740769 - [Thunderbird, OSX 10.6] Intermittent "test_bug325418.js | test failed (with xpcshell return code: 0)" + "TEST-UNEXPECTED-FAIL | undefined | 2 == true"

Status: VERIFIED FIXED

(Core :: XPCOM, defect)

Description

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1332773443.1332778348.25322.gz
MacOSX 10.6 comm-central test xpcshell on 2012/03/26 07:50:43
s: tb2-darwin10-slave60

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1333064813.1333069752.32654.gz
MacOSX 10.6 comm-central test xpcshell on 2012/03/29 16:46:53
s: tb2-darwin10-slave61
{
TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | test failed (with xpcshell return code: 0), see following log:
>>>>>>>

TEST-INFO | (xpcshell/head.js) | test 1 pending

TEST-INFO | (xpcshell/head.js) | test 2 pending

TEST-INFO | (xpcshell/head.js) | test 2 finished

TEST-INFO | (xpcshell/head.js) | running event loop

TEST-UNEXPECTED-FAIL | undefined | 2 == true - See following stack:
JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/head.js :: do_throw :: line 462
JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/head.js :: _do_check_eq :: line 556
JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/head.js :: do_check_eq :: line 577
JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/head.js :: do_check_true :: line 591
JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js :: <TOP_LEVEL> :: line 14

TEST-INFO | (xpcshell/head.js) | exiting test
<<<<<<<
}

Comment 1

[Serge Gautherie (:sgautherie)]

Attachment: (Av1) Fix multiple issues in test_bug325418.js [Checked in: Comments 3 and 19]

Comment 2

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 610850 [details] [diff] [review]
(Av1) Fix multiple issues in test_bug325418.js
[Checked in: Comments 3 and 19]

r=me

Comment 3

[Serge Gautherie (:sgautherie)]

Comment on attachment 610850 [details] [diff] [review]
(Av1) Fix multiple issues in test_bug325418.js
[Checked in: Comments 3 and 19]

https://hg.mozilla.org/mozilla-central/rev/da0d07b5ca1e


[Approval Request Comment]
Regression caused by (bug #): Bug 325418.
User impact if declined:  None, but (random-orange on) broken(!) test.
Testing completed (on m-c, etc.): This comment.
Risk to taking this patch (and alternatives if risky): None, test-only.
String changes made by this patch: None.

Comment 4

[Alex Keybl [:akeybl]]

Comment on attachment 610850 [details] [diff] [review]
(Av1) Fix multiple issues in test_bug325418.js
[Checked in: Comments 3 and 19]

[Triage Comment]
Test-only fix. Approved for Aurora 13 and Beta 12.

Comment 5

[Serge Gautherie (:sgautherie)]

Attachment: (Bv1) Convert test_bug325418.js file to LF end-of-lines, Fix nsITimer.idl documentation [Checked in: Comment 12]

Comment 6

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 613464 [details] [diff] [review]
(Bv1) Convert test_bug325418.js file to LF end-of-lines, Fix nsITimer.idl documentation
[Checked in: Comment 12]

r=me

Comment 7

[Serge Gautherie (:sgautherie)]

Attachment: (Cv1) test_bug325418.js: Further improvements to reduce/diagnose random (too long delay) failures

Patch A did not fully fix this "orange":

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1333900948.1333905770.17458.gz
MacOSX 10.6 comm-central test xpcshell on 2012/04/08 09:02:28
s: tb2-darwin10-slave62
{
TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | test failed (with xpcshell return code: 0), see following log:

TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | 3 == 1 - See following stack:

JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js :: observeTC :: line 16
}

***

Here is a patch which should help further,
though I wonder whether it's a (random) box (slowness) issue :-|

Comment 8

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 613466 [details] [diff] [review]
(Cv1) test_bug325418.js: Further improvements to reduce/diagnose random (too long delay) failures

Please describe the actual changes made in your checkin comment.  That would have the side benefit of letting me know what you're trying to accomplish, exactly.

