All the tests under https://dvcs.w3.org/hg/html/file/1d30da826739/tests/approved/canvas need to be imported to webkit. The failed ones will be added to the expected failed tests list. Individual bugs to be added for each failed test.
Created attachment 241462 [details] Patch
Two bugs were logged to track the tests which failed and the tests which need to be modified such that they can be automatically verified. https://bugs.webkit.org/show_bug.cgi?id=138672 https://bugs.webkit.org/show_bug.cgi?id=138676
Created attachment 241497 [details] Patch
Comment on attachment 241497 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=241497&action=review rs=me for the actual tests > LayoutTests/TestExpectations:219 > +# Imported W3C canvas failed tests > +webkit.org/b/138672 imported/w3c/canvas/2d.composite.uncovered.pattern.destination-atop.html [ Skip ] > +webkit.org/b/138672 imported/w3c/canvas/2d.pattern.crosscanvas.html [ Skip ] Do these fail everywhere consistently? If so, they should probably be [ Failure ] rather than [ Skip ] (this means we'll notice progressions) > LayoutTests/TestExpectations:331 > +# Imported W3C canvas tests need to be changed such that they can be utomatically verifid Typo in automatically. > LayoutTests/imported/w3c/ChangeLog:1665 > + * common/canvas-frame.css: Added. > + (body): > + (p): > + (p:first-child): > + (h1): I usually remove all the cruft under CSS file changes. It isn't always accurate. Just having the file name is enough in this case.
Created attachment 241513 [details] Patch
(In reply to comment #4) > Comment on attachment 241497 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=241497&action=review > > rs=me for the actual tests > > > LayoutTests/TestExpectations:219 > > +# Imported W3C canvas failed tests > > +webkit.org/b/138672 imported/w3c/canvas/2d.composite.uncovered.pattern.destination-atop.html [ Skip ] > > +webkit.org/b/138672 imported/w3c/canvas/2d.pattern.crosscanvas.html [ Skip ] > > Do these fail everywhere consistently? > > If so, they should probably be [ Failure ] rather than [ Skip ] > > (this means we'll notice progressions) > Done. I changed [ Skip ] with [ Failure ] and I added expected files for the failed tests. > > LayoutTests/TestExpectations:331 > > +# Imported W3C canvas tests need to be changed such that they can be utomatically verifid > > Typo in automatically. > Fixed. > > LayoutTests/imported/w3c/ChangeLog:1665 > > + * common/canvas-frame.css: Added. > > + (body): > > + (p): > > + (p:first-child): > > + (h1): > > I usually remove all the cruft under CSS file changes. It isn't always > accurate. Just having the file name is enough in this case. Done.
Comment on attachment 241513 [details] Patch rs=me
Comment on attachment 241513 [details] Patch Rejecting attachment 241513 [details] from commit-queue. Number of test failures exceeded the failure limit. Full output: http://webkit-queues.appspot.com/results/5383446117482496
Created attachment 241526 [details] Archive of layout-test-results from webkit-cq-02 for mac-mountainlion The attached test failures were seen while running run-webkit-tests on the commit-queue. Bot: webkit-cq-02 Port: mac-mountainlion Platform: Mac OS X 10.8.5
Comment on attachment 241513 [details] Patch Attachment 241513 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/6677766724386816 Number of test failures exceeded the failure limit.
Created attachment 241527 [details] Archive of layout-test-results from webkit-ews-09 for mac-mountainlion-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: webkit-ews-09 Port: mac-mountainlion-wk2 Platform: Mac OS X 10.8.5
Comment on attachment 241513 [details] Patch Attachment 241513 [details] did not pass mac-ews (mac): Output: http://webkit-queues.appspot.com/results/6636305257594880 New failing tests: imported/w3c/canvas/2d.line.join.parallel.html imported/w3c/canvas/2d.strokeRect.zero.5.html imported/w3c/canvas/2d.gradient.interpolate.colouralpha.html
Created attachment 241550 [details] Archive of layout-test-results from webkit-ews-02 for mac-mountainlion The attached test failures were seen while running run-webkit-tests on the mac-ews. Bot: webkit-ews-02 Port: mac-mountainlion Platform: Mac OS X 10.8.5
Created attachment 241598 [details] Patch
I added the failed tests to the TestExpectations because they seem to be flaky. They have been passing on my machine but they fail on Bot. I will investigate them with the other failures.
Created attachment 241612 [details] Patch
My source tree was outdated. I have to sync it and resubmit a new patch.
Comment on attachment 241612 [details] Patch the EWS is now passing. This patch should be ready to go.
Comment on attachment 241612 [details] Patch I'll land it manually.
Comment on attachment 241612 [details] Patch Clearing flags on attachment: 241612 Committed r176144: <http://trac.webkit.org/changeset/176144>
All reviewed patches have been landed. Closing bug.
Many of these tests use non-existent background.png: https://build.webkit.org/results/Apple%20Yosemite%20Debug%20WK2%20(Tests)/r176694%20(747)/imported/w3c/canvas/2d.imageData.put.null-stderr.txt Adding it will make tests a lot less noisy, faster, and presumably more of them will pass.
(In reply to comment #22) > Many of these tests use non-existent background.png: > > https://build.webkit.org/results/Apple%20Yosemite%20Debug%20WK2%20(Tests)/ > r176694%20(747)/imported/w3c/canvas/2d.imageData.put.null-stderr.txt > > Adding it will make tests a lot less noisy, faster, and presumably more of > them will pass. Moving those tests to imported/w3c/web-platform-tests should take care of those issues.
(In reply to comment #23) > (In reply to comment #22) > > Many of these tests use non-existent background.png: > > > > https://build.webkit.org/results/Apple%20Yosemite%20Debug%20WK2%20(Tests)/ > > r176694%20(747)/imported/w3c/canvas/2d.imageData.put.null-stderr.txt > > > > Adding it will make tests a lot less noisy, faster, and presumably more of > > them will pass. > > Moving those tests to imported/w3c/web-platform-tests should take care of > those issues. This issue has been fixed in https://bugs.webkit.org/show_bug.cgi?id=140712. The problem was in not adopting the same directory structure as w3c did. The css file imported/w3c/common/canvas-tests.css was referencing an image from imported/w3c/resources/background.png while it was in imported/w3c/canvas/resources/backgorund.png. The fix was to move the common resources from imported/w3c/canvas/resources/ to imported/w3c/resources/.
(In reply to comment #24) > (In reply to comment #23) > > (In reply to comment #22) > > > Many of these tests use non-existent background.png: > > > > > > https://build.webkit.org/results/Apple%20Yosemite%20Debug%20WK2%20(Tests)/ > > > r176694%20(747)/imported/w3c/canvas/2d.imageData.put.null-stderr.txt > > > > > > Adding it will make tests a lot less noisy, faster, and presumably more of > > > them will pass. > > > > Moving those tests to imported/w3c/web-platform-tests should take care of > > those issues. > > This issue has been fixed in https://bugs.webkit.org/show_bug.cgi?id=140712. > The problem was in not adopting the same directory structure as w3c did. The > css file imported/w3c/common/canvas-tests.css was referencing an image from > imported/w3c/resources/background.png while it was in > imported/w3c/canvas/resources/backgorund.png. The fix was to move the common > resources from imported/w3c/canvas/resources/ to imported/w3c/resources/. Oh, good. There may be still some similar issues though, like the fonts folder. Checking a few canvas tests listed in bug 138676, they seem to pass when imported using import-w3c-test in imported/w3c/web-platform-tests. That may be worth a try?