intel/compiler: Be more conservative about subgroup sizes in GL

The rules for gl_SubgroupSize in Vulkan require that it be a constant
that can be queried through the API.  However, all GL requires is that
it's a uniform.  Instead of always claiming that the subgroup size in
the shader is 32 in GL like we have to do for Vulkan, claim 8 for
geometry stages, the maximum for fragment shaders, and the actual size
for compute.

Reviewed-by: Caio Marcelo de Oliveira Filho <[email protected]>

8 files changed, 50 insertions, 13 deletions

diff --git a/src/intel/compiler/brw_compiler.h b/src/intel/compiler/brw_compiler.h
index 8f8cca79deb..ba95df5b7dc 100644
--- a/src/intel/compiler/brw_compiler.h
+++ b/src/intel/compiler/brw_compiler.h
@@ -203,9 +203,18 @@ struct brw_sampler_prog_key_data {
    float scale_factors[32];
 };
 
+/** An enum representing what kind of input gl_SubgroupSize is. */
+enum PACKED brw_subgroup_size_type
+{
+   BRW_SUBGROUP_SIZE_API_CONSTANT,  /**< Vulkan behavior */
+   BRW_SUBGROUP_SIZE_UNIFORM,       /**< OpenGL behavior */
+};
+
 struct brw_base_prog_key {
    unsigned program_string_id;
 
+   enum brw_subgroup_size_type subgroup_size_type;
+
    struct brw_sampler_prog_key_data tex;
 };
 
diff --git a/src/intel/compiler/brw_fs.cpp b/src/intel/compiler/brw_fs.cpp
index 30e04ddc318..4ab50aad11f 100644
--- a/src/intel/compiler/brw_fs.cpp
+++ b/src/intel/compiler/brw_fs.cpp
@@ -7967,7 +7967,9 @@ brw_compile_fs(const struct brw_compiler *compiler, void *log_data,
 {
    const struct gen_device_info *devinfo = compiler->devinfo;
 
-   brw_nir_apply_key(shader, compiler, &key->base, true);
+   unsigned max_subgroup_size = unlikely(INTEL_DEBUG & DEBUG_DO32) ? 32 : 16;
+
+   brw_nir_apply_key(shader, compiler, &key->base, max_subgroup_size, true);
    brw_nir_lower_fs_inputs(shader, devinfo, key);
    brw_nir_lower_fs_outputs(shader);
 
@@ -8228,7 +8230,7 @@ compile_cs_to_nir(const struct brw_compiler *compiler,
                   unsigned dispatch_width)
 {
    nir_shader *shader = nir_shader_clone(mem_ctx, src_shader);
-   brw_nir_apply_key(shader, compiler, &key->base, true);
+   brw_nir_apply_key(shader, compiler, &key->base, dispatch_width, true);
 
    NIR_PASS_V(shader, brw_nir_lower_cs_intrinsics, dispatch_width);
 
diff --git a/src/intel/compiler/brw_nir.c b/src/intel/compiler/brw_nir.c
index 6bac6636c47..9e4b33c8b49 100644
--- a/src/intel/compiler/brw_nir.c
+++ b/src/intel/compiler/brw_nir.c
@@ -809,13 +809,6 @@ brw_postprocess_nir(nir_shader *nir, const struct brw_compiler *compiler,
 
    UNUSED bool progress; /* Written by OPT */
 
-   const nir_lower_subgroups_options subgroups_options = {
-      .subgroup_size = BRW_SUBGROUP_SIZE,
-      .ballot_bit_size = 32,
-      .lower_subgroup_masks = true,
-   };
-   OPT(nir_lower_subgroups, &subgroups_options);
-
    OPT(brw_nir_lower_mem_access_bit_sizes);
 
    do {
@@ -973,16 +966,48 @@ brw_nir_apply_sampler_key(nir_shader *nir,
    return nir_lower_tex(nir, &tex_options);
 }
 
+static unsigned
+get_subgroup_size(const struct brw_base_prog_key *key,
+                  unsigned max_subgroup_size)
+{
+   switch (key->subgroup_size_type) {
+   case BRW_SUBGROUP_SIZE_API_CONSTANT:
+      /* We have to use the global constant size. */
+      return BRW_SUBGROUP_SIZE;
+
+   case BRW_SUBGROUP_SIZE_UNIFORM:
+      /* It has to be uniform across all invocations but can vary per stage
+       * if we want.  This gives us a bit more freedom.
+       *
+       * For compute, brw_nir_apply_key is called per-dispatch-width so this
+       * is the actual subgroup size and not a maximum.  However, we only
+       * invoke one size of any given compute shader so it's still guaranteed
+       * to be uniform across invocations.
+       */
+      return max_subgroup_size;
+   }
+
+   unreachable("Invalid subgroup size type");
+}
+
 void
 brw_nir_apply_key(nir_shader *nir,
                   const struct brw_compiler *compiler,
                   const struct brw_base_prog_key *key,
+                  unsigned max_subgroup_size,
                   bool is_scalar)
 {
    bool progress = false;
 
    OPT(brw_nir_apply_sampler_key, compiler, &key->tex);
 
+   const nir_lower_subgroups_options subgroups_options = {
+      .subgroup_size = get_subgroup_size(key, max_subgroup_size),
+      .ballot_bit_size = 32,
+      .lower_subgroup_masks = true,
+   };
+   OPT(nir_lower_subgroups, &subgroups_options);
+
    if (progress)
       brw_nir_optimize(nir, compiler, is_scalar, false);
 }
diff --git a/src/intel/compiler/brw_nir.h b/src/intel/compiler/brw_nir.h
index 5f1843c6280..d3130476ab1 100644
--- a/src/intel/compiler/brw_nir.h
+++ b/src/intel/compiler/brw_nir.h
@@ -140,6 +140,7 @@ void brw_nir_apply_tcs_quads_workaround(nir_shader *nir);
 void brw_nir_apply_key(nir_shader *nir,
                        const struct brw_compiler *compiler,
                        const struct brw_base_prog_key *key,
+                       unsigned max_subgroup_size,
                        bool is_scalar);
 
 enum brw_reg_type brw_type_for_nir_type(const struct gen_device_info *devinfo,
diff --git a/src/intel/compiler/brw_shader.cpp b/src/intel/compiler/brw_shader.cpp
index 686c5c3a46d..5f86e251c33 100644
--- a/src/intel/compiler/brw_shader.cpp
+++ b/src/intel/compiler/brw_shader.cpp
@@ -1244,7 +1244,7 @@ brw_compile_tes(const struct brw_compiler *compiler,
    nir->info.inputs_read = key->inputs_read;
    nir->info.patch_inputs_read = key->patch_inputs_read;
 
-   brw_nir_apply_key(nir, compiler, &key->base, is_scalar);
+   brw_nir_apply_key(nir, compiler, &key->base, 8, is_scalar);
    brw_nir_lower_tes_inputs(nir, input_vue_map);
    brw_nir_lower_vue_outputs(nir);
    brw_postprocess_nir(nir, compiler, is_scalar);
diff --git a/src/intel/compiler/brw_vec4.cpp b/src/intel/compiler/brw_vec4.cpp
index 5f7decd5410..2014f95f9c7 100644
--- a/src/intel/compiler/brw_vec4.cpp
+++ b/src/intel/compiler/brw_vec4.cpp
@@ -2845,7 +2845,7 @@ brw_compile_vs(const struct brw_compiler *compiler, void *log_data,
                char **error_str)
 {
    const bool is_scalar = compiler->scalar_stage[MESA_SHADER_VERTEX];
-   brw_nir_apply_key(shader, compiler, &key->base, is_scalar);
+   brw_nir_apply_key(shader, compiler, &key->base, 8, is_scalar);
 
    const unsigned *assembly = NULL;
 
diff --git a/src/intel/compiler/brw_vec4_gs_visitor.cpp b/src/intel/compiler/brw_vec4_gs_visitor.cpp
index c45e284e07e..b7476afb466 100644
--- a/src/intel/compiler/brw_vec4_gs_visitor.cpp
+++ b/src/intel/compiler/brw_vec4_gs_visitor.cpp
@@ -639,7 +639,7 @@ brw_compile_gs(const struct brw_compiler *compiler, void *log_data,
                        &c.input_vue_map, inputs_read,
                        shader->info.separate_shader);
 
-   brw_nir_apply_key(shader, compiler, &key->base, is_scalar);
+   brw_nir_apply_key(shader, compiler, &key->base, 8, is_scalar);
    brw_nir_lower_vue_inputs(shader, &c.input_vue_map);
    brw_nir_lower_vue_outputs(shader);
    brw_postprocess_nir(shader, compiler, is_scalar);
diff --git a/src/intel/compiler/brw_vec4_tcs.cpp b/src/intel/compiler/brw_vec4_tcs.cpp
index d82439f2771..a107a9b6630 100644
--- a/src/intel/compiler/brw_vec4_tcs.cpp
+++ b/src/intel/compiler/brw_vec4_tcs.cpp
@@ -397,7 +397,7 @@ brw_compile_tcs(const struct brw_compiler *compiler,
                             nir->info.outputs_written,
                             nir->info.patch_outputs_written);
 
-   brw_nir_apply_key(nir, compiler, &key->base, is_scalar);
+   brw_nir_apply_key(nir, compiler, &key->base, 8, is_scalar);
    brw_nir_lower_vue_inputs(nir, &input_vue_map);
    brw_nir_lower_tcs_outputs(nir, &vue_prog_data->vue_map,
                              key->tes_primitive_mode);