tcp: don't backtrack to sacked skbs
authorIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Sat, 28 Feb 2009 04:44:27 +0000 (04:44 +0000)
committerDavid S. Miller <davem@davemloft.net>
Mon, 2 Mar 2009 11:00:08 +0000 (03:00 -0800)
commitac11ba753f3aa839292c1a3b6c971c637ad2e839
tree36bedcc6c6d6be721f2b9c5c0e5b8bf66ca112a5
parent79cb5def8dcad1a978841118bd14e155b3fd5589
tcp: don't backtrack to sacked skbs

Backtracking to sacked skbs is a horrible performance killer
since the hint cannot be advanced successfully past them...
...And it's totally unnecessary too.

In theory this is 2.6.27..28 regression but I doubt anybody
can make .28 to have worse performance because of other TCP
improvements.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4/tcp_output.c