Bug 111514 - Layout Test plugins/plugin-clip-subframe.html is failing
Summary: Layout Test plugins/plugin-clip-subframe.html is failing
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: John Bauman
URL:
Keywords:
Depends on: 109879
Blocks:
  Show dependency treegraph
 
Reported: 2013-03-05 19:51 PST by Hayato Ito
Modified: 2013-03-14 17:41 PDT (History)
8 users (show)

See Also:


Attachments
Patch (3.29 KB, patch)
2013-03-08 00:53 PST, John Bauman
no flags Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Hayato Ito 2013-03-05 19:51:11 PST
The following layout test is failing on WebKit Mac and Win.

plugins/plugin-clip-subframe.html

Dashboard:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=plugins%2Fplugin-clip-subframe.html

Probable cause:
I cannot tell a exact revision from the dashboard.

The test was introduced 6 days ago in:
   http://trac.webkit.org/changeset/144236  Plugin in iframe may not display
Comment 1 Hayato Ito 2013-03-05 20:01:18 PST
Updated a chromium TestExpectations:

http://trac.webkit.org/changeset/144869
webkit.org/b/111514 [ Mac ] plugins/plugin-clip-subframe.html [ Failure ]
Comment 2 John Bauman 2013-03-08 00:53:28 PST
Created attachment 192170 [details]
Patch
Comment 3 Eric Seidel (no email) 2013-03-11 16:41:27 PDT
We should not need to force layouts from JavaScript.  If that does de-flake this, then we're missing a layout somewhere!
Comment 4 John Bauman 2013-03-11 16:48:20 PDT
Ok, I'm not certain this deflakes it on Win7 (dbg)(2) (as I can't reproduce the flakiness myself), but it seemed likely that this would fix the issue. I figured that it's probably doing layout at paint time, and since I'm using setTimeout in the test that means it's not doing a second paint/layout by the time the test finishes. I don't know if there's a better explanation.
Comment 5 Ryosuke Niwa 2013-03-14 01:06:48 PDT
*** Bug 112324 has been marked as a duplicate of this bug. ***
Comment 6 Ryosuke Niwa 2013-03-14 01:08:10 PDT
I've confirmed that this test change fixes the bug 112324.

If this signifies a real bug in WebCore, please file a separate bug and fix it there. It's not acceptable to have a flaky test like this one.
Comment 7 Ryosuke Niwa 2013-03-14 01:10:16 PDT
In fact, the bug 112324 might have been the culprit of Mac WK2 EWS being so behind.
Comment 8 John Bauman 2013-03-14 01:17:36 PDT
Actually, I'm not certain that this patch fixes the type of flakiness found in bug 1123234. Are you sure you weren't just lucky when testing it? I'll make a patch that's certain to fix that flakiness, just in case this one doesn't help.
Comment 9 Ryosuke Niwa 2013-03-14 01:21:49 PDT
(In reply to comment #8)
> Actually, I'm not certain that this patch fixes the type of flakiness found in bug 1123234. Are you sure you weren't just lucky when testing it? I'll make a patch that's certain to fix that flakiness, just in case this one doesn't help.

Yeah. I was seeing the 100% sample failure rate on my MBP with Mac WK2. With this change, its sample failure rate is 0%.
Comment 10 WebKit Review Bot 2013-03-14 01:41:10 PDT
Comment on attachment 192170 [details]
Patch

Clearing flags on attachment: 192170

Committed r145794: <http://trac.webkit.org/changeset/145794>
Comment 11 WebKit Review Bot 2013-03-14 01:41:14 PDT
All reviewed patches have been landed.  Closing bug.
Comment 12 Ryosuke Niwa 2013-03-14 17:41:03 PDT
You're right. plugins/plugin-clip-subframe.html is still flaky on some bot :(