nir/algebraic: Lower CS derivatives to zero when no group defined

In compute shaders if no derivative group is defined, the derivatives
will always be zero.  Specified in NV_compute_shader_derivatives.

To make the check more convenient, add a "info" local variable to the
generated code so we can refer to it in the Python rules.  (Jason)

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

diff --git a/src/compiler/nir/nir_algebraic.py b/src/compiler/nir/nir_algebraic.py
index d4b3bb5957f4c42d0a0950600fdc5219a1143832..4772d4d123dc1ae7787a256dcb97037fa8b42247 100644 (file)
--- a/src/compiler/nir/nir_algebraic.py
+++ b/src/compiler/nir/nir_algebraic.py
@@ -770,7 +770,9 @@ ${pass_name}(nir_shader *shader)
    bool progress = false;
    bool condition_flags[${len(condition_list)}];
    const nir_shader_compiler_options *options = shader->options;
+   const shader_info *info = &shader->info;
    (void) options;
+   (void) info;
 
    % for index, condition in enumerate(condition_list):
    condition_flags[${index}] = ${condition};

diff --git a/src/compiler/nir/nir_opt_algebraic.py b/src/compiler/nir/nir_opt_algebraic.py
index d67379958440f779166e0885519ff83d5d57467c..b8d418c2e490c4d58fe2459d21cb7d91319e4ac3 100644 (file)
--- a/src/compiler/nir/nir_opt_algebraic.py
+++ b/src/compiler/nir/nir_opt_algebraic.py
@@ -972,6 +972,18 @@ for op in ['fadd', 'fmul', 'iadd', 'imul']:
       ((op, ('bcsel(is_used_once)', a, '#b', c), '#d'), ('bcsel', a, (op, b, d), (op, c, d)))
    ]
 
+# For derivatives in compute shaders, GLSL_NV_compute_shader_derivatives
+# states:
+#
+#     If neither layout qualifier is specified, derivatives in compute shaders
+#     return zero, which is consistent with the handling of built-in texture
+#     functions like texture() in GLSL 4.50 compute shaders.
+for op in ['fddx', 'fddx_fine', 'fddx_coarse',
+           'fddy', 'fddy_fine', 'fddy_coarse']:
+   optimizations += [
+      ((op, 'a'), 0.0, 'info->stage == MESA_SHADER_COMPUTE && info->cs.derivative_group == DERIVATIVE_GROUP_NONE')
+]
+
 # This section contains "late" optimizations that should be run before
 # creating ffmas and calling regular optimizations for the final time.
 # Optimizations should go here if they help code generation and conflict