perf record: Fix BTF type checks in the off-cpu profiling
authorNamhyung Kim <namhyung@kernel.org>
Fri, 22 Sep 2023 23:44:44 +0000 (16:44 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 20 Nov 2023 10:52:10 +0000 (11:52 +0100)
commit1e03a269599588aeecae5b54c64a7aac4922c840
tree7a7cfc2d7f0b628876a618de0f08b993652beb1f
parent1c4eb1bc39d3090da7235f84164d753d20fdc93e
perf record: Fix BTF type checks in the off-cpu profiling

[ Upstream commit 0e501a65d35bf72414379fed0e31a0b6b81ab57d ]

The BTF func proto for a tracepoint has one more argument than the
actual tracepoint function since it has a context argument at the
begining.  So it should compare to 5 when the tracepoint has 4
arguments.

  typedef void (*btf_trace_sched_switch)(void *, bool, struct task_struct *, struct task_struct *, unsigned int);

Also, recent change in the perf tool would use a hand-written minimal
vmlinux.h to generate BTF in the skeleton.  So it won't have the info
of the tracepoint.  Anyway it should use the kernel's vmlinux BTF to
check the type in the kernel.

Fixes: b36888f71c85 ("perf record: Handle argument change in sched_switch")
Reviewed-by: Ian Rogers <irogers@google.com>
Acked-by: Song Liu <song@kernel.org>
Cc: Hao Luo <haoluo@google.com>
CC: bpf@vger.kernel.org
Link: https://lore.kernel.org/r/20230922234444.3115821-1-namhyung@kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
tools/perf/util/bpf_off_cpu.c