Bug 66432

Summary: [chromium] LayerChromium::updateCompositorResources should use an explicit GC3D
Product: WebKit Reporter: Adrienne Walker <enne>
Component: WebCore Misc.Assignee: Adrienne Walker <enne>
Status: RESOLVED FIXED    
Severity: Normal CC: enne, husky, jamesr, nduca
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Bug Depends on:    
Bug Blocks: 66430    
Attachments:
Description Flags
Patch jamesr: review+

Description Adrienne Walker 2011-08-17 16:48:26 PDT
LayerChromium::updateCompositorResources is a commit-time function.  Therefore, it should use a GraphicsContext3D that is passed explicitly in as a function arg, rather than one that it implicitly has access to via LayerRendererChromium.
Comment 1 Adrienne Walker 2011-08-19 11:25:56 PDT
Created attachment 104532 [details]
Patch
Comment 2 James Robinson 2011-08-19 11:38:03 PDT
Comment on attachment 104532 [details]
Patch

Cool
Comment 3 Adrienne Walker 2011-08-19 12:23:41 PDT
(In reply to comment #2)
> (From update of attachment 104532 [details])
> Cool

Hrm.  I though tthis was going to be sufficient, but in looking at this further, it looks like I might need to s/GraphicsContext3D/LayerRendererChromium/ in this patch so that I can get the skia context from it from it for the accelerated painting update.

Alternatively, maybe this isn't patch isn't so big that I can just wrap this fix into bug 66433.
Comment 4 James Robinson 2011-08-19 13:10:59 PDT
It should be fine to do this and leave the layerRenderer()->context() getter alone, so the old codepaths work, and then later go and remove the getters.
Comment 5 Adrienne Walker 2011-08-19 14:20:38 PDT
(In reply to comment #4)
> It should be fine to do this and leave the layerRenderer()->context() getter alone, so the old codepaths work, and then later go and remove the getters.

Ok, I think I can make it work and will just land it as-is.
Comment 6 Adrienne Walker 2011-08-19 14:24:30 PDT
Committed r93438: <http://trac.webkit.org/changeset/93438>