fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-05 20:28:04 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

pan/bi: Run nir_lower_all_phis_to_scalar() late

Browse source 
We were running this in the preprocess step and then trusting that it
would clean up everything before we got to the back-end.  However, we
were running the entire optimization loop in between as well as drivers
potentially adding stuff (since panvk has it's own passes after
postprocess).  Instead, this should be one of the last things run, right
before we go into the back-end.

Cc: mesa-stable
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Acked-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38821>
This commit is contained in:
Faith Ekstrand 2025-11-17 15:06:04 -05:00 • committed by Marge Bot
parent 5f22dfa2e1
commit 2bd282a968
1 changed files with 11 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
src/panfrost/compiler/bifrost/bifrost_compile.c
View file

@ -6364,7 +6364,6 @@ bifrost_postprocess_nir(nir_shader *nir, unsigned gpu_id)
NIR_PASS(_, nir, nir_lower_alu_width, bi_vectorize_filter, &gpu_id);
NIR_PASS(_, nir, nir_lower_load_const_to_scalar);
NIR_PASS(_, nir, nir_lower_phis_to_scalar, bi_vectorize_filter, &gpu_id);
NIR_PASS(_, nir, nir_lower_flrp, 16 | 32 | 64, false /* always_precise */);
NIR_PASS(_, nir, nir_lower_var_copies);
NIR_PASS(_, nir, nir_lower_alu);
@ -6984,6 +6983,17 @@ bifrost_compile_shader_nir(nir_shader *nir,
bi_optimize_nir(nir, inputs->gpu_id, inputs->robust2_modes);
{
bool scalar_phis_pass = false;
unsigned gpu_id = inputs->gpu_id;
NIR_PASS(scalar_phis_pass, nir, nir_lower_phis_to_scalar,
bi_vectorize_filter, &gpu_id);
if (scalar_phis_pass) {
NIR_PASS(_, nir, nir_opt_copy_prop);
NIR_PASS(_, nir, nir_opt_dce);
}
}
info->tls_size = nir->scratch_size;
pan_nir_collect_varyings(nir, info, PAN_MEDIUMP_VARY_32BIT);