Web Inspector: inspector/storage-panel-dom-storage-update.html is flakey on the bots

[Joseph Pecoraro]

inspector/storage-panel-dom-storage-update.html is flakey on the bots.

It looks like the test is relying on InspectorTest.runAfterPendingDispatches(…) to run code after DOM Storage add/update/clears. However, those are not pending dispatches / results of actions, but DOMStorage agent events. So we should instead listen for pending events.

[Radar WebKit Bug Importer]

<rdar://problem/14881605>

[Joseph Pecoraro]

Created attachment 210147 [details]
[PATCH] Proposed Fix

The test could probably be simplified even more (removing most of the runAfterPendingDispatches could probably be removed) but ideally this would be converted to a real protocol test. I think this should fix the flakey ness though.

[WebKit Commit Bot]

Comment on attachment 210147 [details]
[PATCH] Proposed Fix

Clearing flags on attachment: 210147

Committed r154907: <http://trac.webkit.org/changeset/154907>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Joseph Pecoraro]

Ahh, WebKit2 bot still complains about this:
<http://build.webkit.org/results/Apple%20MountainLion%20Release%20WK2%20(Tests)/r154907%20(11253)/results.html>

Investigating now that I can reproduce with -2.

[Joseph Pecoraro]

So adding an extra log makes the test not timeout right now which makes very little sense at the moment.

            function viewUpdatedAfterAddition()
            {
                InspectorTest.addResult("viewUpdatedAfterAddition");
                InspectorTest.runAfterPendingDispatches(function() {

[Joseph Pecoraro]

(In reply to comment #6)
> So adding an extra log makes the test not timeout right now which makes very little sense at the moment.
> 
>             function viewUpdatedAfterAddition()
>             {
>                 InspectorTest.addResult("viewUpdatedAfterAddition");
>                 InspectorTest.runAfterPendingDispatches(function() {

So apparently removing this runAfterPendingDispatches fixed things, as long as I then made running the event handlers aware of this. I have no idea why this caused any issue on WebKit 2. I have little interest in investigating that further right now when it was unnecessary anyways.

[Joseph Pecoraro]

Landed follow-up fix: <http://trac.webkit.org/changeset/154911>

[Alexey Proskuryakov]

This change made it so that the test times out 100% of the time on WebKit2. The test also sometimes times out on WK1. Marked accordingly in <http://trac.webkit.org/r156002>.

This doesn't happen when new Web Inspector is enabled (I don't know if the test runs properly then, but it's reported as passing).

[Timothy Hatcher]

Moving to the right component.

[Brian Burg]

Closing as invalid, as this bug pertains to the old inspector UI and/or its tests.
Please file a new bug (https://www.webkit.org/new-inspector-bug) if the bug/feature/issue is still relevant to WebKit trunk.