Mesa (main): nir: Produce correct results for atan with NaN

[GitLab Mirror]

Module: Mesa
Branch: main
Commit: 1cb3d1a6ae027b5045e47ccf7e551bd81fc3cab2
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=1cb3d1a6ae027b5045e47ccf7e551bd81fc3cab2

Author: Ian Romanick <ian.d.romanick at intel.com>
Date:   Fri Oct 29 10:50:55 2021 -0700

nir: Produce correct results for atan with NaN

Properly handling NaN adversely affects several hundred shaders in
shader-db (lots of Skia and a few others from various synthetic
benchmarks) and fossil-db (mostly Talos and some Doom 2016).  Only apply
the NaN handling work-around when the shader demands it.

v2: Add comment explaining the 1.0*y_over_x.  Suggested by Caio.

Reviewed-by: Caio Oliveira <caio.oliveira at intel.com>
Fixes: 2098ae16c8b ("nir/builder: Move nir_atan and nir_atan2 from SPIR-V translator")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13999>

---

 src/compiler/nir/nir_builtin_builder.c | 23 ++++++++++++++++++++++-
 1 file changed, 22 insertions(+), 1 deletion(-)

diff --git a/src/compiler/nir/nir_builtin_builder.c b/src/compiler/nir/nir_builtin_builder.c
index da332e72de9..b975157cfb1 100644
--- a/src/compiler/nir/nir_builtin_builder.c
+++ b/src/compiler/nir/nir_builtin_builder.c
@@ -223,7 +223,28 @@ nir_atan(nir_builder *b, nir_ssa_def *y_over_x)
                   tmp);
 
    /* sign fixup */
-   return nir_fmul(b, tmp, nir_fsign(b, y_over_x));
+   nir_ssa_def *result = nir_fmul(b, tmp, nir_fsign(b, y_over_x));
+
+   /* The fmin and fmax above will filter out NaN values.  This leads to
+    * non-NaN results for NaN inputs.  Work around this by doing
+    *
+    *    !isnan(y_over_x) ? ... : y_over_x;
+    */
+   if (b->exact ||
+       nir_is_float_control_signed_zero_inf_nan_preserve(b->shader->info.float_controls_execution_mode, bit_size)) {
+      const bool exact = b->exact;
+
+      b->exact = true;
+      nir_ssa_def *is_not_nan = nir_feq(b, y_over_x, y_over_x);
+      b->exact = exact;
+
+      /* The extra 1.0*y_over_x ensures that subnormal results are flushed to
+       * zero.
+       */
+      result = nir_bcsel(b, is_not_nan, result, nir_fmul_imm(b, y_over_x, 1.0));
+   }
+
+   return result;
 }
 
 nir_ssa_def *