Bug 61545 - [chromium] Be robust to empty viewports in the compositor
Summary: [chromium] Be robust to empty viewports in the compositor
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: WebCore Misc. (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Adrienne Walker
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-05-26 11:01 PDT by Adrienne Walker
Modified: 2011-05-26 12:06 PDT (History)
3 users (show)

See Also:

Attachments
Patch (4.05 KB, patch)
2011-05-26 11:13 PDT, Adrienne Walker 		jamesr: review+
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 Adrienne Walker 2011-05-26 11:01:19 PDT 
[chromium] Be robust to empty viewports in the compositor
Comment 1 Adrienne Walker 2011-05-26 11:13:10 PDT 
Created attachment 95008 [details]
Patch
Comment 2 Adrienne Walker 2011-05-26 11:17:03 PDT 
See: http://code.google.com/p/chromium/issues/detail?id=83961

The order of events that causes this is that the layer renderer is created and a doComposite event is called while the layer renderer (and the main frame) have an empty viewport rect.

This might be able to be fixed at a higher level too, but I think this code should be robust to the possibility of an empty content rect.
Comment 3 James Robinson 2011-05-26 11:26:38 PDT 
Comment on attachment 95008 [details]
Patch

R=me. the m_updateRect = IntRect() before the early out confused me for a bit, maybe stick the clear inside the branch?
Comment 4 Adrienne Walker 2011-05-26 11:29:07 PDT 
(In reply to comment #3)
> (From update of attachment 95008 [details])
> R=me. the m_updateRect = IntRect() before the early out confused me for a bit, maybe stick the clear inside the branch?

Thanks.  I'll clean that up before landing.
Comment 5 Adrienne Walker 2011-05-26 12:06:58 PDT 
Committed r87409: <http://trac.webkit.org/changeset/87409>