diff options
| author | José Fonseca <jfonseca@vmware.com> | 2010-09-04 19:51:54 +0100 |
|---|---|---|
| committer | José Fonseca <jfonseca@vmware.com> | 2010-09-05 10:17:51 +0100 |
| commit | d278ddc00966b6348eb4703b12166c05cf539635 (patch) | |
| tree | 11b1b4ef26c8b8224097a3487509bc6c32470337 /src/gallium/drivers/llvmpipe/lp_bld_interp.h | |
| parent | b8684b2458bc9bdcfd6b43dc7c2b8c2d485105fd (diff) | |
llvmpipe: Fix perspective divide interpolation.
Intuition != mathematics, so this time I actually worked out the right
formula for first order approximation of perspective interpolation.
Ironically, per quad divide actually makes things slower when compared
with per pixel divide -- probably because the divide hardware unit is
rarely used, whereas the multiply unit is typically already saturated
and the first order approximation imply more multiplications.
Diffstat (limited to 'src/gallium/drivers/llvmpipe/lp_bld_interp.h')
| -rw-r--r-- | src/gallium/drivers/llvmpipe/lp_bld_interp.h | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/src/gallium/drivers/llvmpipe/lp_bld_interp.h b/src/gallium/drivers/llvmpipe/lp_bld_interp.h index 2905513301..3054030f73 100644 --- a/src/gallium/drivers/llvmpipe/lp_bld_interp.h +++ b/src/gallium/drivers/llvmpipe/lp_bld_interp.h @@ -64,6 +64,8 @@ struct lp_build_interp_soa_context LLVMValueRef a [1 + PIPE_MAX_SHADER_INPUTS][NUM_CHANNELS]; LLVMValueRef dadq[1 + PIPE_MAX_SHADER_INPUTS][NUM_CHANNELS]; + LLVMValueRef oow; + LLVMValueRef attribs[1 + PIPE_MAX_SHADER_INPUTS][NUM_CHANNELS]; /* |