[webkit-reviews] review denied: [Bug 56825] [Qt] fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html fails : [Attachment 86922] Ooops, last patch had invalid git diff

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Fri Mar 25 06:18:45 PDT 2011


Benjamin Poulain <benjamin at webkit.org> has denied Jarkko Sakkinen
<jarkko.j.sakkinen at gmail.com>'s request for review:
Bug 56825: [Qt]
fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html fails
https://bugs.webkit.org/show_bug.cgi?id=56825

Attachment 86922: Ooops, last patch had invalid git diff
https://bugs.webkit.org/attachment.cgi?id=86922&action=review

------- Additional Comments from Benjamin Poulain <benjamin at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=86922&action=review

quick review

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:313
> +    else
> +	   m_glWidget = new QGLWidget();

Shouldn't we get rid of this? Just set m_valid = false and return?
If the viewport does not have QGLWidget -> too bad for you, you won't enjoy
webgl :)

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:-325
> -    m_attrs.alpha = format.alpha();

Why can you skip checking the alpha from the GL surface?
You use m_attrs.alpha in reshape(), but as far as I can see, it is initialized
with the default value.

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:-327
> -    m_attrs.depth = format.depth();
> -    m_attrs.stencil = format.stencil();

Ditto :)

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:322
> +    // No need for widgets viewport. We use QGLWidget only to get a separate

> +    // OpenGL context.

The comment can be on one line, we don't limit lines at 80 characters in
WebKit.

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:428
> +	   m_valid = false;

This assignement is useless, the condition to enter this branch is
if (!m_valid)

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:507
> +    // Construct canvas FBO

Comments must be sentence, so they must end with a full stop.

> Source/WebCore/platform/graphics/qt/GraphicsContext3DQt.cpp:512
> +#if !defined(QT_OPENGL_ES_2)

Do you have to #ifdef here, it looks like m_attrs.stencil is always gonna be
false for QT_OPENGL_ES_2 because of the way it is initialized.

I am not aware of the problem of the stencil buffer with OpenGL ES 2. So my
question is: would the following call make sense if stencil problems were fixed
of GL ES 2?


More information about the webkit-reviews mailing list