nir: Add a new subgroups lowering pass

This commit pulls nir_lower_read_invocations_to_scalar along with most
of the guts of nir_opt_intrinsics (which mostly does subgroup lowering)
into a new nir_lower_subgroups pass.  There are various other bits of
subgroup lowering that we're going to want to do so it makes a bit more
sense to keep it all together in one pass.  We also move it in i965 to
happen after nir_lower_system_values to ensure that because we want to
handle the subgroup mask system value intrinsics here.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

src/compiler/Makefile.sources

@@ -233,11 +233,11 @@ NIR_FILES = \
 	nir/nir_lower_passthrough_edgeflags.c \
 	nir/nir_lower_patch_vertices.c \
 	nir/nir_lower_phis_to_scalar.c \
-	nir/nir_lower_read_invocation_to_scalar.c \
 	nir/nir_lower_regs_to_ssa.c \
 	nir/nir_lower_returns.c \
 	nir/nir_lower_samplers.c \
 	nir/nir_lower_samplers_as_deref.c \
+	nir/nir_lower_subgroups.c \
 	nir/nir_lower_system_values.c \
 	nir/nir_lower_tex.c \
 	nir/nir_lower_to_source_mods.c \

src/compiler/nir/meson.build

@@ -120,11 +120,11 @@ files_libnir = files(
   'nir_lower_passthrough_edgeflags.c',
   'nir_lower_patch_vertices.c',
   'nir_lower_phis_to_scalar.c',
-  'nir_lower_read_invocation_to_scalar.c',
   'nir_lower_regs_to_ssa.c',
   'nir_lower_returns.c',
   'nir_lower_samplers.c',
   'nir_lower_samplers_as_deref.c',
+  'nir_lower_subgroups.c',
   'nir_lower_system_values.c',
   'nir_lower_tex.c',
   'nir_lower_to_source_mods.c',

src/compiler/nir/nir.h

@@ -1835,9 +1835,6 @@ typedef struct nir_shader_compiler_options {
    bool lower_extract_byte;
    bool lower_extract_word;
 
-   bool lower_vote_trivial;
-   bool lower_subgroup_masks;
-
    /**
     * Does the driver support real 32-bit integers?  (Otherwise, integers
     * are simulated by floats.)
@@ -2479,6 +2476,15 @@ bool nir_lower_samplers(nir_shader *shader,
 bool nir_lower_samplers_as_deref(nir_shader *shader,
                                  const struct gl_shader_program *shader_program);
 
+typedef struct nir_lower_subgroups_options {
+   bool lower_to_scalar:1;
+   bool lower_vote_trivial:1;
+   bool lower_subgroup_masks:1;
+} nir_lower_subgroups_options;
+
+bool nir_lower_subgroups(nir_shader *shader,
+                         const nir_lower_subgroups_options *options);
+
 bool nir_lower_system_values(nir_shader *shader);
 
 typedef struct nir_lower_tex_options {

src/compiler/nir/nir_lower_read_invocation_to_scalar.c

@@ -1,112 +0,0 @@
-/*
- * Copyright © 2017 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- */
-
-#include "nir.h"
-#include "nir_builder.h"
-
-/** @file nir_lower_read_invocation_to_scalar.c
- *
- * Replaces nir_intrinsic_read_invocation/nir_intrinsic_read_first_invocation
- * operations with num_components != 1 with individual per-channel operations.
- */
-
-static void
-lower_read_invocation_to_scalar(nir_builder *b, nir_intrinsic_instr *intrin)
-{
-   b->cursor = nir_before_instr(&intrin->instr);
-
-   nir_ssa_def *value = nir_ssa_for_src(b, intrin->src[0], intrin->num_components);
-   nir_ssa_def *reads[4];
-
-   for (unsigned i = 0; i < intrin->num_components; i++) {
-      nir_intrinsic_instr *chan_intrin =
-         nir_intrinsic_instr_create(b->shader, intrin->intrinsic);
-      nir_ssa_dest_init(&chan_intrin->instr, &chan_intrin->dest,
-                        1, intrin->dest.ssa.bit_size, NULL);
-      chan_intrin->num_components = 1;
-
-      /* value */
-      chan_intrin->src[0] = nir_src_for_ssa(nir_channel(b, value, i));
-      /* invocation */
-      if (intrin->intrinsic == nir_intrinsic_read_invocation)
-         nir_src_copy(&chan_intrin->src[1], &intrin->src[1], chan_intrin);
-
-      nir_builder_instr_insert(b, &chan_intrin->instr);
-
-      reads[i] = &chan_intrin->dest.ssa;
-   }
-
-   nir_ssa_def_rewrite_uses(&intrin->dest.ssa,
-                            nir_src_for_ssa(nir_vec(b, reads,
-                                                    intrin->num_components)));
-   nir_instr_remove(&intrin->instr);
-}
-
-static bool
-nir_lower_read_invocation_to_scalar_impl(nir_function_impl *impl)
-{
-   bool progress = false;
-   nir_builder b;
-   nir_builder_init(&b, impl);
-
-   nir_foreach_block(block, impl) {
-      nir_foreach_instr_safe(instr, block) {
-         if (instr->type != nir_instr_type_intrinsic)
-            continue;
-
-         nir_intrinsic_instr *intrin = nir_instr_as_intrinsic(instr);
-
-         if (intrin->num_components == 1)
-            continue;
-
-         switch (intrin->intrinsic) {
-         case nir_intrinsic_read_invocation:
-         case nir_intrinsic_read_first_invocation:
-            lower_read_invocation_to_scalar(&b, intrin);
-            progress = true;
-            break;
-         default:
-            break;
-         }
-      }
-   }
-
-   if (progress) {
-      nir_metadata_preserve(impl, nir_metadata_block_index |
-                                  nir_metadata_dominance);
-   }
-   return progress;
-}
-
-bool
-nir_lower_read_invocation_to_scalar(nir_shader *shader)
-{
-   bool progress = false;
-
-   nir_foreach_function(function, shader) {
-      if (function->impl)
-         progress |= nir_lower_read_invocation_to_scalar_impl(function->impl);
-   }
-
-   return progress;
-}