Comment 9

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1333939806.1333944991.7137.gz
MacOSX 10.6 comm-central test xpcshell on 2012/04/08 19:50:06
s: tb2-darwin10-slave62
TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | 3 == 1 - See following stack:


http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1334011952.1334017314.10437.gz
MacOSX 10.6 comm-central test xpcshell on 2012/04/09 15:52:32
s: tb2-darwin10-slave60
TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | 3 == 1 - See following stack:

Comment 10

[Serge Gautherie (:sgautherie)]

Comment on attachment 613466 [details] [diff] [review]
(Cv1) test_bug325418.js: Further improvements to reduce/diagnose random (too long delay) failures

*"Duplicate" observer, just in (unlikely) case the first timer would be firing.
*Init gStartTime[12] later, to reduce the recorded delay a bit.
*Use Date.now() instead of (new Date()).getTime(): simpler (& faster?).
*"timer = null;" a bit later, as not needed earlier, fwiw.
*Nits.

Comment 11

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 613466 [details] [diff] [review]
(Cv1) test_bug325418.js: Further improvements to reduce/diagnose random (too long delay) failures

That should all go in the checkin comment.

Comment 12

[Serge Gautherie (:sgautherie)]

Comment on attachment 613464 [details] [diff] [review]
(Bv1) Convert test_bug325418.js file to LF end-of-lines, Fix nsITimer.idl documentation
[Checked in: Comment 12]

http://hg.mozilla.org/mozilla-central/rev/6fe5b0271cd1

Comment 13

[Serge Gautherie (:sgautherie)]

Attachment: (Cv1a) test_bug325418.js: Create separate observers, Initialize gStartTime[12] later, Use Date.now() instead of (new Date()).getTime(), "timer = null;" a bit later, Nits [Checked in: Comment 17]

Cv1, with comment 8 suggestion(s).

Comment 14

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 613474 [details] [diff] [review]
(Cv1a) test_bug325418.js: Create separate observers, Initialize gStartTime[12] later, Use Date.now() instead of (new Date()).getTime(), "timer = null;" a bit later, Nits
[Checked in: Comment 17]

r=me

Comment 15

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

(In reply to Serge Gautherie (:sgautherie) from comment #12)
> Comment on attachment 613464 [details] [diff] [review]
> (Bv1) Convert test_bug325418.js file to LF end-of-lines, Fix nsITimer.idl
> documentation
> [Checked in: Comment 12]
> 
> http://hg.mozilla.org/mozilla-central/rev/6fe5b0271cd1

This was an inappropriate use of DONTBUILD.

Comment 16

[Serge Gautherie (:sgautherie)]

(In reply to Ms2ger from comment #15)
> This was an inappropriate use of DONTBUILD.

Why?

Comment 17

[Serge Gautherie (:sgautherie)]

Comment on attachment 613474 [details] [diff] [review]
(Cv1a) test_bug325418.js: Create separate observers, Initialize gStartTime[12] later, Use Date.now() instead of (new Date()).getTime(), "timer = null;" a bit later, Nits
[Checked in: Comment 17]

https://hg.mozilla.org/mozilla-central/rev/2374ebd5026f

Comment 18

[Serge Gautherie (:sgautherie)]

Though I could tweak code order a bit again to reduce actual delay a bit more,
this test looks fine as it is now wrt its code.

V.Fixed, wrt that.

*****

(In reply to Serge Gautherie (:sgautherie) from comment #7)
> I wonder whether it's a (random) box (slowness) issue :-|

Still happening (as I expected :-|):

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1334080146.1334085169.24313.gz
MacOSX 10.6 comm-central test xpcshell on 2012/04/10 10:49:06
s: tb2-darwin10-slave62
{
TEST-UNEXPECTED-FAIL | /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js | 2 == 1 - See following stack:

JS frame :: /builds/slave/comm-central-macosx64-opt-unittest-xpcshell/build/xpcshell/tests/xpcom/tests/unit/test_bug325418.js :: observeTC2 :: line 39
}

If someone wants to investigate the specific issue of these TB on OSX 10.6 builds,
please file a follow-up bug.

Comment 19

[Jens Hatlak (:InvisibleSmiley)]

Comment on attachment 610850 [details] [diff] [review]
(Av1) Fix multiple issues in test_bug325418.js
[Checked in: Comments 3 and 19]

http://hg.mozilla.org/releases/mozilla-aurora/rev/4c7f697b3697

Comment 20

[Serge Gautherie (:sgautherie)]

Missed Gecko 12 :-|