1093967 - Implement BlitFramebuffer

Status: RESOLVED FIXED

(Core :: Canvas: WebGL, defect)

Description

[Jeff Gilbert [:jgilbert]]

Attachment: 0001-Update-BindFramebuffer-and-GetParameter.patch

Comment 1

[Jeff Gilbert [:jgilbert]]

Attachment: 0002-Fix-all-instances-of-mBoundFramebuffer.patch

Comment 2

[Jeff Gilbert [:jgilbert]]

Attachment: 0003-Implement-BlitFramebuffer.patch

Comment 3

[David Anderson [:dvander] - inactive, e-mail if emergency]

Comment on attachment 8517116 [details] [diff] [review]
0002-Fix-all-instances-of-mBoundFramebuffer.patch

Review of attachment 8517116 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/canvas/WebGLContextValidate.cpp
@@ +1799,5 @@
> +        break;
> +    }
> +
> +    if (MOZ_LIKELY(isValid)) {
> +        return true;

I'd just return true in the above cases, unless something else will need to be validated in a later patch.

Comment 4

[David Anderson [:dvander] - inactive, e-mail if emergency]

Comment on attachment 8517117 [details] [diff] [review]
0003-Implement-BlitFramebuffer.patch

Review of attachment 8517117 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/canvas/WebGL2ContextFramebuffers.cpp
@@ +276,5 @@
> +     *   and destination depth and stencil buffer formats do not match.
> +     *
> +     * jgilbert: The wording is such that if only DEPTH_BUFFER_BIT is specified,
> +     * the stencil formats must match. This seems wrong. It could be a spec bug,
> +     * or I could be missing an interaction in one of the earlier paragraphs.

Yeah, I read this as "if X or Y, then x or y should match respectively."

Comment 5

[Jeff Gilbert [:jgilbert]]

The only unambiguous reading of the passage is problematic, though. 'Reasonable interpretation' doesn't really cut it for specs.

Comment 6

[Jeff Gilbert [:jgilbert]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/88cf7f2fe48f

Comment 7

[Wes Kocher (:KWierso)]

Backed out in https://hg.mozilla.org/integration/mozilla-inbound/rev/8d48c34f44ed for build bustage:
https://treeherder.mozilla.org/logviewer.html#?job_id=5300122&repo=mozilla-inbound

Comment 8

[Jeff Gilbert [:jgilbert]]

(In reply to Wes Kocher (:KWierso) from comment #7)
> Backed out in
> https://hg.mozilla.org/integration/mozilla-inbound/rev/8d48c34f44ed for
> build bustage:
> https://treeherder.mozilla.org/logviewer.html#?job_id=5300122&repo=mozilla-
> inbound

https://hg.mozilla.org/integration/mozilla-inbound/rev/32e93a36551d

The bustage was from conflicts with the shader rewrite stuff, which was also backed out.

Comment 9

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/32e93a36551d