From: Connor Abbott <cwabbott0@gmail.com>
Date: Thu, 18 Oct 2018 13:39:13 +0000 (+0200)
Subject: ac: Fix loading a dvec3 from an SSBO
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=27fe3f5b5a18c7ae404ac933dae6a9adcb2d5f7b;p=mesa.git

ac: Fix loading a dvec3 from an SSBO

The comment was wrong, since the loop above casts to a type with the
correct bitsize already.

Fixes: 7e7ee82698247d8f93fe37775b99f4838b0247dd ("ac: add support for 16bit buffer loads")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
---

diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c
index 402cf2d6655..ee75e2890dd 100644
--- a/src/amd/common/ac_nir_to_llvm.c
+++ b/src/amd/common/ac_nir_to_llvm.c
@@ -1685,8 +1685,8 @@ static LLVMValueRef visit_load_buffer(struct ac_nir_context *ctx,
 		};
 
 		if (num_bytes > 16 && num_components == 3) {
-			/* we end up with a v4f32 and v2f32 but shuffle fails on that */
-			results[1] = ac_build_expand_to_vec4(&ctx->ac, results[1], 2);
+			/* we end up with a v2i64 and i64 but shuffle fails on that */
+			results[1] = ac_build_expand(&ctx->ac, results[1], 1, 2);
 		}
 
 		LLVMValueRef swizzle = LLVMConstVector(masks, num_components);