Bug 22730

Summary:

Adding test to verify that an active timer does not cause a mem leak by keeping protected JS objects

Product:

WebKit

Reporter:

Dmitry Titov <dimich>

Component:

Tools / Tests

Assignee:

Nobody <webkit-unassigned>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

ap, darin

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

All

Attachments:

Description	Flags
proposed patch	darin: review+
Same test, fixed the typo	none

Dmitry Titov

Reported 2008-12-07 18:05:20 PST

This is a test for bug 22710.

Attachments
proposed patch (3.50 KB, patch) 2008-12-07 18:07 PST, Dmitry Titov	darin: review+	Details Formatted Diff Diff
Same test, fixed the typo (3.50 KB, patch) 2008-12-08 11:43 PST, Dmitry Titov	no flags	Details Formatted Diff Diff
View All Add attachment proposed patch, testcase, etc.

Dmitry Titov

Comment 1 2008-12-07 18:07:19 PST

Created attachment 25833 [details] proposed patch

Alexey Proskuryakov

Comment 2 2008-12-07 23:18:44 PST

Tests that use GCController.getJSObjectCount() break when running "run-webkit-tests --threaded", so I'm not sure if they are a gain or a loss. Perhaps we should remove the --threaded option altogether, now that multithreading is tested in fast/workers?

Alexey Proskuryakov

Comment 3 2008-12-07 23:46:33 PST

Now that I think about it more, GCController.getJSObjectCount() will work fine in all tests when I land my patch that makes JSCreateGlobalObject use a unique context group.

Darin Adler

Comment 4 2008-12-08 10:24:14 PST

Comment on attachment 25833 [details] proposed patch Test looks good, r=me There is a typo in a comment: "shoudl".

Dmitry Titov

Comment 5 2008-12-08 11:43:36 PST

Created attachment 25847 [details] Same test, fixed the typo Fixed the typo.

Alexey Proskuryakov

Comment 6 2008-12-08 23:49:25 PST

Committed revision 39131.

Note You need to log in before you can comment on or make changes to this bug.