lima/ppir: enable vectorize optimization

pp has vector units and some operations can be optimized when bundled
together.
Benchmarking this with piglit shaders shows that the instruction count
can be greatly reduced on many examples with vectorize.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
This commit is contained in:
Erico Nunes 2019-08-10 22:46:02 +02:00
parent 2a8a81d109
commit 4379dcc12d
1 changed files with 5 additions and 0 deletions

View File

@ -169,6 +169,11 @@ lima_program_optimize_fs_nir(struct nir_shader *s)
NIR_PASS_V(s, nir_lower_regs_to_ssa);
NIR_PASS_V(s, nir_lower_tex, &tex_options);
do {
progress = false;
NIR_PASS(progress, s, nir_opt_vectorize);
} while (progress);
do {
progress = false;