Bug 219256 – GraphicsContextGL should have robust multivalue setters

• Component: WebGL
• Status: RESOLVED FIXED

[Kimmo Kinnunen]

GraphicsContextGL should have robust multivalue setters

Instead of passing `size, ptr`
we should pass `GCGLSpan(ptr, size)`
in all functions taking buffer parameters.

[Kimmo Kinnunen]

Created attachment 414831 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414832 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414833 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414834 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414837 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414839 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414840 [details]
Patch

[Dean Jackson]

Comment on attachment 414840 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=414840&action=review

Very nice.

> Source/WebCore/html/canvas/WebGL2RenderingContext.cpp:1132
> -    m_context->getExtensions().texImage2DRobustANGLE(target, level, internalformat, width, height, border, format, type, 0, reinterpret_cast<void*>(offset));
> +    m_context->texImage2D(target, level, internalformat, width, height, border, format, type, offset);

Was this intentional? Does texImage2D call the robust ANGLE form itself?

[Kimmo Kinnunen]

Created attachment 414895 [details]
Patch

[Kimmo Kinnunen]

(In reply to Dean Jackson from comment #8)
> > Source/WebCore/html/canvas/WebGL2RenderingContext.cpp:1132
> > -    m_context->getExtensions().texImage2DRobustANGLE(target, level, internalformat, width, height, border, format, type, 0, reinterpret_cast<void*>(offset));
> > +    m_context->texImage2D(target, level, internalformat, width, height, border, format, type, offset);
> 
> Was this intentional? Does texImage2D call the robust ANGLE form itself?

Yes. Sorry for not describing it in the ChangeLog and not carrying the change through its logical conclusion.

I uploaded a new patch, which sohuld clarify this in the ChangeLog and in the code. In the code, I moved all the texImage functions from the ExtensionsGL to GraphicsContextGL.

[Kimmo Kinnunen]

Created attachment 414896 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414897 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414898 [details]
Patch

[Kimmo Kinnunen]

Created attachment 414899 [details]
Patch

[EWS]

Committed r270185: <https://trac.webkit.org/changeset/270185>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 414899 [details].

[Radar WebKit Bug Importer]

<rdar://problem/71761224>

[Philippe Normand]

Comment on attachment 414899 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=414899&action=review

> Source/WebCore/platform/graphics/GraphicsTypesGL.h:72
> +    GCGLSpan(const GCGLSpan<U, Extent>& other, std::enable_if_t<std::is_convertible_v<U(*)[], T(*)[]>, nullptr_t> = nullptr)

Source/WebCore/platform/graphics/GraphicsTypesGL.h:72:104: error: unknown type name 'nullptr_t'; did you mean 'std::nullptr_t'?
    GCGLSpan(const GCGLSpan<U, Extent>& other, std::enable_if_t<std::is_convertible_v<U(*)[], T(*)[]>, nullptr_t> = nullptr)
                                                                                                       ^~~~~~~~~
                                                                                                       std::nullptr_t
/usr/bin/../lib/gcc/x86_64-unknown-linux-gnu/10.2.0/../../../../include/c++/10.2.0/x86_64-unknown-linux-gnu/bits/c++config.h:264:29: note: 'std::nullptr_t' declared here
  typedef decltype(nullptr) nullptr_t;
                            ^

[Philippe Normand]

r270189

[Kimmo Kinnunen]

Thanks Philippe