Fix PostProcessing shader garbage on screen. #882
Conversation
| } | ||
| else | ||
| { | ||
| std::unique_ptr<unsigned char[]> clear_colour(new unsigned char[width * height * 4]()); |
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
Why do you check for gl4.4 for such an easy extension? You'll lock out mesa for years :/ Also, I don't see why we should depend on an optional extension for non-performance critical loads. imo it's fine here to just call glClear before the common drawings. We already have a bound framebuffer there. If you want to save some GPU cycles, add a flag if this texture was already cleared. |
Sonicadvance1
force-pushed
the
fix-pp-blackness
branch
4 times, most recently
from
7f7ec8d
to
dd27608
Compare
|
If this fixes 7574, it also fixes 7521 because 7574 is a duplicate of my bug report. I'm happy to have this fixed, though. |
|
LGTM if it doesn't impact performance too badly (glClear can end up serializing stuff on some drivers?). @JMC47 care to quickly test in a few games, at both low and high IR? |
|
|
||
| #include "VideoBackends/OGL/GLExtensions/gl_common.h" | ||
|
|
||
| extern PFNGLBINDBUFFERSBASEPROC glBindBuffersBase; |
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
I haven't got why you include two unused extension, but otherwise: LGTM |
We were generating a texture without ever setting the data to a known value. This happened on the old code as well, just that PP shaders are receiving some love and people are using it and noticing some of its issues.
Sonicadvance1
force-pushed
the
fix-pp-blackness
branch
from
dd27608
to
ca68526
Compare
|
LGTM |
Fix PostProcessing shader garbage on screen.
This fixes issue 7574 where we are sampling from uninitialized texture data which can basically be anything from VRAM.