See: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#master=ChromiumWebkit&tests=canvas-arc-360-winding.html&showExpectations=true Diff: @@ -4,13 +4,13 @@ PASS data[0] is 0 -PASS data[1] is 0 +FAIL data[1] should be 0. Was 255. PASS data[2] is 0 PASS data[0] is 0 PASS data[1] is 255 PASS data[2] is 0 PASS data[0] is 0 -PASS data[1] is 0 +FAIL data[1] should be 0. Was 255. PASS data[2] is 0 PASS successfullyParsed is true
Forgot a link to the revision that added the test: http://trac.webkit.org/changeset/71936
We just render a big green rectangle on this in Skia instead of concentric circles.
any action here? as a developer i describe this bug as "chrome can't draw a torus shape with ctx.arc(), while firefox/ie9 can". this "testcase fails" description would indicate something's off by a pixel or philosophically wrong in an obscure case, but this is a bad bug! a common-case-is-visibly-cattywampus bug! i claim the root cause is in WebCore/platform/graphics/skia/PathSkia.cpp -- Path::addArc just adds an oval if sweepDegrees covers a whole circle. it disregards anticlockwise and the sign of sweepDegrees in this case.
http://trac.webkit.org/export/80302/trunk/LayoutTests/fast/canvas/canvas-arc-360-winding.html
Created attachment 84737 [details] respect anticlockwise in Path::addArc()
holy crap, that was the fastest response from any dev team i've ever seen. thanks :)
Comment on attachment 84737 [details] respect anticlockwise in Path::addArc() Looks sane, r=me
Comment on attachment 84737 [details] respect anticlockwise in Path::addArc() View in context: https://bugs.webkit.org/attachment.cgi?id=84737&action=review > WebCore/ChangeLog:9 > + LayoutTests/fast/canvas/canvas-arc-360-winding.html Shouldn't the TEXT entry for this in LayoutTests/platform/chromium/test_expectations.txt be removed, now that we expect it to pass?
(In reply to comment #8) > (From update of attachment 84737 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=84737&action=review > > > WebCore/ChangeLog:9 > > + LayoutTests/fast/canvas/canvas-arc-360-winding.html > > Shouldn't the TEXT entry for this in LayoutTests/platform/chromium/test_expectations.txt be removed, now that we expect it to pass? D'oh. Yes of course. @Mike: Please update the patch accordingly.
Created attachment 84773 [details] Patch
This time with the updated test_expectations.txt
Comment on attachment 84773 [details] Patch Rejecting attachment 84773 [details] from commit-queue. Failed to run "['./Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '--bot-id=cr-jail-7', 'build-..." exit_code: 2 Last 500 characters of output: ....................................................................................................................................................................................................................... inspector ........... inspector/audits . inspector/audits/audits-panel-functional.html -> failed Exiting early after 1 failures. 12758 tests run. 299.28s total testing time 12757 test cases (99%) succeeded 1 test case (<1%) had incorrect layout 8 test cases (<1%) had stderr output Full output: http://queues.webkit.org/results/8087853
Comment on attachment 84773 [details] Patch Flake
The commit-queue encountered the following flaky tests while processing attachment 84773 [details]: inspector/audits/audits-panel-functional.html bug 55776 (authors: apavlov@chromium.org, pfeldman@chromium.org, and rniwa@webkit.org) The commit-queue is continuing to process your patch.
Comment on attachment 84773 [details] Patch Clearing flags on attachment: 84773 Committed r80409: <http://trac.webkit.org/changeset/80409>
All reviewed patches have been landed. Closing bug.
Cool!