bpf, trace: add BPF_F_CURRENT_CPU flag for bpf_perf_event_read
authorDaniel Borkmann <daniel@iogearbox.net>
Tue, 28 Jun 2016 10:18:25 +0000 (12:18 +0200)
committerDavid S. Miller <davem@davemloft.net>
Thu, 30 Jun 2016 09:54:40 +0000 (05:54 -0400)
commit6816a7ffce32e999601825ddfd887f36d3052932
tree99a35abec2ab665d9cd72ae8f143b544d5eef923
parentd79313303181d357d293453fb8486bdc87bfd2f4
bpf, trace: add BPF_F_CURRENT_CPU flag for bpf_perf_event_read

Follow-up commit to 1e33759c788c ("bpf, trace: add BPF_F_CURRENT_CPU
flag for bpf_perf_event_output") to add the same functionality into
bpf_perf_event_read() helper. The split of index into flags and index
component is also safe here, since such large maps are rejected during
map allocation time.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
include/uapi/linux/bpf.h
kernel/trace/bpf_trace.c