softpipe: separate case for PIPE_PRIM_POLYGON in sp_vbuf_draw()
Because of flat shading, we can't use same code as PIPE_PRIM_TRIANGLE_FAN.
This is a follow-on to commit a59575d8fb.
(cherry picked from commit 086ecea179)
This is about a 30% performance win in OA with high settings on my GM45,
and experiments with 915GM indicate that it'll be around a 20% win there.
Currently, 915-class hardware is seriously hurt by the fact that we use
fence regs to control the tiling even for 3D instructions that could live
without them, so we spend a bunch of time waiting on previous rendering in
order to pull fences off. Thus, the texture_tiling driconf option defaults
off there for now.
swrast: always do span clipping in _swrast_write_rgba_span()
It's possible for mis-behaving vertex programs to produce vertex data
with very large/NaN values. This doesn't get handled reliably by the
clipper code so we may try to rasterize triangles that extend beyond
the viewport/window. Always clip spans to avoid invalid memory accesses
later.
This was originally taken from i915 and it shows.
Basically most the stuff in r300_render.c was never needed and
shouldn't have worked in the first place
util: Fix 24 to 32 bit expansion binary arithmetic expression.
When approaching y = x * 0xffffffff / 0xffffff with bit arithmetic, the
8 least significant bits of y should come from the
8 most significant bits of x.
intel: Clip to window after calling Driver.TexImage2D
This prevents the width / height from being clipped to the window size before
the texture is allocated. This matches intelCopyTexImage1D.
This should fix bug #21227
Signed-off-by: Ian Romanick <ian.romanick@intel.com>
(cherry picked from commit 129f311673)
intel: Clip to window after calling Driver.TexImage2D
This prevents the width / height from being clipped to the window size before
the texture is allocated. This matches intelCopyTexImage1D.
This should fix bug #21227
Signed-off-by: Ian Romanick <ian.romanick@intel.com>