[Mesa-dev] [PATCH 08/23] radeonsi: only use LLVMBuildLoad once when updating color outputs at the end

[Marek Olšák]

From: Marek Olšák <marek.olsak at amd.com>

without LLVMBuildStore.

So:
- do LLVMBuildLoad
- update the values as necessary
- export
---
 src/gallium/drivers/radeonsi/si_shader.c | 67 ++++++++++----------------------
 1 file changed, 20 insertions(+), 47 deletions(-)

diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c
index 73a34ac..13e5140 100644
--- a/src/gallium/drivers/radeonsi/si_shader.c
+++ b/src/gallium/drivers/radeonsi/si_shader.c
@@ -1334,24 +1334,8 @@ static void si_llvm_init_export_args(struct lp_build_tgsi_context *bld_base,
 		memcpy(&args[5], values, sizeof(values[0]) * 4);
 }
 
-/* Load from output pointers and initialize arguments for the shader export intrinsic */
-static void si_llvm_init_export_args_load(struct lp_build_tgsi_context *bld_base,
-					  LLVMValueRef *out_ptr,
-					  unsigned target,
-					  LLVMValueRef *args)
-{
-	struct gallivm_state *gallivm = bld_base->base.gallivm;
-	LLVMValueRef values[4];
-	int i;
-
-	for (i = 0; i < 4; i++)
-		values[i] = LLVMBuildLoad(gallivm->builder, out_ptr[i], "");
-
-	si_llvm_init_export_args(bld_base, values, target, args);
-}
-
 static void si_alpha_test(struct lp_build_tgsi_context *bld_base,
-			  LLVMValueRef alpha_ptr)
+			  LLVMValueRef alpha)
 {
 	struct si_shader_context *si_shader_ctx = si_shader_context(bld_base);
 	struct gallivm_state *gallivm = bld_base->base.gallivm;
@@ -1363,8 +1347,7 @@ static void si_alpha_test(struct lp_build_tgsi_context *bld_base,
 		LLVMValueRef alpha_pass =
 			lp_build_cmp(&bld_base->base,
 				     si_shader_ctx->shader->key.ps.alpha_func,
-				     LLVMBuildLoad(gallivm->builder, alpha_ptr, ""),
-				     alpha_ref);
+				     alpha, alpha_ref);
 		LLVMValueRef arg =
 			lp_build_select(&bld_base->base,
 					alpha_pass,
@@ -1383,12 +1366,12 @@ static void si_alpha_test(struct lp_build_tgsi_context *bld_base,
 	}
 }
 
-static void si_scale_alpha_by_sample_mask(struct lp_build_tgsi_context *bld_base,
-					  LLVMValueRef alpha_ptr)
+static LLVMValueRef si_scale_alpha_by_sample_mask(struct lp_build_tgsi_context *bld_base,
+						  LLVMValueRef alpha)
 {
 	struct si_shader_context *si_shader_ctx = si_shader_context(bld_base);
 	struct gallivm_state *gallivm = bld_base->base.gallivm;
-	LLVMValueRef coverage, alpha;
+	LLVMValueRef coverage;
 
 	/* alpha = alpha * popcount(coverage) / SI_NUM_SMOOTH_AA_SAMPLES */
 	coverage = LLVMGetParam(si_shader_ctx->radeon_bld.main_fn,
@@ -1406,9 +1389,7 @@ static void si_scale_alpha_by_sample_mask(struct lp_build_tgsi_context *bld_base
 				 lp_build_const_float(gallivm,
 					1.0 / SI_NUM_SMOOTH_AA_SAMPLES), "");
 
