Closed Bug 595062 Opened 14 years ago Closed 13 years ago

Intermittent mochitest-ipcplugins failures: test_cocoa_focus.html | Test timed out then multiple failures in test_cocoa_window_focus.html

Categories

(Core Graveyard :: Plug-ins, defect)

All
macOS
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: philor, Assigned: mounir)

References

Details

(Keywords: intermittent-failure)

Attachments

(3 files, 2 obsolete files)

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1284087956.1284089580.4390.gz#err1
Rev3 MacOSX Snow Leopard 10.6.2 mozilla-central opt test mochitest-other on 2010/09/09 20:05:56
s: talos-r3-snow-018

33 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_focus.html | Test timed out.
37 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Activation state should be: activated - got false, expected true
39 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Activation state should be: activated - got false, expected true
42 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Window focus event count should be 2 - got 1, expected 2
44 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Window focus event count should be 2 - got 1, expected 2
45 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Activation state should be: activated - got false, expected true
46 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Window focus event count should be 3 - got 1, expected 3
47 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Activation state should be: activated - got false, expected true
48 ERROR TEST-UNEXPECTED-FAIL | /tests/modules/plugin/test/test_cocoa_window_focus.html | Window focus event count should be 3 - got 1, expected 3
Very likely caused by the change from bug 369306.
I will try to investigate that.
Assignee: nobody → mounir.lamouri
Hardware: x86_64 → All
Emu boots is popurlar increasingly as Ugg boots. They are genuine sheepskin boots. We have all the fashionable styles availably who can help you to got the cheap ugg boots!!
http://www.emu-boots.org/
Attached patch Patch v1 (obsolete) (deleted) — Splinter Review
Attempt to fix the orange.
Attachment #477409 - Flags: review?(jst)
Status: NEW → ASSIGNED
Attachment #477409 - Flags: review?(jst) → review+
I think Steven Michaud has a modification for this test posted, cc'ing him.
(In reply to comment #57)

> I think Steven Michaud has a modification for this test posted,
> cc'ing him.

Yes I do, at bug 601182 (attachment 484147 [details] [diff] [review]).

But I don't think it's terribly likely to fix this bug.  And in any
case it can't land until the patch for bug 601182 also lands.
Attached patch Patch v1.1 (deleted) — Splinter Review
Attachment #477409 - Attachment is obsolete: true
Fix pushed:
http://hg.mozilla.org/mozilla-central/rev/dab8e1613a67

I will wait a few days and if nothing appears, I will close it.
I was a bit to quick in my previous comment. The patch was pushed but some tests were still running without that patch. Now, no tests on mozilla-central should be running without that patch.
(In reply to comment #83)

Mounir:  Judging by the number of recent mozilla-central test failures, your patch doesn't seem to have fixed this bug.
(In reply to comment #99)
> (In reply to comment #83)
> 
> Mounir:  Judging by the number of recent mozilla-central test failures, your
> patch doesn't seem to have fixed this bug.

Indeed. Even if some results come from the tryserver and tracemonkey which were not up to date.
I will try to have a look today to see if I can find something else.
Attached patch Patch 2 (deleted) — Splinter Review
So, I've seen two issues here:
1. when focusing the main window back, I didn't use an event handler. This might not be a problem but the event handler is better for correctness.
2. the plugin also use an event handler to update its states so we need to be sure to not check them before the update. Using executeSoon() should fix that.

I hope this patch would be the one :)
Attachment #487607 - Flags: review?(jst)
Attachment #487607 - Flags: review?(jst) → review+
Pushed the new fix:
http://hg.mozilla.org/mozilla-central/rev/f0129b7f9190

Marking the bug as fixed. I will reopen if the fix isn't working.
Status: ASSIGNED → RESOLVED
Closed: 14 years ago
Resolution: --- → FIXED
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Argh :(

Could that be the reason of this random orange:
2010-11-01 14:50:09.232 firefox-bin[544:903] Mozilla has caught an Obj-C exception [NSInternalInconsistencyException: Error (1000) creating CGSWindow]
NPP_Destroy
Mounir, you might want to try what I did in a revision of this test
which I haven't yet landed (attachment 484147 [details] [diff] [review], see bug 601182 comment
#21):

I got rid of 'onload="runTests()"' from the <body> tag and replaced it
with:

  window.opener.SimpleTest.waitForFocus(runTests, window);

When Josh first suggested my patch might fix this problem (in comment
#57 above) I didn't think this was likely.  Now I'm not so sure.
Attached patch Patch #3 (obsolete) (deleted) — Splinter Review
Let's try with waitForFocus().
Attachment #490465 - Flags: review?(jst)
Comment on attachment 490465 [details] [diff] [review]
Patch #3

The failure seems permanent or very frequent with this patch. I will check on a Mac what could be the cause. Canceling the review in the mean time.
Attachment #490465 - Flags: review?(jst)
Attached patch Patch #3 (deleted) — Splinter Review
Attachment #490465 - Attachment is obsolete: true
Attachment #490544 - Flags: review?(jst)
Attachment #490544 - Flags: review?(jst) → review+
Pushed:
http://hg.mozilla.org/mozilla-central/rev/68df5bb6da6c

Some MacOS X mochitests are still running on try without this patch applied.
Let's hope this patch will fix the issue.
Status: REOPENED → RESOLVED
Closed: 14 years ago14 years ago
Resolution: --- → FIXED
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Ok, I think my last patch fixed this random orange.
After my patch, all failures are after "browser_relative.js" failing and this did not happen since a month even if bug 573259 (browser_relative intermittent) still happen. I think that, for some reason, bug 573259 no longer make this test failing.

Feel free to reopen if you do not agree.
Status: REOPENED → RESOLVED
Closed: 14 years ago14 years ago
Resolution: --- → FIXED
Looks like comment 211 is this bug. I reopen this bug. If you don't think the log isn't this bug, please re-close this bug.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Obviously, comment 211 was a false positive.
Status: REOPENED → RESOLVED
Closed: 14 years ago13 years ago
Resolution: --- → FIXED
Whiteboard: [orange]
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: