Closed Bug 719186 Opened 13 years ago Closed 3 years ago

Intermittent docshell/test/test_bug413310.html | finished in a non-clean fashion (in /tests/docshell/test/test_bug384014.html)

Categories

(Core :: DOM: Navigation, defect, P5)

Product:
Core
Component:
DOM: Navigation
Type:
defect

Tracking

()

Status:
RESOLVED INCOMPLETE
Tracking Flags:
Tracking Status
firefox12 --- wontfix

People

(Reporter: sgautherie, Unassigned)

References

(
URL
)

Details

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

Attachments

(2 files)

guess fix
(deleted), patch
bzbarsky
: review+
Details | Diff | Splinter Review
export
(deleted), patch
Details | Diff | Splinter Review
No description provided.
Component: General → Document Navigation
QA Contact: general → docshell
Is there a log for this failure?
Oh, forget that! https://tbpl.mozilla.org/php/getParsedLog.php?id=8639231&tree=Firefox Rev4 MacOSX Snow Leopard 10.6 mozilla-central debug test mochitests-2/5 on 2012-01-18 09:23:09 PST for push 19af1aac33ac { 573 INFO TEST-UNEXPECTED-FAIL | /tests/docshell/test/test_bug413310.html | finished in a non-clean fashion (in /tests/docshell/test/test_bug384014.html) }
(In reply to Serge Gautherie (:sgautherie) from comment #2) > Oh, forget that! "Oh, I forgot to copy that!"
So this test calls history.back(), and that seems to navigate the test frame back as opposed to the iframe inside the test. Boris, I remember we had a bug similar to this a while ago which you fixed. Could you also look into this?
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey-Aurora/1335343933.1335346096.5960.gz WINNT 5.2 comm-aurora debug test mochitests-2/5 on 2012/04/25 01:52:13 s: cn-sea-qm-win2k3-01
Whiteboard: [orange]
Failures (now) happen on all platforms, not x86_64 MacOSX only. ***** Recent reports often include some JS 'TypeError'. Hopefully, they could help to figure out what causes this bug. (In reply to TBPL Robot from comment #105) > TypeError: infos is undefined (In reply to TBPL Robot from comment #112) > TypeError: this.downloads[data.manifest] is undefined (In reply to TBPL Robot from comment #113) > TypeError: this.downloads[data.manifest] is undefined (In reply to TBPL Robot from comment #114) > TypeError: this.downloads[data.manifest] is undefined > TypeError: infos is null (In reply to TBPL Robot from comment #115) > TypeError: this.downloads[data.manifest] is undefined (In reply to TBPL Robot from comment #116) > TypeError: this.downloads[data.manifest] is undefined (In reply to TBPL Robot from comment #118) > ImportError: cannot import name OrderedDict (In reply to TBPL Robot from comment #120) > ImportError: cannot import name OrderedDict
OS: Mac OS X → All
Hardware: x86_64 → All
Boris, about comment 4 and comment 5.
Flags: needinfo?(bzbarsky)
Olli, do you have time to look into this?
Flags: needinfo?(bzbarsky) → needinfo?(bugs)
Test disabled until someone looks into the failures: https://hg.mozilla.org/integration/mozilla-inbound/rev/8ab4f1b026aa
Whiteboard: [test disabled][leave open]
Attached patch guess fix (deleted) — — Splinter Review
Doing any new page load or session history stuff during load event listener may lead to surprising results. https://tbpl.mozilla.org/?tree=Try&rev=489170531aa4
Flags: needinfo?(bugs)
(In reply to Olli Pettay [:smaug] from comment #168) > Doing any new page load or session history stuff during load event listener > may lead to surprising results. Worth a try :-) May be try SimpleTest.executeSoon() ? Maybe add a comment to the test itself to explain why a "delay" is needed?
No reason to use SimpleTest.executeSoon. Apparently the test is now in m-1, so I retriggered it few times.
Attachment #8453428 - Flags: review?(bzbarsky)
Comment on attachment 8453428 [details] [diff] [review] guess fix Seems ok, though maybe executSoon instead of setTimeout?
Attachment #8453428 - Flags: review?(bzbarsky) → review+
I really don't know why to use executeSoon.
(in other words, I prefer using APIs which are exposed to normal web pages.)
executeSoon just makes it clear that you don't care about timers, just want a separate task.
Attached patch export (deleted) — — Splinter Review
Well, since I pushed to try with setTimeout, not worth to change.
Assignee: nobody → bugs
sorry had to backout this change for e10s test failure like https://tbpl.mozilla.org/php/getParsedLog.php?id=43590098&tree=Mozilla-Inbound
Assignee: bugs → nobody
(In reply to Carsten Book [:Tomcat] from comment #177) > sorry had to backout this change for e10s test failure like > https://tbpl.mozilla.org/php/getParsedLog.php?id=43590098&tree=Mozilla- > Inbound Ftr, {{ Ubuntu VM 12.04 mozilla-inbound opt test mochitest-e10s-2 on 2014-07-11 00:54:09 PDT for push 2b7493acc268 slave: tst-linux32-spot-652 00:57:24 INFO - 129 INFO TEST-PASS | /tests/docshell/test/test_bug413310.html | Done new load 00:57:24 INFO - 130 INFO TEST-PASS | /tests/docshell/test/test_bug413310.html | Unexpected subframe location after new load 00:57:24 INFO - 131 INFO TEST-PASS | /tests/docshell/test/test_bug413310.html | History traversal didn't fire onload: bfcache issues! 00:57:24 INFO - 132 INFO TEST-PASS | /tests/docshell/test/test_bug413310.html | Unexpected subframe location 00:57:24 INFO - 133 INFO TEST-UNEXPECTED-FAIL | /tests/docshell/test/test_bug413310.html | Did a load when going back? - got POST 1405065443969, expected POST 1405065444146 }}
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3
Priority: P3 → P5

Bulk closing some old intermittents.

Status: NEW → RESOLVED
Closed: 3 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: