perf record: Switch user option to use BPF filter
authorIan Rogers <irogers@google.com>
Wed, 4 Jun 2025 17:45:38 +0000 (10:45 -0700)
committerNamhyung Kim <namhyung@kernel.org>
Mon, 9 Jun 2025 18:18:17 +0000 (11:18 -0700)
commit1151208e702267ad1ce2f24aa9d21deb47fa17f9
tree4bf5e1d3d56114cf67eb58a4be026d1336d2b82d
parent466db4275edd35b7a9af7c82575bcb3289f2c9c0
perf record: Switch user option to use BPF filter

Finding user processes by scanning /proc is inherently racy and
results in perf_event_open failures. Use a BPF filter to drop samples
where the uid doesn't match. Ensure adding the BPF filter forces
system-wide.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250604174545.2853620-5-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
tools/perf/builtin-record.c