]> git.karo-electronics.de Git - linux-beck.git/commit
bpf: fix cb access in socket filter programs
authorAlexei Starovoitov <ast@plumgrid.com>
Wed, 7 Oct 2015 17:55:41 +0000 (10:55 -0700)
committerDavid S. Miller <davem@davemloft.net>
Sun, 11 Oct 2015 11:40:05 +0000 (04:40 -0700)
commitff936a04e5f28b7e0455be0e7fa91334f89e4b44
treef23b1f44945600ce2631560eb1f93364606b62b6
parentd49ae37c613f7cc1f4c1fd6ef073d60c32e000dd
bpf: fix cb access in socket filter programs

eBPF socket filter programs may see junk in 'u32 cb[5]' area,
since it could have been used by protocol layers earlier.

For socket filter programs used in af_packet we need to clean
20 bytes of skb->cb area if it could be used by the program.
For programs attached to TCP/UDP sockets we need to save/restore
these 20 bytes, since it's used by protocol layers.

Remove SK_RUN_FILTER macro, since it's no longer used.

Long term we may move this bpf cb area to per-cpu scratch, but that
requires addition of new 'per-cpu load/store' instructions,
so not suitable as a short term fix.

Fixes: d691f9e8d440 ("bpf: allow programs to write to certain skb fields")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
include/linux/bpf.h
include/linux/filter.h
kernel/bpf/verifier.c
net/core/filter.c
net/packet/af_packet.c