tcp: fix wrong checksum calculation on MTU probing
authorDouglas Caetano dos Santos <douglascs@taghos.com.br>
Thu, 22 Sep 2016 18:52:04 +0000 (15:52 -0300)
committerWilly Tarreau <w@1wt.eu>
Fri, 10 Feb 2017 10:03:46 +0000 (11:03 +0100)
commitd318f82f1c4e3823253fcc9dc2053a46ef7b14a6
treef65b7e9838d458b4d96fe2597c73daf5400b2221
parent93522d310c31a54973def2a6ea15d3b418045f49
tcp: fix wrong checksum calculation on MTU probing

commit 2fe664f1fcf7c4da6891f95708a7a56d3c024354 upstream.

With TCP MTU probing enabled and offload TX checksumming disabled,
tcp_mtu_probe() calculated the wrong checksum when a fragment being copied
into the probe's SKB had an odd length. This was caused by the direct use
of skb_copy_and_csum_bits() to calculate the checksum, as it pads the
fragment being copied, if needed. When this fragment was not the last, a
subsequent call used the previous checksum without considering this
padding.

The effect was a stale connection in one way, as even retransmissions
wouldn't solve the problem, because the checksum was never recalculated for
the full SKB length.

Signed-off-by: Douglas Caetano dos Santos <douglascs@taghos.com.br>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Willy Tarreau <w@1wt.eu>
net/ipv4/tcp_output.c