648b9865d8
We need to explicitly round when converting colors from float to uint because multiplying a normalized float by 255 might not result in a whole number. (The exact result here may vary depending on your drivers/hardware.) Ideally, we shouldn't be using floating point here, but fixing that is a much more complicated patch. Fixes gxtest TEV tests using Intel HD 4000. |
||
---|---|---|
.. | ||
AudioCommon | ||
Common | ||
Core | ||
DiscIO | ||
DolphinWX | ||
InputCommon | ||
VideoBackends | ||
VideoCommon | ||
CMakeLists.txt |