Repository navigation
perf check: skip a call graph perf cannot record on the machine - #3711
Merged
Merged
Conversation
On the GitHub arm runner perf refuses --call-graph=dwarf with cpu-clock:u, saying the PMU cannot sample, so the aarch64 gdb and perf job stopped there with exit 255 after the fp rows passed. The script now records true once with each call graph first and skips one that fails with a SKIP line that gives perf's reason.
tamnd
added
kind/bug
11 tasks done
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The "linux gdb and perf on aarch64-linux-gnu" job has failed on every run of main. The fp rows pass, then perf refuses
--call-graph=dwarfwithcpu-clock:uon the GitHub arm runner ("PMU Hardware doesn't support sampling/overflow-interrupts") and the script stops with exit 255.tests/tools/perf.sh now records
trueonce with each call graph before the real runs, and a call graph that perf cannot record there is skipped with a SKIP line that gives perf's own reason. The fp comparison still runs and still fails the job when the cc stacks are worse.On the Linux box the script records both call graphs and passes as before. With a perf on PATH that refuses dwarf the same way the runner does, it prints the fp rows, then "SKIP dwarf: perf cannot record this call graph here: cpu-clock:uH: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat'", and exits 0.