diff options
author | Iago Toral Quiroga <itoral@igalia.com> | 2015-01-20 17:09:59 +0100 |
---|---|---|
committer | Iago Toral Quiroga <itoral@igalia.com> | 2015-02-03 13:19:36 +0100 |
commit | 5dfb085ff325df3dbefda515f06106469babbefc (patch) | |
tree | acf8e618b2e412ca3846a98d96e9484841ae4496 /src/glsl/ir_optimization.h | |
parent | c27d23f0c8e952d8ec340c4c9cd3ad026cd6187b (diff) | |
download | external_mesa3d-5dfb085ff325df3dbefda515f06106469babbefc.zip external_mesa3d-5dfb085ff325df3dbefda515f06106469babbefc.tar.gz external_mesa3d-5dfb085ff325df3dbefda515f06106469babbefc.tar.bz2 |
glsl: Improve precision of mod(x,y)
Currently, Mesa uses the lowering pass MOD_TO_FRACT to implement
mod(x,y) as y * fract(x/y). This implementation has a down side though:
it introduces precision errors due to the fract() operation. Even worse,
since the result of fract() is multiplied by y, the larger y gets the
larger the precision error we produce, so for large enough numbers the
precision loss is significant. Some examples on i965:
Operation Precision error
-----------------------------------------------------
mod(-1.951171875, 1.9980468750) 0.0000000447
mod(121.57, 13.29) 0.0000023842
mod(3769.12, 321.99) 0.0000762939
mod(3769.12, 1321.99) 0.0001220703
mod(-987654.125, 123456.984375) 0.0160663128
mod( 987654.125, 123456.984375) 0.0312500000
This patch replaces the current lowering pass with a different one
(MOD_TO_FLOOR) that follows the recommended implementation in the GLSL
man pages:
mod(x,y) = x - y * floor(x/y)
This implementation eliminates the precision errors at the expense of
an additional add instruction on some systems. On systems that can do
negate with multiply-add in a single operation this new implementation
would come at no additional cost.
v2 (Ian Romanick)
- Do not clone operands because when they are expressions we would be
duplicating them and that can lead to suboptimal code.
Fixes the following 16 dEQP tests:
dEQP-GLES3.functional.shaders.builtin_functions.precision.mod.mediump_*
dEQP-GLES3.functional.shaders.builtin_functions.precision.mod.highp_*
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Diffstat (limited to 'src/glsl/ir_optimization.h')
-rw-r--r-- | src/glsl/ir_optimization.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/glsl/ir_optimization.h b/src/glsl/ir_optimization.h index 34e0b4b..912d910 100644 --- a/src/glsl/ir_optimization.h +++ b/src/glsl/ir_optimization.h @@ -34,7 +34,7 @@ #define EXP_TO_EXP2 0x04 #define POW_TO_EXP2 0x08 #define LOG_TO_LOG2 0x10 -#define MOD_TO_FRACT 0x20 +#define MOD_TO_FLOOR 0x20 #define INT_DIV_TO_MUL_RCP 0x40 #define BITFIELD_INSERT_TO_BFM_BFI 0x80 #define LDEXP_TO_ARITH 0x100 |