Bug 79748 - [Chromium] We should reset mouse position before calling eventSender.mouseMove in LayoutTest
Summary: [Chromium] We should reset mouse position before calling eventSender.mouseMov...
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: yosin
URL:
Keywords:
Depends on:
Blocks: 70298
  Show dependency tree / graph
 
Reported: 2012-02-27 23:16 PST by yosin
Modified: 2012-03-04 19:33 PST (History)
2 users (show)

See Also:

Attachments
Patch 1 (4.60 KB, patch)
2012-02-27 23:38 PST, yosin 		no flags Details | Formatted Diff | Diff
Patch 2 (3.18 KB, patch)
2012-02-28 00:38 PST, yosin 		no flags Details | Formatted Diff | Diff
Show Obsolete (1) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description yosin 2012-02-27 23:16:33 PST 
It seems flaky layout tests in fast/forms/number sometimes didn't receive mouse move event. This may be the root cause of flaky.
I update following tests adding reset mouse position before sending mouse move event.
 * input-number-events.html
 * input-number-large-padding.html
 * input-spinbutton-capturing.html
I'm not sure this update remove flake. But, I believe it is worth to try and make out test robust.
If this update won't fix flaky, I'll seek another for root of flaky.
Comment 1 yosin 2012-02-27 23:38:25 PST 
Created attachment 129198 [details]
Patch 1
Comment 2 Kent Tamura 2012-02-27 23:56:54 PST 
Comment on attachment 129198 [details]
Patch 1

View in context: https://bugs.webkit.org/attachment.cgi?id=129198&action=review

> LayoutTests/platform/chromium/test_expectations.txt:2154
> -BUGCR54330 MAC : fast/forms/number/input-spinbutton-capturing.html = PASS TEXT
> -BUGCR54331 MAC : fast/forms/number/input-number-events.html = PASS TEXT
>  

We're not sure that mouse pointer position is the root cause of the flakiness yet.  So do not remove these lines.

> LayoutTests/platform/chromium/test_expectations.txt:-3340
> -BUGWK70298 WIN LINUX : fast/forms/number/input-number-events.html = TEXT PASS
> -BUGWK70298 : fast/forms/number/input-number-large-padding.html = TEXT PASS
> -BUGWK70298 WIN LINUX : fast/forms/number/input-spinbutton-capturing.html = TEXT PASS

ditto.
Comment 3 yosin 2012-02-28 00:38:55 PST 
Created attachment 129208 [details]
Patch 2
Comment 4 Kent Tamura 2012-02-28 01:03:55 PST 
Comment on attachment 129208 [details]
Patch 2

ok.
Let's land this and see the flakiness dashboard.
Comment 5 WebKit Review Bot 2012-02-28 02:19:19 PST 
The commit-queue encountered the following flaky tests while processing attachment 129208 [details]:

css3/filters/effect-invert-hw.html bug 79639 (author: cmarrin@apple.com)
fast/files/file-reader-event-listener.html bug 79769 (author: jianli@chromium.org)
The commit-queue is continuing to process your patch.
Comment 6 WebKit Review Bot 2012-02-28 02:21:21 PST 
Comment on attachment 129208 [details]
Patch 2

Clearing flags on attachment: 129208

Committed r109093: <http://trac.webkit.org/changeset/109093>
Comment 7 WebKit Review Bot 2012-02-28 02:21:25 PST 
All reviewed patches have been landed.  Closing bug.
Comment 8 Kent Tamura 2012-03-04 18:26:17 PST 
Ok, they look really stable after r109093.  I'll remove the flakiness expectation from test_expectations.txt.
Comment 9 Kent Tamura 2012-03-04 19:33:48 PST 
  (In reply to comment #8)
> Ok, they look really stable after r109093.  I'll remove the flakiness expectation from test_expectations.txt.

Done: http://trac.webkit.org/changeset/109692