Closed Bug 815646 Opened 12 years ago Closed 11 years ago

Intermittent accessible/states/test_link.html | Doubled event { event type: document load complete, target: 'undefined' } in test with ID = 'link + 'link_traversed' clicked.'.

Categories

(Core :: Disability Access APIs, defect)

All
Linux
defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla20
Tracking Status
firefox19 --- fixed
firefox20 --- fixed

People

(Reporter: emorley, Assigned: surkov)

References

Details

(Keywords: intermittent-failure)

Attachments

(2 files)

+++ This bug was initially created as a clone of Bug #810554 +++
Similar to bug 810554 except without the initial timeout.

Rev3 Fedora 12 mozilla-inbound opt test mochitest-other on 2012-11-26 10:13:13 PST for push 7e97c5392d81

slave: talos-r3-fed-005

https://tbpl.mozilla.org/php/getParsedLog.php?id=17345817&tree=Mozilla-Inbound

{
9332 ERROR TEST-UNEXPECTED-FAIL | chrome://mochitests/content/a11y/accessible/states/test_link.html | Doubled event { event type: document load complete, target:  'undefined' } in test with ID = 'link + 'link_traversed' clicked.'.
}

With more context:
{
9313 INFO TEST-START | chrome://mochitests/content/a11y/accessible/states/test_link.html
9314 INFO TEST-INFO | chrome://mochitests/content/a11y/accessible/states/test_link.html | must wait for load
9315 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_href' !
9316 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for ['null node', address: [object Text], role: text leaf, name: 'link', address: [xpconnect wrapped nsIAccessible]]!
9317 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_click' !
9318 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_mousedown' !
9319 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_mouseup' !
9320 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_arialink' !
9321 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_ariabutton' !
9322 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | state bits should not be present in ID  'link_ariabutton' !
9323 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_notlink' !
9324 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | state bits should not be present in ID  'link_notlink' !
9325 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for  'link_traversed' !
9326 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | state bits should not be present in ID  'link_traversed' !
9327 INFO TEST-INFO | chrome://mochitests/content/a11y/accessible/states/test_link.html | Invoke the 'link + 'link_traversed' clicked.' test { expected 'state change' event; expected 'document load complete' event;  }
9328 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | Wrong extra state bit of the statechange event.
9329 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | Wrong state of the statechange event.
9330 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | Wrong state of statechange event state
9331 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | wrong state bits for ['a@id="link_traversed" node', address: http://www.example.com/, role: link, name: 'example.com', address: [xpconnect wrapped nsIAccessible]]!
9332 ERROR TEST-UNEXPECTED-FAIL | chrome://mochitests/content/a11y/accessible/states/test_link.html | Doubled event { event type: document load complete, target:  'undefined' } in test with ID = 'link + 'link_traversed' clicked.'.
9333 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | test with ID = 'link + 'link_traversed' clicked.' failed. No state change event.
9334 INFO TEST-PASS | chrome://mochitests/content/a11y/accessible/states/test_link.html | test with ID = 'link + 'link_traversed' clicked.' failed. No document load complete event.
9335 INFO TEST-END | chrome://mochitests/content/a11y/accessible/states/test_link.html | finished in 811ms
}
Attached patch enable loggingSplinter Review
it seems that was a recent change that documents having no presshell get document loading notificaitons (or something was changed in UI).
Attachment #686479 - Flags: review?(trev.saunders)
Attachment #686479 - Flags: review?(trev.saunders) → review+
enable logging (phase 2) http://hg.mozilla.org/integration/mozilla-inbound/rev/5061278bcf35
Whiteboard: [leave open]
about:blank document gets created before example.com document (a document open in new window on link click). From screen reader user point of view that could mean extra noise in temporary documents announcement. As far as I know we don't have any idea whether about:blank document is temporary or persistent one to filter the case.

Jamie, is flashing about:blank documents loadings (document complete load event + busy state change event) a problem for AT?
Flags: needinfo?(jamie)
If it ever gets focus, it will create extraneous "noise". If it doesn't get focus, it shouldn't.
Flags: needinfo?(jamie)
Attached patch fix the testSplinter Review
It doesn't seem (per Jamie's comment) we have a real problem here so let's fix the test.
Assignee: nobody → surkov.alexander
Status: NEW → ASSIGNED
Attachment #687666 - Flags: review?(trev.saunders)
Comment on attachment 687666 [details] [diff] [review]
fix the test

any idea why this isn't an issue in other tests?
(In reply to Trevor Saunders (:tbsaunde) from comment #15)
> Comment on attachment 687666 [details] [diff] [review]
> fix the test
> 
> any idea why this isn't an issue in other tests?

You mean tests that listen document_load_complete events? Maybe we should wait longer.
ping?
(In reply to alexander :surkov from comment #16)
> (In reply to Trevor Saunders (:tbsaunde) from comment #15)
> > Comment on attachment 687666 [details] [diff] [review]
> > fix the test
> > 
> > any idea why this isn't an issue in other tests?
> 
> You mean tests that listen document_load_complete events? Maybe we should

yes

> wait longer.

not sure what you mean by that.
(In reply to Trevor Saunders (:tbsaunde) from comment #19)

> > wait longer.
> 
> not sure what you mean by that.

I meant this failure doesn't happen very often so maybe we will see other tests failing one day.

I don't really see how they are different so that they shouldn't fail.
(In reply to alexander :surkov from comment #20)
> (In reply to Trevor Saunders (:tbsaunde) from comment #19)
> 
> > > wait longer.
> > 
> > not sure what you mean by that.
> 
> I meant this failure doesn't happen very often so maybe we will see other
> tests failing one day.
> 
> I don't really see how they are different so that they shouldn't fail.

ok, not sure I like it but I guess we can live with it.
Attachment #687666 - Flags: review?(trev.saunders) → review+
(In reply to Trevor Saunders (:tbsaunde) from comment #22)
> (In reply to alexander :surkov from comment #20)
> > (In reply to Trevor Saunders (:tbsaunde) from comment #19)
> > 
> > > > wait longer.
> > > 
> > > not sure what you mean by that.
> > 
> > I meant this failure doesn't happen very often so maybe we will see other
> > tests failing one day.
> > 
> > I don't really see how they are different so that they shouldn't fail.
> 
> ok, not sure I like it but I guess we can live with it.

at least you should try :) I really don't see alternatives, due to specialties of document loadings we can get a temporary document created, that's what log shows us, per Jamie's comment it shouldn't affect on them so there's no practical reason to fix this behavior and thus we need to fix a failing test. Also there's no practical reason to fix other similar tests because they may fail since we don't have known failures. Fixing a test is not ideal solution but it seems a practical one.
https://hg.mozilla.org/mozilla-central/rev/f586bd6769d3
Status: ASSIGNED → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla20
You need to log in before you can comment on or make changes to this bug.