The following layout test is failing on Mac and Win debug:
This test has been flaky before, see http://trac.webkit.org/changeset/94634
Committed r119277: <http://trac.webkit.org/changeset/119277>
Committed expectations update.
Do you have a stack trace to post?
(In reply to comment #3)
> Do you have a stack trace to post?
Unfortunately Chromium bots delete symbols in WebKit/WebCore so we can't see anything useful. We should really change that though... it makes impossible to diagnose the problems :(
Was this crashing? If it was just failing.
I could make test result more stable somehow.
Created attachment 192277 [details]
I believe this patch will de-flake the test. Without this patch, locally (cr linux debug) the test fails 6-20 times when run with --repeat-each=10 --iterations=10. With this patch is passes 100/100 times.
Comment on attachment 192277 [details]
Should there be a TestExpectations change to re-enable this test?
(In reply to comment #8)
> (From update of attachment 192277 [details])
> Should there be a TestExpectations change to re-enable this test?
I accidentally landed that part already in: http://trac.webkit.org/changeset/145259/trunk/LayoutTests/platform/chromium/TestExpectations
Committed r145263: <http://trac.webkit.org/changeset/145263>