[gstreamer-bugs] [Bug 615696] Shaders using GLSL 1.20 without #version.

[GStreamer (bugzilla.gnome.org)]

https://bugzilla.gnome.org/show_bug.cgi?id=615696
  GStreamer | gst-plugins-gl | git

Filippo Argiolas <fargiolas> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #6 from Filippo Argiolas <fargiolas at gnome.org> 2010-04-30 14:25:22 UTC ---
(In reply to comment #5)
> Looks like nouveau is claiming support for GLSL without actually supporting
> even the required features of the GLSL in GLES2.  The new compiler will be
> making that easier to support, but until then I keep the 915 version of the
> same problem hidden under a driconf option :)

About nouveau they actually tell you there is no support for anything 3D
related and they want no bug report about it :) So I guess it's just a matter
of waiting them to support what's missing.

About i915, thank you for pointing me to driconf :) As you might already know
I've finally found a good use for my netbook and started (and now finished)
porting all the shaders in gst-gl to work there too.

> Now, we're dreaming of making Mesa super smart and watch the uniforms used for
> a shader, then spawn a thread to go off and recompile a version where constant
> uniforms are turned into constants so we can have that version on hand if you
> keep your uniforms constant.  That's a loooong way down the road, though.

Ok, I've removed "constant" uniforms in some shader but not all yet since I'd
like to move all the shaders in dedicated files and bake them with constant
parameters at runtime but that will be a long time task as I really don't have
time now for not-so-funny things like these :P
Thank you again for all the precious information that helped me to better
understand how everything works.

> The next thing I wanted to check for this patch series was whether the if
> statement was helping even for the kernel that was 4/9 0.0s.  My bet is that
> removing the if statement will be a big win.

I've pushed a bunch of changes to make everything work on i915. Removed array
constructors, removed if statements, removed atan (which let some subtle IF
sneak in) calls, optimized shaders that used too much instructions, avoided
extra indirections, and probably more I don't remember. Hope I didn't forget
anything or introduce any regression :P
Making everything work with a limited instruction subset and limited hardware
was indeed great, it made me spot a lot of stupid things that I did when I
wrote the shaders and forced me to make everything run faster :)

So I think this bug is fixed, please let me know if you find anything wrong or
want to teach me anything else :)

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.