bpf: sockmap, fix skmsg recvmsg handler to track size correctly
authorJohn Fastabend <john.fastabend@gmail.com>
Tue, 16 Oct 2018 17:36:01 +0000 (10:36 -0700)
committerDaniel Borkmann <daniel@iogearbox.net>
Wed, 17 Oct 2018 00:29:15 +0000 (02:29 +0200)
commit3f4c3127d332000530349db4843deece27fe5e0c
treee9eb7a109cedfc88885cd074bf0364884402c1dc
parent9032c10e94fc44bb3245dc7f0d52aeeb3e8715b6
bpf: sockmap, fix skmsg recvmsg handler to track size correctly

When converting sockmap to new skmsg generic data structures we missed
that the recvmsg handler did not correctly use sg.size and instead was
using individual elements length. The result is if a sock is closed
with outstanding data we omit the call to sk_mem_uncharge() and can
get the warning below.

[   66.728282] WARNING: CPU: 6 PID: 5783 at net/core/stream.c:206 sk_stream_kill_queues+0x1fa/0x210

To fix this correct the redirect handler to xfer the size along with
the scatterlist and also decrement the size from the recvmsg handler.
Now when a sock is closed the remaining 'size' will be decremented
with sk_mem_uncharge().

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
include/linux/skmsg.h
net/ipv4/tcp_bpf.c