| Age | Commit message (Collapse) | Author |
|---|
|
|
|
We want to use the pipe_buffer_* inlines everywhere, but a pipe context
is not always available nor is it needed.
|
|
|
|
long shaders
|
|
1. #define MAX_INPUTS/OUTPUTS/TEMPS/etc with better values.
2. Add assertions in aos_get_x86() to check register file indexes
3. Assert that constant regs haven't changed after running SSE code.
|
|
|
|
Second loop was never run. Spotted by Keith.
|
|
The freeSlabs list is not really needed as we free empty slabs immediately.
Time based cached is done separately.
|
|
In the hope of MSVC inline some more functions, but without much result.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
In p_thread.h replace _glthread_* functions with new pipe_* functions.
Remove other old cruft.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Also, rename p_tile.[ch] to u_tile.[ch]
|
|
|
|
|
|
|
|
|
|
The two functions are identical. Removed align_int() from p_util.h
|
|
|
|
More can be done...
|
|
|
|
|
|
New code surrounded with #if FAST_MATH to allow comparing against original
code if we need to debug.
|
|
So far, optimized/low-precision versions of exp(), exp2(), log2(), pow().
|
|
We were sometimes referencing an invalid vertex.
Fixes progs/trivial/line-clip.c test among others.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Implement NVIDIA [-1;+1] saturate mode.
|
|
|
|
|
|
|
|
|
|
Fixes a few glean glsl regressions.
|
