From 36f0f0318275f65f8744ec6f9471702e2f58e6d5 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Tue, 6 Sep 2016 19:45:51 -0700 Subject: nir: Allow opt_peephole_sel to be more aggressive in flattening IFs. VC4 was running into a major performance regression from enabling control flow in the glmark2 conditionals test, because of short if statements containing an ffract. This pass seems like it was was trying to ensure that we only flattened IFs that should be entirely a win by guaranteeing that there would be fewer bcsels than there were MOVs otherwise. However, if the number of ALU ops is small, we can avoid the overhead of branching (which itself costs cycles) and still get a win, even if it means moving real instructions out of the THEN/ELSE blocks. For now, just turn on aggressive flattening on vc4. i965 will need some tuning to avoid regressions. It does looks like this may be useful to replace freedreno code. Improves glmark2 -b conditionals:fragment-steps=5:vertex-steps=0 from 47 fps to 95 fps on vc4. vc4 shader-db: total instructions in shared programs: 101282 -> 99543 (-1.72%) instructions in affected programs: 17365 -> 15626 (-10.01%) total uniforms in shared programs: 31295 -> 31172 (-0.39%) uniforms in affected programs: 3580 -> 3457 (-3.44%) total estimated cycles in shared programs: 225182 -> 223746 (-0.64%) estimated cycles in affected programs: 26085 -> 24649 (-5.51%) v2: Update shader-db output. Reviewed-by: Ian Romanick (v1) --- src/gallium/drivers/vc4/vc4_program.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src/gallium/drivers/vc4/vc4_program.c') diff --git a/src/gallium/drivers/vc4/vc4_program.c b/src/gallium/drivers/vc4/vc4_program.c index 986a1ff..81c6716 100644 --- a/src/gallium/drivers/vc4/vc4_program.c +++ b/src/gallium/drivers/vc4/vc4_program.c @@ -1430,7 +1430,7 @@ vc4_optimize_nir(struct nir_shader *s) NIR_PASS(progress, s, nir_opt_dce); NIR_PASS(progress, s, nir_opt_dead_cf); NIR_PASS(progress, s, nir_opt_cse); - NIR_PASS(progress, s, nir_opt_peephole_select); + NIR_PASS(progress, s, nir_opt_peephole_select, 8); NIR_PASS(progress, s, nir_opt_algebraic); NIR_PASS(progress, s, nir_opt_constant_folding); NIR_PASS(progress, s, nir_opt_undef); -- cgit v1.1