| Summary: | detectFailures in run-jsc-stress-tests can report everything was ok if too many tests fail | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Mark Hahnenberg <mhahnenberg> | ||||||
| Component: | JavaScriptCore | Assignee: | Mark Hahnenberg <mhahnenberg> | ||||||
| Status: | RESOLVED FIXED | ||||||||
| Severity: | Normal | CC: | commit-queue | ||||||
| Priority: | P2 | ||||||||
| Version: | 528+ (Nightly build) | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Attachments: |
|
||||||||
Created attachment 226050 [details]
Patch
Created attachment 226067 [details]
Patch
Comment on attachment 226067 [details] Patch Clearing flags on attachment: 226067 Committed r165239: <http://trac.webkit.org/changeset/165239> All reviewed patches have been landed. Closing bug. |
This is due to the following line: output = sshRead("cd #{$remoteDirectory}/#{$outputDir.basename}/.runner && (ls test_fail_* 2> /dev/null || true)") ls test_fail_* can fail if there are too many files that match. We then swallow that error, which makes everything seem ok. The fix is to use find instead of ls like we do in shell-runner.sh.