radeonsi: compile non-GS middle parts of shaders immediately if enabled

Still disabled.

Only prologs & epilogs are compiled in draw calls, but each variant of those
is compiled only once per process.

VS is always compiled as hw VS.
TES is always compiled as hw VS.

LS and ES stages are always compiled on demand.

Reviewed-by: Nicolai Hähnle <[email protected]>

1 files changed, 46 insertions, 12 deletions

diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c
index 2f27da4f420..e3ba1f97252 100644
--- a/src/gallium/drivers/radeonsi/si_shader.c
+++ b/src/gallium/drivers/radeonsi/si_shader.c
@@ -4815,11 +4815,11 @@ static void si_init_shader_ctx(struct si_shader_context *ctx,
 	bld_base->op_actions[TGSI_OPCODE_MIN].intr_name = "llvm.minnum.f32";
 }
 
-static int si_compile_tgsi_shader(struct si_screen *sscreen,
-				  LLVMTargetMachineRef tm,
-				  struct si_shader *shader,
-				  bool is_monolithic,
-				  struct pipe_debug_callback *debug)
+int si_compile_tgsi_shader(struct si_screen *sscreen,
+			   LLVMTargetMachineRef tm,
+			   struct si_shader *shader,
+			   bool is_monolithic,
+			   struct pipe_debug_callback *debug)
 {
 	struct si_shader_selector *sel = shader->selector;
 	struct si_shader_context ctx;
@@ -5854,15 +5854,48 @@ int si_shader_create(struct si_screen *sscreen, LLVMTargetMachineRef tm,
 		     struct si_shader *shader,
 		     struct pipe_debug_callback *debug)
 {
+	struct si_shader *mainp = shader->selector->main_shader_part;
 	int r;
 
-	/* Compile TGSI. */
-	r = si_compile_tgsi_shader(sscreen, tm, shader,
-				   sscreen->use_monolithic_shaders, debug);
-	if (r)
-		return r;
+	/* LS and ES are always compiled on demand. */
+	if (!mainp ||
+	    (shader->selector->type == PIPE_SHADER_VERTEX &&
+	     (shader->key.vs.as_es || shader->key.vs.as_ls)) ||
+	    (shader->selector->type == PIPE_SHADER_TESS_EVAL &&
+	     shader->key.tes.as_es)) {
+		/* Monolithic shader (compiled as a whole, has many variants,
+		 * may take a long time to compile).
+		 */
+		r = si_compile_tgsi_shader(sscreen, tm, shader, true, debug);
+		if (r)
+			return r;
+	} else {
+		/* The shader consists of 2-3 parts:
+		 *
+		 * - the middle part is the user shader, it has 1 variant only
+		 *   and it was compiled during the creation of the shader
+		 *   selector
+		 * - the prolog part is inserted at the beginning
+		 * - the epilog part is inserted at the end
+		 *
+		 * The prolog and epilog have many (but simple) variants.
+		 */
 
-	if (!sscreen->use_monolithic_shaders) {
+		/* Copy the compiled TGSI shader data over. */
+		shader->is_binary_shared = true;
+		shader->binary = mainp->binary;
+		shader->config = mainp->config;
+		shader->num_input_sgprs = mainp->num_input_sgprs;
+		shader->num_input_vgprs = mainp->num_input_vgprs;
+		shader->face_vgpr_index = mainp->face_vgpr_index;
+		memcpy(shader->vs_output_param_offset,
+		       mainp->vs_output_param_offset,
+		       sizeof(mainp->vs_output_param_offset));
+		shader->uses_instanceid = mainp->uses_instanceid;
+		shader->nr_pos_exports = mainp->nr_pos_exports;
+		shader->nr_param_exports = mainp->nr_param_exports;
+
+		/* Select prologs and/or epilogs. */
 		switch (shader->selector->type) {
 		case PIPE_SHADER_VERTEX:
 			if (!si_shader_select_vs_parts(sscreen, tm, shader, debug))
@@ -5927,5 +5960,6 @@ void si_shader_destroy(struct si_shader *shader)
 
 	r600_resource_reference(&shader->bo, NULL);
 
-	radeon_shader_binary_clean(&shader->binary);
+	if (!shader->is_binary_shared)
+		radeon_shader_binary_clean(&shader->binary);
 }