From: Nicolai Hähnle <nicolai.haehnle@amd.com>
Date: Fri, 17 Nov 2017 12:40:18 +0000 (+0100)
Subject: radeonsi/gfx10: distinguish between merged shaders and multi-part shaders
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=e86256c51229c135bc4861018c16b9a8f1c255a6;p=mesa.git

radeonsi/gfx10: distinguish between merged shaders and multi-part shaders

Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
---

diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c
index fdb4346f7fd..b6410c62448 100644
--- a/src/gallium/drivers/radeonsi/si_shader.c
+++ b/src/gallium/drivers/radeonsi/si_shader.c
@@ -83,7 +83,8 @@ static bool llvm_type_is_64bit(struct si_shader_context *ctx,
 	return false;
 }
 
-static bool is_merged_shader(struct si_shader_context *ctx)
+/** Whether the shader runs as a combination of multiple API shaders */
+static bool is_multi_part_shader(struct si_shader_context *ctx)
 {
 	if (ctx->screen->info.chip_class <= GFX8)
 		return false;
@@ -94,6 +95,12 @@ static bool is_merged_shader(struct si_shader_context *ctx)
 	       ctx->type == PIPE_SHADER_GEOMETRY;
 }
 
+/** Whether the shader runs on a merged HW stage (LSHS or ESGS) */
+static bool is_merged_shader(struct si_shader_context *ctx)
+{
+	return ctx->shader->key.as_ngg || is_multi_part_shader(ctx);
+}
+
 void si_init_function_info(struct si_function_info *fninfo)
 {
 	fninfo->num_params = 0;
@@ -6543,7 +6550,7 @@ static void si_build_wrapper_function(struct si_shader_context *ctx,
 
 		/* Merged shaders are executed conditionally depending
 		 * on the number of enabled threads passed in the input SGPRs. */
-		if (is_merged_shader(ctx) && part == 0) {
+		if (is_multi_part_shader(ctx) && part == 0) {
 			LLVMValueRef ena, count = initial[3];
 
 			count = LLVMBuildAnd(builder, count,
@@ -6605,7 +6612,7 @@ static void si_build_wrapper_function(struct si_shader_context *ctx,
 
 		ret = ac_build_call(&ctx->ac, parts[part], in, num_params);
 
-		if (is_merged_shader(ctx) &&
+		if (is_multi_part_shader(ctx) &&
 		    part + 1 == next_shader_first_part) {
 			lp_build_endif(&if_state);