006-12-03T05:23:32+00:00 Al Viro viro@zeniv.linux.org.uk 2006-11-15T05:36:14+00:00 44bb93633f57a55979f3c2589b10fd6a2bfc7c08 Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
[NET]: Annotate callers of csum_fold() in net/* 2006-12-03T05:23:27+00:00 Al Viro viro@zeniv.linux.org.uk 2006-11-15T05:24:49+00:00 d3bc23e7ee9db8023dff5a86bb3b0069ed018789 Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
[IPV6]: Assorted trivial endianness annotations. 2006-12-03T05:22:50+00:00 Al Viro viro@zeniv.linux.org.uk 2006-11-15T04:51:49+00:00 714e85be3557222bc25f69c252326207c900a7db Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
[TCP]: Don't set SKB owner in tcp_transmit_skb(). 2006-12-03T05:21:52+00:00 David S. Miller davem@sunset.davemloft.net 2006-11-10T03:58:25+00:00 931731123a103cfb3f70ac4b7abfc71d94ba1f03 The data itself is already charged to the SKB, doing the skb_set_owner_w() just generates a lot of noise and extra atomics we don't really need. Lmbench improvements on lat_tcp are minimal: before: TCP latency using localhost: 23.2701 microseconds TCP latency using localhost: 23.1994 microseconds TCP latency using localhost: 23.2257 microseconds after: TCP latency using localhost: 22.8380 microseconds TCP latency using localhost: 22.9465 microseconds TCP latency using localhost: 22.8462 microseconds Signed-off-by: David S. Miller <davem@davemloft.net> 
The data itself is already charged to the SKB, doing
the skb_set_owner_w() just generates a lot of noise and
extra atomics we don't really need.

Lmbench improvements on lat_tcp are minimal:

before:
TCP latency using localhost: 23.2701 microseconds
TCP latency using localhost: 23.1994 microseconds
TCP latency using localhost: 23.2257 microseconds

after:
TCP latency using localhost: 22.8380 microseconds
TCP latency using localhost: 22.9465 microseconds
TCP latency using localhost: 22.8462 microseconds

Signed-off-by: David S. Miller <davem@davemloft.net>
[NET]: Turn nfmark into generic mark 2006-12-03T05:21:38+00:00 Thomas Graf tgraf@suug.ch 2006-11-09T23:19:14+00:00 82e91ffef60e6eba9848fe149ce1eecd2b5aef12 nfmark is being used in various subsystems and has become the defacto mark field for all kinds of packets. Therefore it makes sense to rename it to `mark' and remove the dependency on CONFIG_NETFILTER. Signed-off-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net> 
nfmark is being used in various subsystems and has become
the defacto mark field for all kinds of packets. Therefore
it makes sense to rename it to `mark' and remove the
dependency on CONFIG_NETFILTER.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
[IPV4]: struct ip_options annotations 2006-09-29T01:01:53+00:00 Al Viro viro@zeniv.linux.org.uk 2006-09-28T01:28:07+00:00 3ca3c68e76686bee058937ade2b96f4de58ee434 ->faddr is net-endian; annotated as such, variables inferred to be net-endian annotated. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> 
->faddr is net-endian; annotated as such, variables inferred to be net-endian
annotated.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>