net: filter: run cgroup eBPF ingress programs
authorDaniel Mack <daniel@zonque.org>
Wed, 23 Nov 2016 15:52:28 +0000 (16:52 +0100)
committerDavid S. Miller <davem@davemloft.net>
Fri, 25 Nov 2016 21:26:04 +0000 (16:26 -0500)
commitc11cd3a6ec3a817c6b71b00c559e25d855f7e5b4
tree5156c0612dd01bccfa91c89e3862b919f7ad9066
parentf4324551489e8781d838f941b7aee4208e52e8bf
net: filter: run cgroup eBPF ingress programs

If the cgroup associated with the receiving socket has an eBPF
programs installed, run them from sk_filter_trim_cap().

eBPF programs used in this context are expected to either return 1 to
let the packet pass, or != 1 to drop them. The programs have access to
the skb through bpf_skb_load_bytes(), and the payload starts at the
network headers (L3).

Note that cgroup_bpf_run_filter() is stubbed out as static inline nop
for !CONFIG_CGROUP_BPF, and is otherwise guarded by a static key if
the feature is unused.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/filter.c