71653 2011-11-07 00:00:48 -0800 [Chromium] Layout Test platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html is flaky 2013-04-09 16:28:24 -0700 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED WONTFIX http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=platform%2Fchromium-cg-mac%2Fediting%2Finput%2Fime-candidate-window-position.html P2 Normal --- 1 yutak bashi bashi enrica morrita rniwa tkent webkit.review.bot oldest_to_newest 497002 0 yutak 2011-11-07 00:00:48 -0800 The following layout test is flaky (pass / timeout) on all platforms. platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html See http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=platform%2Fchromium-cg-mac%2Fediting%2Finput%2Fime-candidate-window-position.html Probable cause: Unknown, as it's not clear when this flakiness started. This test is less flaky on DEPS bots. 497014 1 bashi 2011-11-07 01:18:24 -0800 Umm, I don't have any idea. Maybe recent editing changes trigger this flaky? 497219 2 rniwa 2011-11-07 09:03:26 -0800 My suspicion is that doTest() is called before iframe is loaded, and doTest is encountering some javascript error and end up not calling finishJSTest. We should probably move finishJSTest above if with setTimeout(~, 0) so that we can see the diff. 497709 3 113983 bashi 2011-11-07 20:58:02 -0800 Created attachment 113983 Patch 497710 4 bashi 2011-11-07 21:00:16 -0800 (In reply to comment #2) > My suspicion is that doTest() is called before iframe is loaded, and doTest is encountering some javascript error and end up not calling finishJSTest. > > We should probably move finishJSTest above if with setTimeout(~, 0) so that we can see the diff. It seems you are right. I changed the test to call doTest() after iframe is loaded and iterated layout test 10 times, but no timeout occurred. 497712 5 113983 tkent 2011-11-07 21:03:01 -0800 Comment on attachment 113983 Patch ok, let's land this and watch the dashboard. 497713 6 rniwa 2011-11-07 21:03:48 -0800 (In reply to comment #4) > It seems you are right. I changed the test to call doTest() after iframe is loaded and iterated layout test 10 times, but no timeout occurred. Could you reproduce the issue (intermittent failure) without your patch? It's still odd that load event is fired before the iframe is loaded though. HTML5 spec is explicit about delaying load event until the iframe's content is loaded. Adam, do you know anything about this? 497756 7 113983 webkit.review.bot 2011-11-07 22:50:09 -0800 Comment on attachment 113983 Patch Clearing flags on attachment: 113983 Committed r99524: <http://trac.webkit.org/changeset/99524> 497757 8 webkit.review.bot 2011-11-07 22:50:13 -0800 All reviewed patches have been landed. Closing bug. 497841 9 tkent 2011-11-08 01:27:16 -0800 (In reply to comment #7) > Committed r99524: <http://trac.webkit.org/changeset/99524> It seems this didn't fix the issue. 497845 10 rniwa 2011-11-08 01:33:55 -0800 Hm... http://trac.webkit.org/browser/trunk/LayoutTests/platform/chromium-cg-mac/editing/resources/ime-candidate-window-position-iframe.html and http://trac.webkit.org/browser/trunk/LayoutTests/platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html#L27 both try to focus the same node. So maybe there's some race condition there? Also, can we do setTimeout(finishJSTest, 0) before if statement so that we can at least see the diff if the test were to fail? 872326 11 schenney 2013-04-09 16:28:24 -0700 Marking test failures as WontFix. Bug is still accessible and recording in TestExpectations. 113983 2011-11-07 20:58:02 -0800 2011-11-07 22:50:09 -0800 Patch bug-71653-20111108135801.patch text/plain 2217 bashi U3VidmVyc2lvbiBSZXZpc2lvbjogOTk1MTUKZGlmZiAtLWdpdCBhL0xheW91dFRlc3RzL0NoYW5n ZUxvZyBiL0xheW91dFRlc3RzL0NoYW5nZUxvZwpvbGQgbW9kZSAxMDA3NTUKbmV3IG1vZGUgMTAw NjQ0CmluZGV4IDE5ZDdlZjBiZDMzMmRkYzIzNGQyNjQ1NzhhNDkwYmIxZmNlYzc2ZmEuLmVmZjBm ODFiYTg0NzUzNzljZTBiNzA5ZGIxMjE0ZjllN2YxMWRkNGIKLS0tIGEvTGF5b3V0VGVzdHMvQ2hh bmdlTG9nCisrKyBiL0xheW91dFRlc3RzL0NoYW5nZUxvZwpAQCAtMSwzICsxLDEyIEBACisyMDEx LTExLTA3ICBLZW5pY2hpIElzaGliYXNoaSAgPGJhc2hpQGNocm9taXVtLm9yZz4KKworICAgICAg ICBbQ2hyb21pdW1dIExheW91dCBUZXN0IHBsYXRmb3JtL2Nocm9taXVtLWNnLW1hYy9lZGl0aW5n L2lucHV0L2ltZS1jYW5kaWRhdGUtd2luZG93LXBvc2l0aW9uLmh0bWwgaXMgZmxha3kKKyAgICAg ICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTcxNjUzCisKKyAgICAg ICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgKiBwbGF0Zm9ybS9jaHJv bWl1bS1jZy1tYWMvZWRpdGluZy9pbnB1dC9pbWUtY2FuZGlkYXRlLXdpbmRvdy1wb3NpdGlvbi5o dG1sOiBDYWxscyBkb1Rlc3QoKSBhZnRlciBpZnJhbWUgaXMgbG9hZGVkLgorCiAyMDExLTExLTA3 ICBDaHJpcyBGbGVpemFjaCAgPGNmbGVpemFjaEBhcHBsZS5jb20+CiAKICAgICAgICAgUkVHUkVT U0lPTihyOTkwMjUtcjk5MDM3KTogcGxhdGZvcm0vbWFjL2FjY2Vzc2liaWxpdHkvc2VsZWN0LWVs ZW1lbnQtc2VsZWN0aW9uLXdpdGgtb3B0Z3JvdXBzLmh0bWwgaXMgYXNzZXJ0aW5nCmRpZmYgLS1n aXQgYS9MYXlvdXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1bS1jZy1tYWMvZWRpdGluZy9pbnB1dC9p bWUtY2FuZGlkYXRlLXdpbmRvdy1wb3NpdGlvbi5odG1sIGIvTGF5b3V0VGVzdHMvcGxhdGZvcm0v Y2hyb21pdW0tY2ctbWFjL2VkaXRpbmcvaW5wdXQvaW1lLWNhbmRpZGF0ZS13aW5kb3ctcG9zaXRp b24uaHRtbAppbmRleCA3YzRjMWQ0ZTg0Yzc1ZTZiMmIzMDUzNGQzNzc3YmQ5ZDQyN2IxZWY4Li5h NzE2YTJkMzEwMzkwMWYxMTc5MDU3OGQ4NjI5MWRjNWUxZTI1YTU1IDEwMDY0NAotLS0gYS9MYXlv dXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1bS1jZy1tYWMvZWRpdGluZy9pbnB1dC9pbWUtY2FuZGlk YXRlLXdpbmRvdy1wb3NpdGlvbi5odG1sCisrKyBiL0xheW91dFRlc3RzL3BsYXRmb3JtL2Nocm9t aXVtLWNnLW1hYy9lZGl0aW5nL2lucHV0L2ltZS1jYW5kaWRhdGUtd2luZG93LXBvc2l0aW9uLmh0 bWwKQEAgLTI3LDEyICsyNywxMiBAQCBmdW5jdGlvbiBkb1Rlc3QoKSB7CiB9CiA8L3NjcmlwdD4K IDwvaGVhZD4KLTxib2R5IG9ubG9hZD0iZG9UZXN0KCkiPgorPGJvZHk+CiA8ZGl2PgogVGhpcyB0 ZXN0IGVuc3VyZXMgdGhhdCB0aGUgSU1FIGNhbmRpZGF0ZSB3aW5kb3cgYXBwZWFycyBhdCB0aGUg YXBwcm9wcmlhdGUgcG9zaXRpb24uICBFbnRlciBzb21lIHRleHQgd2l0aCBJTUUgaW4gdGhlIGZv bGxvd2luZyBncmF5IGJveCBhbmQgY29udmVydCB0aGVtLiAgVGhlbiBjaGVjayB3aGV0aGVyIHRo ZSBjYW5kaWRhdGUgd2luZG93IGFwcGVhcnMgdGhlIGJlbG93IG9mIHRoZSBjb21wb3NpdGlvbiB0 ZXh0LiAgSW4gRFJULCB0aGlzIHRlc3QgY2hlY2tzIGl0IGJ5IGNhbGxpbmcgdGV4dElucHV0Q29u dHJvbGxlci5maXJzdFJlY3RGb3JDaGFyYWN0ZXJSYW5nZSgpLgogPC9kaXY+CiA8ZGl2IGlkPSJj b25zb2xlIj48L2Rpdj4KLTxpZnJhbWUgbmFtZT0iaWZyYW1lIiBzdHlsZT0icG9zaXRpb246IGFi c29sdXRlOyB0b3A6IDIwMHB4OyBsZWZ0OiAyMDBweDsiIHNyYz0iLi4vcmVzb3VyY2VzL2ltZS1j YW5kaWRhdGUtd2luZG93LXBvc2l0aW9uLWlmcmFtZS5odG1sIiB3aWR0aD0iMzAwIiBoZWlnaHQ9 IjMwMCI+Cis8aWZyYW1lIG5hbWU9ImlmcmFtZSIgc3R5bGU9InBvc2l0aW9uOiBhYnNvbHV0ZTsg dG9wOiAyMDBweDsgbGVmdDogMjAwcHg7IiBzcmM9Ii4uL3Jlc291cmNlcy9pbWUtY2FuZGlkYXRl LXdpbmRvdy1wb3NpdGlvbi1pZnJhbWUuaHRtbCIgd2lkdGg9IjMwMCIgaGVpZ2h0PSIzMDAiIG9u bG9hZD0iZG9UZXN0KCkiPgogPC9pZnJhbWU+CiA8c2NyaXB0IHNyYz0iLi4vLi4vLi4vLi4vZmFz dC9qcy9yZXNvdXJjZXMvanMtdGVzdC1wb3N0LmpzIj48L3NjcmlwdD4KIDwvYm9keT4K