diff options
| author | Kenneth Graunke <[email protected]> | 2016-01-21 16:37:20 -0800 |
|---|---|---|
| committer | Kenneth Graunke <[email protected]> | 2016-02-08 18:10:34 -0800 |
| commit | 74f956c416d5b0b37b4c2d6b957167bb203502c3 (patch) | |
| tree | dd120bc75edb496fedb54ec2ab201bdfb26199a9 /src/mesa/drivers | |
| parent | 184afd8fd9e7891322224f57a12c2e0fe52b46cb (diff) | |
i965: Use nir_lower_load_const_to_scalar().
I don't know why, but we never hooked up this pass Eric wrote.
Otherwise, you can end up with stupid scalarized code such as:
vec4 ssa_7 = load_const (0.0, 0.0, 0.0, 0.0)
vec4 ssa_8 = ...
vec1 ssa_9 = feq ssa_8, ssa_7
vec1 ssa_10 = feq ssa_8.y, ssa_7.y
vec1 ssa_11 = feq ssa_8, ssa_7.z
vec1 ssa_12 = feq ssa_8.y, ssa_7.w
ssa_8.xyxy == <0, 0, 0, 0> should only take two feq instructions.
shader-db on Skylake:
total instructions in shared programs: 9121153 -> 9120749 (-0.00%)
instructions in affected programs: 32421 -> 32017 (-1.25%)
helped: 277
HURT: 69
total cycles in shared programs: 69003364 -> 69000912 (-0.00%)
cycles in affected programs: 899186 -> 896734 (-0.27%)
helped: 313
HURT: 403
This also prevents regressions when disabling channel expressions.
v2: Don't call opt_cse afterwards (requested by Matt). It should
happen in the optimization loop below anyway.
Signed-off-by: Kenneth Graunke <[email protected]>
Reviewed-by: Eduardo Lima Mitev <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Diffstat (limited to 'src/mesa/drivers')
| -rw-r--r-- | src/mesa/drivers/dri/i965/brw_nir.c | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_nir.c b/src/mesa/drivers/dri/i965/brw_nir.c index 287f935d539..46b51163579 100644 --- a/src/mesa/drivers/dri/i965/brw_nir.c +++ b/src/mesa/drivers/dri/i965/brw_nir.c @@ -482,6 +482,10 @@ brw_preprocess_nir(nir_shader *nir, bool is_scalar) nir = nir_optimize(nir, is_scalar); + if (is_scalar) { + OPT_V(nir_lower_load_const_to_scalar); + } + /* Lower a bunch of stuff */ OPT_V(nir_lower_var_copies); |