Closed Bug 512296 Opened 15 years ago Closed 15 years ago

mochitest-plain: timeout (all tests) in test_menuchecks.xul

Categories

(Core :: XUL, defect)

x86
Linux
defect
Not set
major

Tracking

()

RESOLVED FIXED
mozilla1.9.3a1
Tracking Status
status1.9.2 --- beta1-fixed

People

(Reporter: dbaron, Assigned: Natch)

References

Details

(Keywords: intermittent-failure, Whiteboard: [test which aborts the suite] )

Attachments

(1 file, 3 obsolete files)

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1251134962.1251140722.13052.gz
Linux mozilla-central unit test on 2009/08/24 10:29:22  

...
91512 INFO TEST-PASS | /tests/toolkit/content/tests/widgets/test_menubar.xul | Deactivate menubar with f10 key DOMMenuItemInactive fired
91514 INFO Running /tests/toolkit/content/tests/widgets/test_menuchecks.xul...

command timed out: 300 seconds without output, killing pid 11248
Whiteboard: [orange]
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1251192456.1251198718.6998.gz
Linux mozilla-central unit test on 2009/08/25 02:27:36
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1251302339.1251311021.1197.gz
Linux mozilla-central unit test on 2009/08/26 08:58:59
Summary: random orange: mochitest-plain timeout (all tests) in test_menuchecks.xul → mochitest-plain: timeout (all tests) in test_menuchecks.xul
Severity: normal → major
Whiteboard: [orange] → [test which aborts the suite] [orange]
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1251336413.1251342720.31267.gz
Linux mozilla-central unit test on 2009/08/26 18:26:53
Also 

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox-Unittest/1251337584.1251341368.16640.gz
Linux mozilla-central test mochitests on 2009/08/26 18:46:24

91580 ERROR TEST-UNEXPECTED-FAIL | /tests/toolkit/content/tests/widgets/test_menuchecks.xul | Test timed out.
91583 INFO Error: Unable to restore focus, expect failures and timeouts.
91584 ERROR TEST-UNEXPECTED-FAIL | /tests/toolkit/content/tests/widgets/test_menulist_keynav.xul | cursor down command event fired - got false, expected true
91585 ERROR TEST-UNEXPECTED-FAIL | /tests/toolkit/content/tests/widgets/test_menulist_keynav.xul | cursor down selectedItem - got [object XULElement], expected [object XULElement]
91586 ERROR TEST-UNEXPECTED-FAIL | /tests/toolkit/content/tests/widgets/test_menulist_keynav.xul | cursor down skip disabled command event fired - got false, expected true
leading to 
92764 ERROR TEST-UNEXPECTED-FAIL | (SimpleTest/TestRunner.js) | 4 test timeouts, giving up.
92765 ERROR TEST-UNEXPECTED-FAIL | (SimpleTest/TestRunner.js) | Skipping 36 remaining tests.

And 
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox-Unittest/1251338845.1251341963.23129.gz
Linux mozilla-central test mochitests on 2009/08/26 19:07:25
Attached patch maybe? (obsolete) (deleted) β€” β€” Splinter Review
Might be worth trying this, sending it to the tryserver now. This passes on my machine.
Attachment #396949 - Flags: review?(enndeakin)
This isn't likely to work once bug 506173 is fixed.
(In reply to comment #6)

In won't ever work any less than it works now, the reason I changed it is because this test is essentially just waiting for onload, not onfocus. It _wants_ to be the active window when the test runs, but onfocus doesn't help with that anyhow. It should be the active window regardless, I just put the |window.focus()| call in there as an attempt to set it as active. Once bug 506173 is fixed, there will inevitably have to be a fix for all other instances of this hack, this one can be included.

Tryresults: (Mac, with a warning, unrelated) http://tinderbox.mozilla.org/showlog.cgi?log=MozillaTry/1251346831.1251354655.32266.gz

(Win, clean) http://tinderbox.mozilla.org/showlog.cgi?log=MozillaTry/1251346831.1251353952.24753.gz

(Linux, clean) http://tinderbox.mozilla.org/showlog.cgi?log=MozillaTry/1251346831.1251353157.15987.gz
I don't understand. The test relies on the window being focused right? window.focus() should be asynchronous and doesn't guarantee that the window is active after it returns.
Comment on attachment 396949 [details] [diff] [review]
maybe?

Ah bug 506175 indeed! I was a bit confused about the implications of bug 506173. So, summing up, |window.focus()| is the correct way to activate a window, only it's asynchronous, and you have to wait for onfocus.
Attachment #396949 - Attachment is obsolete: true
Attachment #396949 - Flags: review?(enndeakin)
This is pretty active right now - can we get a fix together?

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox-Unittest/1251814846.1251818386.9832.gz&fulltext=1
Linux mozilla-central test mochitests on 2009/09/01 07:20:46
Attached patch don't open another window (obsolete) (deleted) β€” β€” Splinter Review
Is there any reason this has to be done with a separate window? Seems to run just fine on my machine.
Attachment #397889 - Flags: review?(enndeakin)
Comment on attachment 397889 [details] [diff] [review]
don't open another window

The basic idea of this change is ok as this test was written when the iframe displaying tests was only 100 pixels high, but now it is 300 pixels high which is enough space now.

