Bug 56824

Summary: [Qt] WebGL: renderbufferStorage() fails with internal format GL_RGBA4
Product: WebKit Reporter: Jarkko Sakkinen <jarkko.j.sakkinen>
Component: WebGLAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: abarth, benjamin, commit-queue, eric, jarkko.j.sakkinen, webkit.review.bot
Priority: P2 Keywords: Qt, QtTriaged
Version: 528+ (Nightly build)   
Hardware: PC   
OS: All   
Attachments:
Description Flags
Fallback to GL_RGBA on desktop OpenGL.
benjamin: review-
Fixed defects that benjaminp reported. none

Jarkko Sakkinen
Reported 2011-03-22 05:23:43 PDT
On desktop WebGL method renderbufferStorage() fails with internal format GL_RGBA4 This is shown by the layout test LayoutTests/fast/canvas/webgl/uninitialized-test.html
Attachments
Fallback to GL_RGBA on desktop OpenGL. (1.68 KB, patch)
2011-03-22 05:36 PDT, Jarkko Sakkinen 		benjamin: review-
DetailsFormatted DiffDiff
Fixed defects that benjaminp reported. (2.10 KB, patch)
2011-03-22 07:44 PDT, Jarkko Sakkinen 		no flags
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Jarkko Sakkinen
Comment 1 2011-03-22 05:24:40 PDT
I'm implementing fix in the spirit of this message: http://www.khronos.org/webgl/public-mailing-list/archives/1010/msg00126.html
Jarkko Sakkinen
Comment 2 2011-03-22 05:36:09 PDT
Created attachment 86455 [details] Fallback to GL_RGBA on desktop OpenGL.
Benjamin Poulain
Comment 3 2011-03-22 06:11:27 PDT
Comment on attachment 86455 [details] Fallback to GL_RGBA on desktop OpenGL. View in context: https://bugs.webkit.org/attachment.cgi?id=86455&action=review > Source/WebCore/ChangeLog:10 > + I would prefer you mention the test you are fixing with this patch in the changelog. Just so it is explicit how this change is tested. > Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:1099 > +#if !defined (QT_OPENGL_ES_2) We usually do not put space between defined and the parenthesis. > Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:1100 > + if (internalformat == RGBA4 || internalformat == RGB565 || internalformat == RGB5_A1) What about DEPTH_STENCIL and DEPTH_COMPONENT16? (just saw those have special case in GraphicsContext3DOpenGL).
Jarkko Sakkinen
Comment 4 2011-03-22 06:36:02 PDT
You're right about your comments about GraphicsContext3DOpenGL. In this case we should use the same logic as there. I'll revise the patch. Thanks.
Jarkko Sakkinen
Comment 5 2011-03-22 07:44:41 PDT
Created attachment 86460 [details] Fixed defects that benjaminp reported.
WebKit Commit Bot
Comment 6 2011-03-22 09:21:23 PDT
Comment on attachment 86460 [details] Fixed defects that benjaminp reported. Clearing flags on attachment: 86460 Committed r81672: <http://trac.webkit.org/changeset/81672>
WebKit Commit Bot
Comment 7 2011-03-22 09:21:27 PDT
All reviewed patches have been landed. Closing bug.
WebKit Review Bot
Comment 8 2011-03-22 11:39:53 PDT
http://trac.webkit.org/changeset/81672 might have broken GTK Linux 32-bit Debug The following tests are not passing: inspector/debugger/debug-inlined-scripts.html
Note You need to log in before you can comment on or make changes to this bug.