[TCP]: skb pcount with MTU discovery
authorDavid S. Miller <davem@sunset.davemloft.net>
Mon, 25 Apr 2005 02:12:33 +0000 (19:12 -0700)
committerDavid S. Miller <davem@davemloft.net>
Mon, 25 Apr 2005 02:12:33 +0000 (19:12 -0700)
commitd5ac99a648b8c61d0c7f1c32a8ab7f1dca0123d2
tree3811d84e83dab918c57aeca1081ae343cb97df8b
parent158a0e45b69254a9ee4d7795e3b98d8c959fb799
[TCP]: skb pcount with MTU discovery

The problem is that when doing MTU discovery, the too-large segments in
the write queue will be calculated as having a pcount of >1.  When
tcp_write_xmit() is trying to send, tcp_snd_test() fails the cwnd test
when pcount > cwnd.

The segments are eventually transmitted one at a time by keepalive, but
this can take a long time.

This patch checks if TSO is enabled when setting pcount.

Signed-off-by: John Heffner <jheffner@psc.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
include/net/tcp.h
net/ipv4/tcp_output.c