| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
Don't use a fixed-size array. Saves memory in most cases and avoids
potential overflow for long programs.
|
|
The caller might be requesting 16-bit Z + 8-bit stencil be placed in a
PIPE_FORMAT_S8Z24 or PIPE_FORMAT_Z24S8 buffer.
|
|
|
|
|
|
|
|
|
|
PIPE_OS_WINDOWS is an umbrella for all Windows variants and subsystems,
PIPE_SUBSYSTEM_WINDOWS_USER is just for user-space windows (e.g., OpenGL),
and the thread primitives currently included in p_thread only support the
later.
|
|
Basically, set up one user-space wrapper for all arrays instead of
the individual arrays.
|
|
|
|
|
|
|
|
|
|
|
|
In p_thread.h replace _glthread_* functions with new pipe_* functions.
Remove other old cruft.
|
|
|
|
|
|
Note that SPU vertex transformation is disabled at this time.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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().
|
|
|
|
|
|
|
|
everything into memory).
|
|
|
|
It often happens that new data is written directly to the user buffers
without mapping/unmapping. This hack marks user buffers and dumps them
before passing them to pipe context.
|
|
Because windows limits the ammount of memory that can be mapped.
|