The patch here doesn't cause the test to actually run though, as runTest isn't called.

Also, remove the margins on the hbox.

Finally, the test will need to be updated for bug 513707.
Attachment #397889 - Flags: review?(enndeakin) → review-
Attached patch rev. 2, comments addressed (obsolete) (deleted) β€” β€” Splinter Review
Comments 1 (stupid qref crap) & 2 addressed. Comment 3, not really sure what needs to be addressed. Bug 513707 is really only necessary for tests that open new windows, tests that don't open windows at all expect the focus to be on their window (the testing window, that is), where it should be. At least so it seemed to me.
Attachment #397889 - Attachment is obsolete: true
Attachment #397897 - Flags: review?(enndeakin)
Comment on attachment 397897 [details] [diff] [review]
rev. 2, comments addressed

>+  function runTest()
>+  {
>+    SimpleTest.waitForExplicitFinish();

waitForExplicitFinish should be called outside a function rather than in a load handler.
Attachment #397897 - Flags: review?(enndeakin) → review+
Attached patch patch [Checkin: Comment 18] (deleted) β€” β€” Splinter Review
Thanks. Hopefully this will quiet t-boxes a bit!
Assignee: nobody → highmind63
Attachment #397897 - Attachment is obsolete: true
Status: NEW → ASSIGNED
Keywords: checkin-needed
Linux mozilla-central unit test on 2009/08/31 18:03:34
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1251767014.1251777779.5139.gz

Natch: did you mean to request review on your patch? Would be good to quiet the t-boxen down!
Comment on attachment 397912 [details] [diff] [review]
patch
[Checkin: Comment 18]

Chris, this patch has r+ from Enn already, it's good to go, just need someone to check-in for me.
Attachment #397912 - Flags: review+
http://hg.mozilla.org/mozilla-central/rev/e7aa4e5ec9c0

Should we request 1.9.2 approval for this?
Status: ASSIGNED → RESOLVED
Closed: 15 years ago
Keywords: checkin-needed
Resolution: --- → FIXED
(In reply to comment #18)

I don't know, for the past 36 hours this hasn't shown up once on 1.9.2 tinderboxen, perhaps this needs to be investigated more thoroughly?

Does anyone have a timeframe of when this began? Does anyone know if this happened at all on 1.9.2?
Target Milestone: --- → mozilla1.9.3a1
Attachment #397912 - Attachment description: patch [for checkin] → patch [Checkin: Comment 18]
Flags: in-testsuite+
Here is a 1.9.2 instance:

Linux mozilla-1.9.2 test mochitests
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox3.6-Unittest/1252840440.1252844153.18232.gz
Flags: wanted1.9.2?
And another 1.9.2:

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox3.6-Unittest/1252922522.1252926470.29317.gz
Did something land recently on 1.9.2 that might cause this (focus code, window code, event code, etc.)?

Perhaps it will be easier to figure this out now by cross-referencing the 1.9.2 start and m-c start, unless this has been happening on 1.9.2 for a while and I've been missing it?
Comment on attachment 397912 [details] [diff] [review]
patch
[Checkin: Comment 18]

From the duped bug it seems this is happening pretty frequently on branch, requesting approval.
Attachment #397912 - Flags: approval1.9.2?
Test changes have blanket approval to land on branches. Land away!
Thanks. I'll add the keyword then.
Flags: wanted1.9.2?
Keywords: checkin-needed
Whiteboard: [test which aborts the suite] [orange] → [test which aborts the suite] [orange] [needs 192 landing]
Attachment #397912 - Flags: approval1.9.2?
http://hg.mozilla.org/releases/mozilla-1.9.2/rev/aa4e39e3394a
Keywords: checkin-needed
Whiteboard: [test which aborts the suite] [orange] [needs 192 landing] → [test which aborts the suite] [orange]
1.9.2 is burning. Ehsan, are you around?
backed out
Thanks for covering for me, DΓ£o, and sorry I caught this too late.

The problem was that I mistakenly removed test_menuchecks.xul instead of window_menuchecks.xul from Makefile.in.  Nochum, it'd be great if you can attach a patch which applies cleanly on 1.9.2.
I'll update this later today, guess I figured that the tests wouldn't bitrot or change...
The last patch applies perfectly on 1.9.2 as of now (just updated and tested and it applied fine, no fuzz).
Keywords: checkin-needed
Whiteboard: [test which aborts the suite] [orange] → [test which aborts the suite] [orange] [needs 192 landing]
Landed again on 1.9.2:

http://hg.mozilla.org/releases/mozilla-1.9.2/rev/2a496c6df8c2
Keywords: checkin-needed
Whiteboard: [test which aborts the suite] [orange] [needs 192 landing] → [test which aborts the suite] [orange]
Whiteboard: [test which aborts the suite] [orange] → [test which aborts the suite]
Component: XP Toolkit/Widgets: Menus → XUL
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: