Closed Bug 916143 Opened 12 years ago Closed 7 years ago

Intermittent Windows test_movement_by_words.html | Test timed out, formerly Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 (and many more)

Categories

(Core :: Layout, defect, P5)

Product:
Core
Component:
Layout
Platform:
x86
Windows 8
Type:
defect

Tracking

()

Status:
RESOLVED INCOMPLETE
Milestone:
mozilla28

People

(Reporter: emorley, Unassigned)

Details

(Keywords: intermittent-failure, Whiteboard: [test disabled on Windows] [leave open])

Attachments

(2 files)

SimpleTest.waitForFocus might help...
1.33 KB, patch
Details | Diff | Splinter Review
screenshot from the log in comment 334
288.10 KB, image/png
Details
WINNT 6.2 fx-team pgo test mochitest-5 on 2013-09-12 15:31:09 PDT for push a23d608fb893 slave: t-w864-ix-087 https://tbpl.mozilla.org/php/getParsedLog.php?id=27793502&tree=Fx-Team { 15:35:32 INFO - 1667 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1668 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1669 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1670 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1671 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1672 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1673 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "漢字、漢字。漢字"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1674 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1675 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 15:35:32 INFO - 1676 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1677 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 11 15:35:32 INFO - 1678 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1679 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 15:35:32 INFO - 1680 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1681 INFO TEST-PASS | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] 15:35:32 INFO - 1682 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got [object Text], expected [object Text] 15:35:32 INFO - 1683 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got 0, expected 1 15:35:32 INFO - 1684 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got [object Text], expected [object Text] 15:35:32 INFO - 1685 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Right movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got 0, expected 6 15:35:32 INFO - 1686 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got [object Text], expected [object Text] 15:35:32 INFO - 1687 ERROR TEST-UNEXPECTED-FAIL | /tests/layout/generic/test/test_movement_by_words.html | Left movement broken with eatSpace=undefined in "<b>Hello</b> Kitty"; sel.anchorNode.parentNode=[object HTMLElement] - got 0, expected 1 ... ... ... }
Based on the first comments in this bug, the regression point seems to be an ancestor of a23d608fb893. hg log -r 'reverse(ancestors(a23d608fb893))' may be useful Also noting that this is Windows-only.
Keywords: regressionwindow-wanted
Summary: Intermittent test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 (and many more) → Intermittent Windows test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 (and many more)
Priority: -- → P5
Mats, this is a top orange - could you take a look (especially given the regression-range clue in the previous comments)? Would like to avoid disabling this if possible.
Flags: needinfo?(matspal)
Flags: needinfo?(matspal)
(In reply to Mats Palmgren (:mats) from comment #196) > Created attachment 823682 [details] [diff] [review] > SimpleTest.waitForFocus might help... Thank you for this :-) Did you mean to set a review on it?
Flags: needinfo?(matspal)
Not really, it's just a guess it might fix it... I guess someone needs to take the time to submit it to Try and trigger a bunch of test runs to see if that is the case. If so, then I think we can just land it - I don't think we need to review trivial test-only fixes like this.
Flags: needinfo?(matspal)
(In reply to Mats Palmgren (:mats) from comment #240) > Not really, it's just a guess it might fix it... > I guess someone needs to take the time to submit it to Try and trigger a > bunch of > test runs to see if that is the case. If so, then I think we can just land > it - > I don't think we need to review trivial test-only fixes like this. Is there any reason we can't just push this and see what happens?
(In reply to Ryan VanderMeulen [:RyanVM UTC-5] from comment #289) > Is there any reason we can't just push this and see what happens? Works for me. FYI, this test is not run on Linux in case you intend to trigger a bunch of test runs...
Despite being Win8 PGO only, this is currently our number 4 orange, so something needs to happen here, fix or disable.
And given those choices, pushed in https://hg.mozilla.org/integration/mozilla-inbound/rev/095e5b582f53 and if it doesn't fix it we can always reopen and disable the test.
So on the bright side, that makes it a bit less verbose...
https://hg.mozilla.org/mozilla-central/rev/095e5b582f53
Assignee: nobody → matspal
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla28
Guess I should have thought to [leave open] for the new abbreviated version (or rather, the disabling).
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
The timeout is likely caused by that file explorer window preventing Firefox from gaining focus. It was likely opened by some earlier test.
Assignee: matspal → nobody
The explorer window seems to be present in every screenshot, I've been assuming it's either a releng relic or a Windows relic. How it manages to get above the browser sometimes, I don't know.
Well, every screenshot where it's not behind the browser, like the 330 second timeout ones where we take the screenshot too late, after the browser has already been killed.
Summary: Intermittent Windows test_movement_by_words.html | Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 (and many more) → Intermittent Windows test_movement_by_words.html | Test timed out, formerly Right movement broken with eatSpace=undefined in "Hello Kitty"; sel.anchorNode.parentNode=[object HTMLDivElement] - got 0, expected 6 (and many more)
Fwiw, I ran mochitest-plain locally on Win7 but no file explorer window was opened as far as can see. Perhaps we could hook up one of the Win8 boxes with a monitor and record it with a video camera to figure out which test caused the window to open? Is there a simpler way?
First simplest way is to ask armenzg "hey, do you guys just have a file explorer window open on every t-w864-ix- slave, fresh from the imaging and before any tests run?"
Flags: needinfo?(armenzg)
Philor! You should stop just knowing which question to ask!! :P Yes, we do. It is a side effect of the setup that had to be placed to "login into the desktop rather than the Metro UI". Q set this up and would have more information. Perhaps we should consider upgrading to 8.1 if it fixes the issue.
Boom! Phil is spot on yet again! :-)
Flags: needinfo?(armenzg)
I will remove myself from the CC list. Please file a bug for Releng when you know what we should be doing. I assume that removing the explorer window would be great, however, IIRC I asked Q already (a while ago) about it and it was not possible.
Disabling on Windows for too many intermittent failures (can't see an obvious way to narrow for Win8 PGO only; feel free to target more specifically if that is possible).
Whiteboard: [test disabled on Windows] [leave open]
Mass-closing old bugs I filed that have not had recent activity/no longer affect me.
Status: REOPENED → RESOLVED
Closed: 11 years ago → 7 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: