mirror of
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git
synced 2025-01-01 10:42:11 +00:00
c741c7b5e9
The cpu-list part of this testcase has proven itself to be unreliable.
Sometimes, we get "<not counted>" for system.slice when pinned to CPUs
0 and 1. In such case, the test fails.
Since we cannot simply guarantee that any system.slice load will run
on any arbitrary list of CPUs, except the whole set of all CPUs, let's
rather remove the cpu-list subtest.
Fixes: a84260e314 ("perf test stat_bpf_counters_cgrp: Enhance perf stat cgroup BPF counter test")
Signed-off-by: Michael Petlan <mpetlan@redhat.com>
Cc: vmolnaro@redhat.com
Cc: bpf@vger.kernel.org
Link: https://lore.kernel.org/r/20241101102812.576425-1-mpetlan@redhat.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
67 lines
1.7 KiB
Bash
Executable File
67 lines
1.7 KiB
Bash
Executable File
#!/bin/sh
|
|
# perf stat --bpf-counters --for-each-cgroup test
|
|
# SPDX-License-Identifier: GPL-2.0
|
|
|
|
set -e
|
|
|
|
test_cgroups=
|
|
if [ "$1" = "-v" ]; then
|
|
verbose="1"
|
|
fi
|
|
|
|
# skip if --bpf-counters --for-each-cgroup is not supported
|
|
check_bpf_counter()
|
|
{
|
|
if ! perf stat -a --bpf-counters --for-each-cgroup / true > /dev/null 2>&1; then
|
|
if [ "${verbose}" = "1" ]; then
|
|
echo "Skipping: --bpf-counters --for-each-cgroup not supported"
|
|
perf --no-pager stat -a --bpf-counters --for-each-cgroup / true || true
|
|
fi
|
|
exit 2
|
|
fi
|
|
}
|
|
|
|
# find two cgroups to measure
|
|
find_cgroups()
|
|
{
|
|
# try usual systemd slices first
|
|
if [ -d /sys/fs/cgroup/system.slice ] && [ -d /sys/fs/cgroup/user.slice ]; then
|
|
test_cgroups="system.slice,user.slice"
|
|
return
|
|
fi
|
|
|
|
# try root and self cgroups
|
|
find_cgroups_self_cgrp=$(grep perf_event /proc/self/cgroup | cut -d: -f3)
|
|
if [ -z ${find_cgroups_self_cgrp} ]; then
|
|
# cgroup v2 doesn't specify perf_event
|
|
find_cgroups_self_cgrp=$(grep ^0: /proc/self/cgroup | cut -d: -f3)
|
|
fi
|
|
|
|
if [ -z ${find_cgroups_self_cgrp} ]; then
|
|
test_cgroups="/"
|
|
else
|
|
test_cgroups="/,${find_cgroups_self_cgrp}"
|
|
fi
|
|
}
|
|
|
|
# As cgroup events are cpu-wide, we cannot simply compare the result.
|
|
# Just check if it runs without failure and has non-zero results.
|
|
check_system_wide_counted()
|
|
{
|
|
check_system_wide_counted_output=$(perf stat -a --bpf-counters --for-each-cgroup ${test_cgroups} -e cpu-clock -x, sleep 1 2>&1)
|
|
if echo ${check_system_wide_counted_output} | grep -q -F "<not "; then
|
|
echo "Some system-wide events are not counted"
|
|
if [ "${verbose}" = "1" ]; then
|
|
echo ${check_system_wide_counted_output}
|
|
fi
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
check_bpf_counter
|
|
find_cgroups
|
|
|
|
check_system_wide_counted
|
|
|
|
exit 0
|