769812 - ATI OpenGL driver renders point sprite upside-down

Status: RESOLVED FIXED

(Core :: Graphics: CanvasWebGL, defect)

Description

[Benoit Jacob [:bjacob] (mostly away)]

Attachment: work around upside-down point sprites

This is the reason why this test fails:
https://cvs.khronos.org/svn/repos/registry/trunk/public/webgl/conformance-suites/1.0.1/conformance/glsl/variables/gl-pointcoord.html

Attached patch fixes it, but will break down whenever the driver bug is fixed (in the sense that the upside-down bug will then resurface). I think we should still take it so at least we're not blocked on Apple.

Comment 1

[Kelsey Gilbert [:jgilbert]]

Comment on attachment 638009 [details] [diff] [review]
work around upside-down point sprites

Review of attachment 638009 [details] [diff] [review]:
-----------------------------------------------------------------

Can we try to get ahead on this by querying the value before we set it properly? We could put a DEBUG abort in there which will let us know when this is fixed.

Comment 2

[Benoit Jacob [:bjacob] (mostly away)]

Part of this tryserver push: https://tbpl.mozilla.org/?tree=Try&rev=35d614a1c367

Comment 3

[Benoit Jacob [:bjacob] (mostly away)]

(In reply to Jeff Gilbert [:jgilbert] from comment #1)
> Comment on attachment 638009 [details] [diff] [review]
> work around upside-down point sprites
> 
> Review of attachment 638009 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> Can we try to get ahead on this by querying the value before we set it
> properly? We could put a DEBUG abort in there which will let us know when
> this is fixed.

The bug is not at the level of which of UPPER_LEFT / LOWER_LEFT is current or default. The bug is that no matter what we ask for, the driver actually does the opposite. So it's not something we can easily detect by a GL getter call. We'd have to render a point sprite, readpixels, and look at the result.

Comment 4

[Benoit Jacob [:bjacob] (mostly away)]

http://hg.mozilla.org/integration/mozilla-inbound/rev/8faed44d0af6

Comment 5

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/8faed44d0af6