Mesa (master): ac/nir: init full exec mask for merged shaders.

[Bas Nieuwenhuizen]

Module: Mesa
Branch: master
Commit: 1dda214d9c8f22081448cca64d2ad8876f2f1d94
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=1dda214d9c8f22081448cca64d2ad8876f2f1d94

Author: Dave Airlie <airlied at redhat.com>
Date:   Thu Oct 19 05:29:02 2017 +0100

ac/nir: init full exec mask for merged shaders.

Signed-off-by: Bas Nieuwenhuizen <bas at basnieuwenhuizen.nl>
Reviewed-by: Dave Airlie <airlied at redhat.com>

---

 src/amd/common/ac_llvm_build.c  | 8 ++++++++
 src/amd/common/ac_llvm_build.h  | 1 +
 src/amd/common/ac_nir_to_llvm.c | 3 +++
 3 files changed, 12 insertions(+)

diff --git a/src/amd/common/ac_llvm_build.c b/src/amd/common/ac_llvm_build.c
index 949f181aac..e5cd23e025 100644
--- a/src/amd/common/ac_llvm_build.c
+++ b/src/amd/common/ac_llvm_build.c
@@ -1734,3 +1734,11 @@ void ac_optimize_vs_outputs(struct ac_llvm_context *ctx,
 		*num_param_exports = exports.num;
 	}
 }
+
+void ac_init_exec_full_mask(struct ac_llvm_context *ctx)
+{
+	LLVMValueRef full_mask = LLVMConstInt(ctx->i64, ~0ull, 0);
+	ac_build_intrinsic(ctx,
+			   "llvm.amdgcn.init.exec", ctx->voidt,
+			   &full_mask, 1, AC_FUNC_ATTR_CONVERGENT);
+}
diff --git a/src/amd/common/ac_llvm_build.h b/src/amd/common/ac_llvm_build.h
index f0b5875b42..aa2a2899ab 100644
--- a/src/amd/common/ac_llvm_build.h
+++ b/src/amd/common/ac_llvm_build.h
@@ -281,6 +281,7 @@ void ac_optimize_vs_outputs(struct ac_llvm_context *ac,
 			    uint8_t *vs_output_param_offset,
 			    uint32_t num_outputs,
 			    uint8_t *num_param_exports);
+void ac_init_exec_full_mask(struct ac_llvm_context *ctx);
 #ifdef __cplusplus
 }
 #endif
diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c
index 44de834bd8..0e1c65ae81 100644
--- a/src/amd/common/ac_nir_to_llvm.c
+++ b/src/amd/common/ac_nir_to_llvm.c
@@ -6488,6 +6488,9 @@ LLVMModuleRef ac_translate_nir_to_llvm(LLVMTargetMachineRef tm,
 	ctx.abi.load_ssbo = radv_load_ssbo;
 	ctx.abi.load_sampler_desc = radv_get_sampler_desc;
 
+	if (shader_count >= 2)
+		ac_init_exec_full_mask(&ctx.ac);
+
 	if (ctx.ac.chip_class == GFX9 &&
 	    shaders[shader_count - 1]->stage == MESA_SHADER_TESS_CTRL)
 		ac_nir_fixup_ls_hs_input_vgprs(&ctx);