| Age | Commit message (Collapse) | Author |
|---|
|
Results in a fair speed improvement.
|
|
|
|
This fixes broken blits.
|
|
|
|
|
|
|
|
It's possible the current vs/fs is null when cso_save_vertex/fragment_shader()
is called.
|
|
The return value was never used.
|
|
This reverts commit 1f0f029ba6f22ef4ada01fcdc153da91571a7963.
Which broke the debug build.
|
|
Also, provide a separate flag to say whether the driver can handle
clipping/rhw tasks, in addition to the API flag which indicates they
have already been done.
|
|
|
|
|
|
|
|
|
|
Otherwise we get a symbol clash with the copy in src/gallium/util .
|
|
Also, fast path for re-fencing the same buffer multiple times with the same fence.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
lack of constant vectors handling in execution engine)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PS: pipe_shader_state should probably go away now that it is reduced to a single pointer.
|
|
|
|
cso_restore_* functions are implemented on top of cso_set_*, therefore
they require full knowledge of the current pipe state to work correctly.
Directly calling pipe's set_*_state functions will lead to undefined state.
Also save and restore shaders.
|
|
|
|
|
|
|
