Bug 155154 - WeakBlock::visit() should check for a WeakHandleOwner before consulting mark bits.
Summary: WeakBlock::visit() should check for a WeakHandleOwner before consulting mark ...
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Andreas Kling
URL:
Keywords: Performance
Depends on:
Blocks:
 
Reported: 2016-03-07 19:13 PST by Andreas Kling
Modified: 2016-03-08 09:57 PST (History)
5 users (show)

See Also:

Attachments
Patch (1.98 KB, patch)
2016-03-07 19:13 PST, Andreas Kling 		no flags Details | Formatted Diff | Diff
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 Andreas Kling 2016-03-07 19:13:08 PST 
Many WeakImpl's do not have a WeakHandleOwner, so WeakBlock::visit() is just wasting its time looking up the mark bits for its pointee.
Comment 1 Andreas Kling 2016-03-07 19:13:37 PST 
Created attachment 273260 [details]
Patch
Comment 2 Darin Adler 2016-03-08 09:09:40 PST 
Comment on attachment 273260 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=273260&action=review

> Source/JavaScriptCore/heap/WeakBlock.cpp:121
> +        const JSValue& jsValue = weakImpl->jsValue();

The use of const JSValue& here seems peculiar. Why not just JSValue?
Comment 3 WebKit Commit Bot 2016-03-08 09:56:57 PST 
Comment on attachment 273260 [details]
Patch

Clearing flags on attachment: 273260

Committed r197774: <http://trac.webkit.org/changeset/197774>
Comment 4 WebKit Commit Bot 2016-03-08 09:57:00 PST 
All reviewed patches have been landed.  Closing bug.