perf/tests: Fix tests 84 and 86 Add --metric-only on s390
Subsystems (2)
PATCH 1/1 perf/tests: Fix tests 84 and 86 Add --metric-only on s390
On s390x KVM and z/VM machines the CPU Measurement Facility is not available. Events cycles and instructions do not exist. Running above tests on s390 KVM and z/VM guests always fail with this error:
# ./perf test 84 86 84: perf stat JSON output linter : FAILED! 86: perf stat STD output linter : FAILED! #
Root cause is command
# perf stat -j --metric-only -e instructions,cycles -- true
{"metric-value" : "none"}
#
which fails due to unsupported events and returns "none".
Do not execute this test case on s390 KVM and z/VM machines.Output after:
# ./perf test 84 86 84: perf stat JSON output linter : Ok 86: perf stat STD output linter : Ok #
--- tools/perf/tests/shell/lib/stat_output.sh | 5 +++++ tools/perf/tests/shell/stat+json_output.sh | 5 +++++ 2 files changed, 10 insertions(+)
tools/perf/tests/shell/lib/stat_output.sh +5 -0
@@ -151,6 +151,11 @@ check_per_socket() check_metric_only() { echo -n "Checking $1 output: metric only " + if [ "$(uname -m)" = "s390x" ] && ! grep '^facilities' /proc/cpuinfo | grep -qw 67 + then + echo "[Skip] CPU-measurement counter facility not installed" + return + fi perf stat --metric-only $2 -e instructions,cycles true commachecker --metric-only echo "[Success]"
tools/perf/tests/shell/stat+json_output.sh +5 -0
@@ -176,6 +176,11 @@ check_per_socket() check_metric_only() { echo -n "Checking json output: metric only " + if [ "$(uname -m)" = "s390x" ] && ! grep '^facilities' /proc/cpuinfo | grep -qw 67 + then + echo "[Skip] CPU-measurement counter facility not installed" + return + fi perf stat -j --metric-only -e instructions,cycles -o "${stat_output}" true $PYTHON $pythonchecker --metric-only --file "${stat_output}" echo "[Success]"
Signature · 1 line
2.49.0
Quoted 6 lines · Thomas Richter
Fixes 45a86d017adf ("perf test: Add --metric-only to perf stat output tests")Signed-off-by Thomas Richter <[email protected]>Suggested-by Sumanth Korikkar <[email protected]>Suggested-by Heiko Carstens <[email protected]>Reviewed-by Sumanth Korikkar <[email protected]>Cc Namhyung Kim <[email protected]>
Thanks, applied to perf-tools-next.
But please try to look at how patches are applied, specifically in how commit log messages are rewritten, what is modified in the commit log messages.
Specifically: When we do a 'git log --oneline' what we see should help we find whatever we're trying to find. Twitter (not-X) style.
While I agree with Namhyung that whatever reduces the work a maintainer should (have to) care about, and doing this is just some muscle memory from sending patches to Ingo, I do think that trying to be consistent on how we describe the problem, how the solution being proposed fixes the problem, and then, when that is read, and the code is read, all matches, bingo, patch accepted, tests pass, lets focus on the next issue.
This is not something aimed at you, but its something that takes time when I'm processing patches, maybe I should just ignore this if the code is good enough (I do this more than I want), but I think getting this out of my mind is important.
Lowering the bar invites more people to contribute, but then it bites us later.
- Arnaldo