src/compiler/nir/nir_lower_subgroups.c

@@ -0,0 +1,184 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include "nir.h"
+#include "nir_builder.h"
+
+/**
+ * \file nir_opt_intrinsics.c
+ */
+
+static nir_ssa_def *
+lower_read_invocation_to_scalar(nir_builder *b, nir_intrinsic_instr *intrin)
+{
+   /* This is safe to call on scalar things but it would be silly */
+   assert(intrin->dest.ssa.num_components > 1);
+
+   nir_ssa_def *value = nir_ssa_for_src(b, intrin->src[0],
+                                           intrin->num_components);
+   nir_ssa_def *reads[4];
+
+   for (unsigned i = 0; i < intrin->num_components; i++) {
+      nir_intrinsic_instr *chan_intrin =
+         nir_intrinsic_instr_create(b->shader, intrin->intrinsic);
+      nir_ssa_dest_init(&chan_intrin->instr, &chan_intrin->dest,
+                        1, intrin->dest.ssa.bit_size, NULL);
+      chan_intrin->num_components = 1;
+
+      /* value */
+      chan_intrin->src[0] = nir_src_for_ssa(nir_channel(b, value, i));
+      /* invocation */
+      if (intrin->intrinsic == nir_intrinsic_read_invocation)
+         nir_src_copy(&chan_intrin->src[1], &intrin->src[1], chan_intrin);
+
+      nir_builder_instr_insert(b, &chan_intrin->instr);
+
+      reads[i] = &chan_intrin->dest.ssa;
+   }
+
+   return nir_vec(b, reads, intrin->num_components);
+}
+
+static nir_ssa_def *
+high_subgroup_mask(nir_builder *b,
+                   nir_ssa_def *count,
+                   uint64_t base_mask)
+{
+   /* group_mask could probably be calculated more efficiently but we want to
+    * be sure not to shift by 64 if the subgroup size is 64 because the GLSL
+    * shift operator is undefined in that case. In any case if we were worried
+    * about efficency this should probably be done further down because the
+    * subgroup size is likely to be known at compile time.
+    */
+   nir_ssa_def *subgroup_size = nir_load_subgroup_size(b);
+   nir_ssa_def *all_bits = nir_imm_int64(b, ~0ull);
+   nir_ssa_def *shift = nir_isub(b, nir_imm_int(b, 64), subgroup_size);
+   nir_ssa_def *group_mask = nir_ushr(b, all_bits, shift);
+   nir_ssa_def *higher_bits = nir_ishl(b, nir_imm_int64(b, base_mask), count);
+
+   return nir_iand(b, higher_bits, group_mask);
+}
+
+static nir_ssa_def *
+lower_subgroups_intrin(nir_builder *b, nir_intrinsic_instr *intrin,
+                       const nir_lower_subgroups_options *options)
+{
+   switch (intrin->intrinsic) {
+   case nir_intrinsic_vote_any:
+   case nir_intrinsic_vote_all:
+      if (options->lower_vote_trivial)
+         return nir_ssa_for_src(b, intrin->src[0], 1);
+      break;
+
+   case nir_intrinsic_vote_eq:
+      if (options->lower_vote_trivial)
+         return nir_imm_int(b, NIR_TRUE);
+      break;
+
+   case nir_intrinsic_read_invocation:
+   case nir_intrinsic_read_first_invocation:
+      if (options->lower_to_scalar && intrin->num_components > 1)
+         return lower_read_invocation_to_scalar(b, intrin);
+      break;
+
+   case nir_intrinsic_load_subgroup_eq_mask:
+   case nir_intrinsic_load_subgroup_ge_mask:
+   case nir_intrinsic_load_subgroup_gt_mask:
+   case nir_intrinsic_load_subgroup_le_mask:
+   case nir_intrinsic_load_subgroup_lt_mask: {
+      if (!options->lower_subgroup_masks)
+         return NULL;
+
+      nir_ssa_def *count = nir_load_subgroup_invocation(b);
+
+      switch (intrin->intrinsic) {
+      case nir_intrinsic_load_subgroup_eq_mask:
+         return nir_ishl(b, nir_imm_int64(b, 1ull), count);
+      case nir_intrinsic_load_subgroup_ge_mask:
+         return high_subgroup_mask(b, count, ~0ull);
+      case nir_intrinsic_load_subgroup_gt_mask:
+         return high_subgroup_mask(b, count, ~1ull);
+      case nir_intrinsic_load_subgroup_le_mask:
+         return nir_inot(b, nir_ishl(b, nir_imm_int64(b, ~1ull), count));
+      case nir_intrinsic_load_subgroup_lt_mask:
+         return nir_inot(b, nir_ishl(b, nir_imm_int64(b, ~0ull), count));
+      default:
+         unreachable("you seriously can't tell this is unreachable?");
+      }
+      break;
+   }
+   default:
+      break;
+   }
+
+   return NULL;
+}
+
+static bool
+lower_subgroups_impl(nir_function_impl *impl,
+                     const nir_lower_subgroups_options *options)
+{
+   nir_builder b;
+   nir_builder_init(&b, impl);
+   bool progress = false;
+
+   nir_foreach_block(block, impl) {
+      nir_foreach_instr_safe(instr, block) {
+         if (instr->type != nir_instr_type_intrinsic)
+            continue;
+
+         nir_intrinsic_instr *intrin = nir_instr_as_intrinsic(instr);
+         b.cursor = nir_before_instr(instr);
+
+         nir_ssa_def *lower = lower_subgroups_intrin(&b, intrin, options);
+         if (!lower)
+            continue;
+
+         nir_ssa_def_rewrite_uses(&intrin->dest.ssa, nir_src_for_ssa(lower));
+         nir_instr_remove(instr);
+         progress = true;
+      }
+   }
+
+   return progress;
+}
+
+bool
+nir_lower_subgroups(nir_shader *shader,
+                    const nir_lower_subgroups_options *options)
+{
+   bool progress = false;
+
+   nir_foreach_function(function, shader) {
+      if (!function->impl)
+         continue;
+
+      if (lower_subgroups_impl(function->impl, options)) {
+         progress = true;
+         nir_metadata_preserve(function->impl, nir_metadata_block_index |
+                                               nir_metadata_dominance);
+      }
+   }
+
+   return progress;
+}

