[gstreamer-bugs] [Bug 615696] Shaders using GLSL 1.20 without #version.
GStreamer (bugzilla.gnome.org)
bugzilla at gnome.org
Wed Apr 14 00:56:08 PDT 2010
https://bugzilla.gnome.org/show_bug.cgi?id=615696
GStreamer | gst-plugins-gl | git
Filippo Argiolas <fargiolas> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
CC| |fargiolas at gnome.org
Ever Confirmed|0 |1
--- Comment #1 from Filippo Argiolas <fargiolas at gnome.org> 2010-04-14 07:56:04 UTC ---
Hey Eric,
Thanks for looking at this, having a compiler developer check the shaders is a
lot better than just guessing what could have happened by looking at the result
(when I wrote the shaders mesa glsl stuff was pretty young and bleeding edge).
About blur and laplacian, I thought I had ported them to use the fragment
sources in gst/gl/effects/gstgleffectssources.c. If you look at the convolution
fragments there, the #version 120 requirement had already been removed together
with array constructors.
The kernel[i] != 0 thing was meant to be used in a generic convolution source,
as a little optimization, as I thought avoiding a texture lookup could be worth
some extra comparison.
Same for norm_offset in blur shaders, they were meant to be as generic as
possible.
Now the question is: is the current approach with all the shaders in one file
the best one? this implies, is it better to optimize blur and laplacian for
their particular kernels, remove offsets, remove != 0 comparisons, or just to
port them to the generic unoptimized convolution sources in effectssources.c?
It was done this way because most of the shaders are shared between at least a
couple of effects, sometimes with different parameters and I thought it was
worth having everything in one place instead of duplicating code (apparently I
forgot about filterblur and filterlaplacian, though).
--
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.
More information about the Gstreamer-bugs
mailing list