[webkit-reviews] review granted: [Bug 200852] [WebGPU] Implement GPUErrors for and relax GPUBuffer validation rules : [Attachment 377394] Patch

[bugzilla-daemon at webkit.org]

Dean Jackson <dino at apple.com> has granted Justin Fan <justin_fan at apple.com>'s
request for review:
Bug 200852: [WebGPU] Implement GPUErrors for and relax GPUBuffer validation
rules
https://bugs.webkit.org/show_bug.cgi?id=200852

Attachment 377394: Patch

https://bugs.webkit.org/attachment.cgi?id=377394&action=review




--- Comment #5 from Dean Jackson <dino at apple.com> ---
Comment on attachment 377394
  --> https://bugs.webkit.org/attachment.cgi?id=377394
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=377394&action=review

> Source/WebCore/platform/graphics/gpu/GPUBuffer.h:99
> +    void registerMappingCallback(MappingCallback&&, bool, GPUErrorScopes&);
> +    void unmap(GPUErrorScopes*);
> +    void destroy(GPUErrorScopes*);

Why isn't the GPUErrorScope attached to the buffer itself? Or do different map
and unmap calls have different scopes?

Also, why are these pointers? I see it is because destroy can be called with a
nullptr from the destructor, but maybe we should have a default error scope
that always exists.