src/compiler/nir/nir_opt_intrinsics.c

@@ -28,26 +28,6 @@
  * \file nir_opt_intrinsics.c
  */
 
-static nir_ssa_def *
-high_subgroup_mask(nir_builder *b,
-                   nir_ssa_def *count,
-                   uint64_t base_mask)
-{
-   /* group_mask could probably be calculated more efficiently but we want to
-    * be sure not to shift by 64 if the subgroup size is 64 because the GLSL
-    * shift operator is undefined in that case. In any case if we were worried
-    * about efficency this should probably be done further down because the
-    * subgroup size is likely to be known at compile time.
-    */
-   nir_ssa_def *subgroup_size = nir_load_subgroup_size(b);
-   nir_ssa_def *all_bits = nir_imm_int64(b, ~0ull);
-   nir_ssa_def *shift = nir_isub(b, nir_imm_int(b, 64), subgroup_size);
-   nir_ssa_def *group_mask = nir_ushr(b, all_bits, shift);
-   nir_ssa_def *higher_bits = nir_ishl(b, nir_imm_int64(b, base_mask), count);
-
-   return nir_iand(b, higher_bits, group_mask);
-}
-
 static bool
 opt_intrinsics_impl(nir_function_impl *impl)
 {
@@ -66,22 +46,14 @@ opt_intrinsics_impl(nir_function_impl *impl)
 
