nir/group_loads: rename to nir_opt_group_loads

It's meant to be an optimization pass.

Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36100>

src/compiler/nir/meson.build

@@ -124,7 +124,7 @@ else
   'nir_gather_tcs_info.c',
   'nir_gather_types.c',
   'nir_gather_xfb_info.c',
-  'nir_group_loads.c',
+  'nir_opt_group_loads.c',
   'nir_gs_count_vertices.c',
   'nir_inline_uniforms.c',
   'nir_instr_set.c',

src/compiler/nir/nir.h

@@ -4885,8 +4885,8 @@ typedef enum {
    nir_group_same_resource_only,
 } nir_load_grouping;
 
-bool nir_group_loads(nir_shader *shader, nir_load_grouping grouping,
-                     unsigned max_distance);
+bool nir_opt_group_loads(nir_shader *shader, nir_load_grouping grouping,
+                         unsigned max_distance);
 
 bool nir_shrink_vec_array_vars(nir_shader *shader, nir_variable_mode modes);
 bool nir_split_array_vars(nir_shader *shader, nir_variable_mode modes);

src/compiler/nir/nir_opt_group_loads.c

@@ -454,8 +454,8 @@ process_block(nir_block *block, nir_load_grouping grouping,
  * in a group.
  */
 bool
-nir_group_loads(nir_shader *shader, nir_load_grouping grouping,
-                unsigned max_distance)
+nir_opt_group_loads(nir_shader *shader, nir_load_grouping grouping,
+                    unsigned max_distance)
 {
    /* Temporary space for instruction info. */
    struct util_dynarray infos_scratch;

src/gallium/drivers/i915/i915_screen.c

@@ -222,7 +222,7 @@ i915_optimize_nir(struct nir_shader *s)
    /* Group texture loads together to try to avoid hitting the
     * texture indirection phase limit.
     */
-   NIR_PASS(_, s, nir_group_loads, nir_group_all, ~0);
+   NIR_PASS(_, s, nir_opt_group_loads, nir_group_all, ~0);
 }
 
 static void

src/gallium/drivers/radeonsi/si_shader.c

@@ -1730,7 +1730,7 @@ static void run_late_optimization_and_lowering_passes(struct si_nir_shader_ctx *
    /* This helps LLVM form VMEM clauses and thus get more GPU cache hits.
     * 200 is tuned for Viewperf. It should be done last.
     */
-   NIR_PASS(_, nir, nir_group_loads, nir_group_same_resource_only, 200);
+   NIR_PASS(_, nir, nir_opt_group_loads, nir_group_same_resource_only, 200);
 }
 
 static void get_input_nir(struct si_shader *shader, struct si_nir_shader_ctx *ctx)