Bug 101695 - Uninitialized fields in class JSLock
Summary: Uninitialized fields in class JSLock
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: All All
: P2 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-11-08 20:17 PST by Cosmin Truta
Modified: 2012-11-13 09:17 PST (History)
6 users (show)

See Also:

Attachments
Patch (1001 bytes, patch)
2012-11-08 20:22 PST, Cosmin Truta 		no flags Details | Formatted Diff | Diff
Patch (1.10 KB, patch)
2012-11-10 21:12 PST, Cosmin Truta 		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 Cosmin Truta 2012-11-08 20:17:19 PST 
Valgrind reported that JSLock::m_ownerThread is not initialized. The report indicates a legitimate issue.
A fix will follow.
Comment 1 Cosmin Truta 2012-11-08 20:22:40 PST 
Created attachment 173186 [details]
Patch
Comment 2 Mark Hahnenberg 2012-11-09 14:39:35 PST 
Comment on attachment 173186 [details]
Patch

I think we should also initialize m_lockDropDepth, no?
Comment 3 Cosmin Truta 2012-11-10 21:12:33 PST 
Created attachment 173484 [details]
Patch

(In reply to comment #2)
> I think we should also initialize m_lockDropDepth, no?

You're right. Done.
Comment 4 WebKit Review Bot 2012-11-10 22:04:43 PST 
Comment on attachment 173484 [details]
Patch

Attachment 173484 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/14786720

New failing tests:
inspector-protocol/debugger-terminate-dedicated-worker-while-paused.html
Comment 5 Cosmin Truta 2012-11-11 14:54:20 PST 
Comment on attachment 173484 [details]
Patch

Setting back "cq?". Something seems to be wrong with the Chrome-Linux build.
Comment 6 Mark Hahnenberg 2012-11-13 08:31:20 PST 
(In reply to comment #5)
> (From update of attachment 173484 [details])
> Setting back "cq?". Something seems to be wrong with the Chrome-Linux build.

I think that's just a flaky test. Chrome doesn't use JSC and therefore doesn't depend on JSLock in any way.
Comment 7 Mark Hahnenberg 2012-11-13 08:33:19 PST 
Comment on attachment 173484 [details]
Patch

r=me
Comment 8 WebKit Review Bot 2012-11-13 09:17:41 PST 
Comment on attachment 173484 [details]
Patch

Clearing flags on attachment: 173484

Committed r134430: <http://trac.webkit.org/changeset/134430>
Comment 9 WebKit Review Bot 2012-11-13 09:17:44 PST 
All reviewed patches have been landed.  Closing bug.