Support rendering compute shaders straight to textures (skipping the dummy phase).
authorSteinar H. Gunderson <sgunderson@bigfoot.com>
Tue, 21 Nov 2017 23:04:03 +0000 (00:04 +0100)
committerSteinar H. Gunderson <sgunderson@bigfoot.com>
Tue, 21 Nov 2017 23:05:19 +0000 (00:05 +0100)
commit7366e74b75fa1ac3267709c12417179819b86acb
treec5c8ac14266cbaeee290d7f6f4d7880fd92fa296
parent3393f61ef94cbcb63f189767dfa0772316bd453f
Support rendering compute shaders straight to textures (skipping the dummy phase).

There are lots of limitations currently (only one destination,
only GL_RGBA16F), but it's a good start. Curiously enough,
it doesn't really help anything on the deinterlacing benchmark
for my Haswell, but NVIDIA sees ~15% improvement.
effect_chain.cpp
effect_chain.h
test_util.cpp