i965: Add lowering pass for splitting textureGatherOffsets

Rewrites textureGatherOffsets(s, p, offsets) into

   gvec4(
      textureGatherOffset(s, p, offsets[0]).w,
      textureGatherOffset(s, p, offsets[1]).w,
      textureGatherOffset(s, p, offsets[2]).w,
      textureGatherOffset(s, p, offsets[3]).w
      )

V2: Use ir_builder to be slightly clearer.

Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

src/mesa/drivers/dri/i965/Makefile.sources

@@ -68,6 +68,7 @@ i965_FILES = \
 	brw_gs_surface_state.c \
 	brw_interpolation_map.c \
 	brw_lower_texture_gradients.cpp \
+	brw_lower_offset_array.cpp \
 	brw_misc_state.c \
 	brw_object_purgeable.c \
 	brw_performance_monitor.c \

src/mesa/drivers/dri/i965/brw_context.h

@@ -1686,6 +1686,7 @@ brw_program_reloc(struct brw_context *brw, uint32_t state_offset,
 bool brw_do_cubemap_normalize(struct exec_list *instructions);
 bool brw_lower_texture_gradients(struct brw_context *brw,
                                  struct exec_list *instructions);
+bool brw_do_lower_offset_arrays(struct exec_list *instructions);
 
 struct opcode_desc {
     char    *name;

src/mesa/drivers/dri/i965/brw_lower_offset_array.cpp

@@ -0,0 +1,89 @@
+/*
+ * Copyright © 2013 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.
+ */
+
+/**
+ * \file brw_lower_offset_array.cpp
+ *
+ * IR lower pass to decompose ir_texture ir_tg4 with an array of offsets
+ * into four ir_tg4s with a single ivec2 offset, select the .w component of each,
+ * and return those four values packed into a gvec4.
+ *
+ * \author Chris Forbes <chrisf@ijw.co.nz>
+ */
+
+#include "glsl/glsl_types.h"
+#include "glsl/ir.h"
+#include "glsl/ir_builder.h"
+
+using namespace ir_builder;
+
+class brw_lower_offset_array_visitor : public ir_hierarchical_visitor {
+public:
+   brw_lower_offset_array_visitor()
+   {
+      progress = false;
+   }
+
+   ir_visitor_status visit_leave(ir_texture *ir);
+
+   bool progress;
+};
+
+ir_visitor_status
+brw_lower_offset_array_visitor::visit_leave(ir_texture *ir)
+{
+   if (ir->op != ir_tg4 || !ir->offset || !ir->offset->type->is_array())
+      return visit_continue;
+
+   void *mem_ctx = ralloc_parent(ir);
+
+   ir_variable *var = new (mem_ctx) ir_variable(ir->type, "result", ir_var_auto);
+   base_ir->insert_before(var);
+
+   for (int i = 0; i < 4; i++) {
+      ir_texture *tex = ir->clone(mem_ctx, NULL);
+      tex->offset = new (mem_ctx) ir_dereference_array(tex->offset,
+            new (mem_ctx) ir_constant(i));
+
+      base_ir->insert_before(assign(var, swizzle_w(tex), 1 << i));
+   }
+
+   base_ir->replace_with(new (mem_ctx) ir_dereference_variable(var));
+
+   progress = true;
+   return visit_continue;
+}
+
+extern "C" {
+
+bool
+brw_do_lower_offset_arrays(exec_list *instructions)
+{
+   brw_lower_offset_array_visitor v;
+
+   visit_list_elements(&v, instructions);
+
+   return v.progress;
+}
+
+}

src/mesa/drivers/dri/i965/brw_shader.cpp

@@ -172,6 +172,7 @@ brw_link_shader(struct gl_context *ctx, struct gl_shader_program *shProg)
       do_vec_index_to_cond_assign(shader->ir);
       lower_vector_insert(shader->ir, true);
       brw_do_cubemap_normalize(shader->ir);
+      brw_do_lower_offset_arrays(shader->ir);
       lower_noise(shader->ir);
       lower_quadop_vector(shader->ir, false);