-	alpha = LLVMBuildLoad(gallivm->builder, alpha_ptr, "");
-	alpha = LLVMBuildFMul(gallivm->builder, alpha, coverage, "");
-	LLVMBuildStore(gallivm->builder, alpha, alpha_ptr);
+	return LLVMBuildFMul(gallivm->builder, alpha, coverage, "");
 }
 
 static void si_llvm_emit_clipvertex(struct lp_build_tgsi_context * bld_base,
@@ -2196,8 +2177,8 @@ static void si_llvm_emit_fs_epilogue(struct lp_build_tgsi_context * bld_base)
 	for (i = 0; i < info->num_outputs; i++) {
 		unsigned semantic_name = info->output_semantic_name[i];
 		unsigned semantic_index = info->output_semantic_index[i];
-		unsigned target;
-		LLVMValueRef alpha_ptr;
+		unsigned target, j;
+		LLVMValueRef color[4] = {};
 
 		/* Select the correct target */
 		switch (semantic_name) {
@@ -2215,29 +2196,24 @@ static void si_llvm_emit_fs_epilogue(struct lp_build_tgsi_context * bld_base)
 			continue;
 		case TGSI_SEMANTIC_COLOR:
 			target = V_008DFC_SQ_EXP_MRT + semantic_index;
-			alpha_ptr = si_shader_ctx->radeon_bld.soa.outputs[i][3];
 
-			if (si_shader_ctx->shader->key.ps.clamp_color) {
-				for (int j = 0; j < 4; j++) {
-					LLVMValueRef ptr = si_shader_ctx->radeon_bld.soa.outputs[i][j];
-					LLVMValueRef result = LLVMBuildLoad(builder, ptr, "");
+			for (j = 0; j < 4; j++)
+				color[j] = LLVMBuildLoad(builder,
+							 si_shader_ctx->radeon_bld.soa.outputs[i][j], "");
 
-					result = radeon_llvm_saturate(bld_base, result);
-					LLVMBuildStore(builder, result, ptr);
-				}
-			}
+			if (si_shader_ctx->shader->key.ps.clamp_color)
+				for (j = 0; j < 4; j++)
+					color[j] = radeon_llvm_saturate(bld_base, color[j]);
 
 			if (si_shader_ctx->shader->key.ps.alpha_to_one)
-				LLVMBuildStore(base->gallivm->builder,
-					       base->one, alpha_ptr);
+				color[3] = base->one;
 
 			if (semantic_index == 0 &&
 			    si_shader_ctx->shader->key.ps.alpha_func != PIPE_FUNC_ALWAYS)
-				si_alpha_test(bld_base, alpha_ptr);
+				si_alpha_test(bld_base, color[3]);
 
 			if (si_shader_ctx->shader->key.ps.poly_line_smoothing)
-				si_scale_alpha_by_sample_mask(bld_base, alpha_ptr);
-
+				color[3] = si_scale_alpha_by_sample_mask(bld_base, color[3]);
 			break;
 		default:
 			fprintf(stderr,
@@ -2250,18 +2226,15 @@ static void si_llvm_emit_fs_epilogue(struct lp_build_tgsi_context * bld_base)
 		if (semantic_index == 0 &&
 		    si_shader_ctx->shader->key.ps.last_cbuf > 0) {
 			for (int c = 1; c <= si_shader_ctx->shader->key.ps.last_cbuf; c++) {
-				si_llvm_init_export_args_load(bld_base,
-							      si_shader_ctx->radeon_bld.soa.outputs[i],
-							      V_008DFC_SQ_EXP_MRT + c, args);
+				si_llvm_init_export_args(bld_base, color,
+							 V_008DFC_SQ_EXP_MRT + c, args);
 				lp_build_intrinsic(base->gallivm->builder, "llvm.SI.export",
 						   LLVMVoidTypeInContext(base->gallivm->context),
 						   args, 9, 0);
 			}
 		}
 
-		si_llvm_init_export_args_load(bld_base,
-					      si_shader_ctx->radeon_bld.soa.outputs[i],
-					      target, args);
+		si_llvm_init_export_args(bld_base, color, target, args);
 		if (last_color_export == i) {
 			args[1] = uint->one; /* whether the EXEC mask is valid */
 			args[2] = uint->one; /* DONE bit */
-- 
2.1.4