Closed Bug 1174951 Opened 10 years ago Closed 6 years ago

Intermittent test_XHRDocURI.html | pushStateTest - finished in a non-clean fashion, probably because it didn't call SimpleTest.finish() | test_XHRDocURI.html | undefined assertion nam

Categories

(Core :: DOM: Navigation, defect, P3)

Product:
Core
Component:
DOM: Navigation
Version:
41 Branch
Type:
defect

Tracking

()

Status:
RESOLVED INACTIVE
Tracking Flags:
Tracking Status
firefox39 --- unaffected
firefox40 --- affected
firefox41 --- affected
firefox42 --- affected
firefox-esr31 --- unaffected
firefox-esr38 --- unaffected

People

(Reporter: KWierso, Assigned: smaug)

References

Details

(Keywords: intermittent-failure)

Attachments

(3 files, 1 obsolete file)

with executeSoon
1.46 KB, patch
Details | Diff | Splinter Review
guess fix, v2
640 bytes, patch
Details | Diff | Splinter Review
guess fix, v3
787 bytes, patch
bzbarsky
: review+
Details | Diff | Splinter Review
14:41:03 INFO - 400 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | wrong attr base (doc base and xml:base are same) 14:41:03 INFO - 401 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | wrong attr base (.baseURIObject, doc base and xml:base are same) 14:41:03 INFO - 402 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | wrong base (after <base> changed) 14:41:03 INFO - 403 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | wrong base (.baseURIObject, after <base> changed) 14:41:03 INFO - 404 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | undefined assertion name 14:41:03 INFO - 405 INFO TEST-PASS | dom/base/test/test_XHRDocURI.html | undefined assertion name 14:41:03 INFO - 406 INFO TEST-UNEXPECTED-FAIL | dom/base/test/test_XHRDocURI.html | /tests/dom/base/test/test_XHRDocURI.html/pushStateTest - finished in a non-clean fashion, probably because it didn't call SimpleTest.finish() 14:41:03 INFO - 407 INFO TEST-OK | dom/base/test/test_XHRDocURI.html | took 1007ms 14:41:03 INFO - 408 INFO TEST-UNEXPECTED-FAIL | dom/base/test/test_XHRDocURI.html | undefined assertion name - Result logged after SimpleTest.finish() 14:41:03 INFO - ++DOMWINDOW == 23 (000000D03C13AC00) [pid = 2776] [serial = 2040] [outer = 000000D04C1E9000] 14:41:03 INFO - 409 INFO TEST-START | dom/base/test/test_XHRResponseURL.html 14:41:03 INFO - ++DOMWINDOW == 24 (000000D03C13C800) [pid = 2776] [serial = 2041] [outer = 000000D04C1E9000] 14:41:03 INFO - 410 INFO TEST-PASS | dom/base/test/test_XHRResponseURL.html | XMLHttpRequest should have responseURL attribute 14:41:03 INFO - 411 INFO TEST-PASS | dom/base/test/test_XHRResponseURL.html | responseURL should be empty string if response's url is null 14:41:03 INFO - 412 INFO TEST-PASS | dom/base/test/test_XHRResponseURL.html | assert request succeeded 14:41:03 INFO - 413 INFO TEST-PASS | dom/base/test/test_XHRResponseURL.html | request to same-origin without redirect 14:41:03 INFO - 414 INFO TEST-PASS | dom/base/test/test_XHRResponseURL.html | assert request succeeded
It looks like this test changed recently, in bug 1173215.
Flags: needinfo?(bugs)
Blocks: 1173215
Assignee: nobody → bugs
Flags: needinfo?(bugs)
Attachment #8625738 - Flags: review?(bzbarsky)
Comment on attachment 8625738 [details] [diff] [review] possible fix Probably better to use executeSoon instead of setTimeout there. r=me with that.
Attachment #8625738 - Flags: review?(bzbarsky) → review+
(Don't know why. I explicitly try to avoid using executeSoon in tests, because tests should use web APIs when possible.) Fine, I'll change to executeSoon
Attached patch with executeSoon — — Splinter Review
Attachment #8625738 - Attachment is obsolete: true
and the fix is somewhat educated guess. I haven't been able to reproduce the bug.
https://hg.mozilla.org/mozilla-central/rev/60fe089face1
Status: NEW → RESOLVED
Closed: 10 years ago
status-firefox41: affected → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla41
Status: RESOLVED → REOPENED
status-firefox40: fixed → affected
status-firefox41: fixed → affected
status-firefox42: --- → affected
Flags: needinfo?(bugs)
Resolution: FIXED → ---
Target Milestone: mozilla41 → ---
Oh, fun, that "fix" actually seems to reveal the bug.
But still not reproduceable locally :(
when running the test with other tests, things seem to go wrong more likely.
Comment on attachment 8627379 [details] [diff] [review] guess fix, v3 I'm not really sure what is happening here. Can't reproduce locally, but something around that history.back() causes failures when the test runs within an iframe. If this doesn't work, I'll make the test to open a separate window.
Attachment #8627379 - Flags: review?(bzbarsky)
Comment on attachment 8627379 [details] [diff] [review] guess fix, v3 r=flailing
Attachment #8627379 - Flags: review?(bzbarsky) → review+
(In reply to Treeherder Robot from comment #50) This was on a newer push than comment 48.
Keywords: leave-open
Any other ideas here before I disable the test?
Flags: needinfo?(bugs)
I'd vote for just disabling the test at this point, FWIW.
Flags: needinfo?(bugs)
This test uses SpecialPowers.addPermission, which is racy. That may be causing this failures. Instead, it should use pushPermissions.
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3
The leave-open keyword is there and there is no activity for 6 months. :smaug, maybe it's time to close this bug?
Flags: needinfo?(bugs)
Please reopen if this issue happens still.
Status: REOPENED → RESOLVED
Closed: 10 years ago → 6 years ago
Flags: needinfo?(bugs)
Resolution: --- → INACTIVE
Keywords: leave-open
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: