[Mesa-stable] [PATCH] ac/nir: mark some texture intrinsics as convergent

Rhys Perry pendingchaos02 at gmail.com
Thu May 30 19:52:49 UTC 2019


Otherwise LLVM can sink them and their texture coordinate calculations
into divergent branches.

v2: simplify the conditions on which the intrinsic is marked as convergent

Cc: <mesa-stable at lists.freedesktop.org>
Signed-off-by: Rhys Perry <pendingchaos02 at gmail.com>
Reviewed-By: Bas Nieuwenhuizen <bas at basnieuwenhuizen.nl>
---
 src/amd/common/ac_nir_to_llvm.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c
index 265e3b636c4..b1a191ac24c 100644
--- a/src/amd/common/ac_nir_to_llvm.c
+++ b/src/amd/common/ac_nir_to_llvm.c
@@ -1394,6 +1394,18 @@ static LLVMValueRef build_tex_intrinsic(struct ac_nir_context *ctx,
 	}
 
 	args->attributes = AC_FUNC_ATTR_READNONE;
+	/* Prevent texture instructions with implicit derivatives from being
+	 * sinked into branches. */
+	switch (instr->op) {
+	case nir_texop_tex:
+	case nir_texop_txb:
+	case nir_texop_lod:
+		args->attributes |= AC_FUNC_ATTR_CONVERGENT;
+		break;
+	default:
+		break;
+	}
+
 	return ac_build_image_opcode(&ctx->ac, args);
 }
 
-- 
2.21.0



More information about the mesa-stable mailing list