nir: Use the flrp lowering pass instead of nir_opt_algebraic

I tried to be very careful while updating all the various drivers, but I
don't have any of that hardware for testing. :(

i965 is the only platform that sets always_precise = true, and it is
only set true for fragment shaders.  Gen4 and Gen5 both set lower_flrp32
only for vertex shaders.  For fragment shaders, nir_op_flrp is lowered
during code generation as a(1-c)+bc.  On all other platforms 64-bit
nir_op_flrp and on Gen11 32-bit nir_op_flrp are lowered using the old
nir_opt_algebraic method.

No changes on any other Intel platforms.

v2: Add panfrost changes.

Iron Lake and GM45 had similar results. (Iron Lake shown)
total cycles in shared programs: 188647754 -> 188647748 (<.01%)
cycles in affected programs: 5096 -> 5090 (-0.12%)
helped: 3
HURT: 0
helped stats (abs) min: 2 max: 2 x̄: 2.00 x̃: 2
helped stats (rel) min: 0.12% max: 0.12% x̄: 0.12% x̃: 0.12%

Reviewed-by: Matt Turner <[email protected]>

1 files changed, 24 insertions, 0 deletions

diff --git a/src/mesa/state_tracker/st_glsl_to_nir.cpp b/src/mesa/state_tracker/st_glsl_to_nir.cpp
index 97b2831b880..0a67d4532eb 100644
--- a/src/mesa/state_tracker/st_glsl_to_nir.cpp
+++ b/src/mesa/state_tracker/st_glsl_to_nir.cpp
@@ -304,6 +304,11 @@ void
 st_nir_opts(nir_shader *nir, bool scalar)
 {
    bool progress;
+   unsigned lower_flrp =
+      (nir->options->lower_flrp16 ? 16 : 0) |
+      (nir->options->lower_flrp32 ? 32 : 0) |
+      (nir->options->lower_flrp64 ? 64 : 0);
+
    do {
       progress = false;
 
@@ -332,6 +337,25 @@ st_nir_opts(nir_shader *nir, bool scalar)
       NIR_PASS(progress, nir, nir_opt_algebraic);
       NIR_PASS(progress, nir, nir_opt_constant_folding);
 
+      if (lower_flrp != 0) {
+         bool lower_flrp_progress;
+
+         NIR_PASS(lower_flrp_progress, nir, nir_lower_flrp,
+                  lower_flrp,
+                  false /* always_precise */,
+                  nir->options->lower_ffma);
+         if (lower_flrp_progress) {
+            NIR_PASS(progress, nir,
+                     nir_opt_constant_folding);
+            progress = true;
+         }
+
+         /* Nothing should rematerialize any flrps, so we only need to do this
+          * lowering once.
+          */
+         lower_flrp = 0;
+      }
+
       NIR_PASS(progress, nir, nir_opt_undef);
       NIR_PASS(progress, nir, nir_opt_conditional_discard);
       if (nir->options->max_unroll_iterations) {