[Webkit-unassigned] [Bug 50833] [chromium] Add support to compositor to composite to offscreen texture.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=50833





--- Comment #22 from Kenneth Russell <kbr at google.com>  2010-12-22 12:04:54 PST ---
(In reply to comment #20)
> (In reply to comment #19)
> > Created an attachment (id=77114)
 --> (https://bugs.webkit.org/attachment.cgi?id=77114&action=review) [details] [details]
> > Patch
> 
> I've attached a revised patch which I've manually tested with the WebGL body-viewer demo.
> 
> I've attempted to run the full test suite: the update/build commands had problems (see below), but I manually built debug versions of chrome, test_shell and DumpRenderTree and ran:
> 
> $ ./Tools/Scripts/new-run-webkit-tests --chromium --debug --platform=chromium-gpu
> 
> and I got these results:
> 
>   canvas/philip/tests/2d.imageData.get.source.outside.html -> unexpected pass
>   canvas/philip/tests/2d.shadow.transform.2.html -> unexpected pass
>   fast/canvas/canvas-scale-fillPath-shadow.html -> unexpected pass
>   fast/canvas/canvas-scale-fillRect-shadow.html -> unexpected pass
>   fast/canvas/canvas-scale-strokePath-shadow.html -> unexpected pass
> 22727 tests ran as expected, 5 didn't:                  
> 
> 
> Expected to fail, but passed: (5)
>   canvas/philip/tests/2d.imageData.get.source.outside.html
>   canvas/philip/tests/2d.shadow.transform.2.html
>   fast/canvas/canvas-scale-fillPath-shadow.html
>   fast/canvas/canvas-scale-fillRect-shadow.html
>   fast/canvas/canvas-scale-strokePath-shadow.html

I updated the GPU test expectations yesterday to expect that these five tests will pass, so this is fine.

> Now, I need to determine whether this is in fact a clean set of tests, or if the failure of the update/build commands means it is in fact not to be trusted.
> 
> Thoughts?
> 
> The build/update command failures are:
> -------------------------------------------------------------------------------
> 
> ./Tools/Scripts/update-webkit --chromium
> 
> If I do this on branch 'master' on a freshly checkout-out repo (updated once with git-pull), I get
> 
> Updating OpenSource
> 
> Unable to determine upstream SVN information from working tree history
> 
> Died at ./Tools/Scripts/update-webkit line 132
> 
> ./Tools/Scripts/build-webkit --chromium --debug
> 
> If I try to run this on a branch containing my patch (keeping in mind I couldn't get the update command to run), I get
> 
> make -fMakefile.chromium -j16 BUILDTYPE=Debug all
> make: Makefile.chromium: No such file or directory
> make: *** No rule to make target `Makefile.chromium'. Stop.
> ------------------------------------------------------------------------------

I don't know why this is failing. I use a Subversion checkout of the WebKit tree configured within my Chromium repo per http://www.chromium.org/developers/contributing-to-webkit .

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.