          switch (intrin->intrinsic) {
          case nir_intrinsic_vote_any:
-         case nir_intrinsic_vote_all: {
+         case nir_intrinsic_vote_all:
-            nir_const_value *val = nir_src_as_const_value(intrin->src[0]);
+            if (nir_src_as_const_value(intrin->src[0]))
-            if (!val && !b.shader->options->lower_vote_trivial)
+               replacement = nir_ssa_for_src(&b, intrin->src[0], 1);
-               continue;
-
-            replacement = nir_ssa_for_src(&b, intrin->src[0], 1);
             break;
-         }
+         case nir_intrinsic_vote_eq:
-         case nir_intrinsic_vote_eq: {
+            if (nir_src_as_const_value(intrin->src[0]))
-            nir_const_value *val = nir_src_as_const_value(intrin->src[0]);
+               replacement = nir_imm_int(&b, NIR_TRUE);
-            if (!val && !b.shader->options->lower_vote_trivial)
-               continue;
-
-            replacement = nir_imm_int(&b, NIR_TRUE);
             break;
-         }
          case nir_intrinsic_ballot: {
             assert(b.shader->options->max_subgroup_size != 0);
             if (b.shader->options->max_subgroup_size > 32 ||
@@ -100,37 +72,6 @@ opt_intrinsics_impl(nir_function_impl *impl)
                                                  nir_imm_int(&b, 0));
             break;
          }
-         case nir_intrinsic_load_subgroup_eq_mask:
-         case nir_intrinsic_load_subgroup_ge_mask:
-         case nir_intrinsic_load_subgroup_gt_mask:
-         case nir_intrinsic_load_subgroup_le_mask:
-         case nir_intrinsic_load_subgroup_lt_mask: {
-            if (!b.shader->options->lower_subgroup_masks)
-               break;
-
-            nir_ssa_def *count = nir_load_subgroup_invocation(&b);
-
-            switch (intrin->intrinsic) {
-            case nir_intrinsic_load_subgroup_eq_mask:
-               replacement = nir_ishl(&b, nir_imm_int64(&b, 1ull), count);
-               break;
-            case nir_intrinsic_load_subgroup_ge_mask:
-               replacement = high_subgroup_mask(&b, count, ~0ull);
-               break;
-            case nir_intrinsic_load_subgroup_gt_mask:
-               replacement = high_subgroup_mask(&b, count, ~1ull);
-               break;
-            case nir_intrinsic_load_subgroup_le_mask:
-               replacement = nir_inot(&b, nir_ishl(&b, nir_imm_int64(&b, ~1ull), count));
-               break;
-            case nir_intrinsic_load_subgroup_lt_mask:
-               replacement = nir_inot(&b, nir_ishl(&b, nir_imm_int64(&b, ~0ull), count));
-               break;
-            default:
-               unreachable("you seriously can't tell this is unreachable?");
-            }
-            break;
-         }
          default:
             break;
          }

src/intel/compiler/brw_compiler.c

@@ -57,7 +57,6 @@ static const struct nir_shader_compiler_options scalar_nir_options = {
    .lower_unpack_snorm_4x8 = true,
    .lower_unpack_unorm_2x16 = true,
    .lower_unpack_unorm_4x8 = true,
-   .lower_subgroup_masks = true,
    .max_subgroup_size = 32,
    .max_unroll_iterations = 32,
 };
@@ -80,7 +79,6 @@ static const struct nir_shader_compiler_options vector_nir_options = {
    .lower_unpack_unorm_2x16 = true,
    .lower_extract_byte = true,
    .lower_extract_word = true,
-   .lower_vote_trivial = true,
    .max_unroll_iterations = 32,
 };
 
@@ -99,7 +97,6 @@ static const struct nir_shader_compiler_options vector_nir_options_gen6 = {
    .lower_unpack_unorm_2x16 = true,
    .lower_extract_byte = true,
    .lower_extract_word = true,
-   .lower_vote_trivial = true,
    .max_unroll_iterations = 32,
 };
 

src/intel/compiler/brw_nir.c

@@ -620,7 +620,6 @@ brw_preprocess_nir(const struct brw_compiler *compiler, nir_shader *nir)
 
    OPT(nir_lower_tex, &tex_options);
    OPT(nir_normalize_cubemap_coords);
-   OPT(nir_lower_read_invocation_to_scalar);
 
    OPT(nir_lower_global_vars_to_local);
 
@@ -637,6 +636,13 @@ brw_preprocess_nir(const struct brw_compiler *compiler, nir_shader *nir)
 
    OPT(nir_lower_system_values);
 
+   const nir_lower_subgroups_options subgroups_options = {
+      .lower_to_scalar = true,
+      .lower_subgroup_masks = true,
+      .lower_vote_trivial = !is_scalar,
+   };
+   OPT(nir_lower_subgroups, &subgroups_options);
+
    OPT(nir_lower_clip_cull_distance_arrays);
 
    nir_variable_mode indirect_mask = 0;