[llvm] Fix attributes of llvm.amdgcn.make.buffer.rsrc (PR #68609)
Krzysztof Drewniak via llvm-commits
llvm-commits at lists.llvm.org
Mon Oct 9 09:33:58 PDT 2023
https://github.com/krzysz00 created https://github.com/llvm/llvm-project/pull/68609
Per discussion on https://github.com/llvm/llvm-project/pull/68453 , the make_buffer_rsrc intrinsic was incorrectly marked noclobber. This commit fixes the issue.
>From 3a3844a40c18e6fc717cb29e1448c3b3bafd50ec Mon Sep 17 00:00:00 2001
From: Krzysztof Drewniak <Krzysztof.Drewniak at amd.com>
Date: Mon, 9 Oct 2023 16:31:49 +0000
Subject: [PATCH] Fix attributes of llvm.amdgcn.make.buffer.rsrc
Per discussion on https://github.com/llvm/llvm-project/pull/68453 ,
the make_buffer_rsrc intrinsic was incorrectly marked noclobber. This
commit fixes the issue.
---
llvm/include/llvm/IR/IntrinsicsAMDGPU.td | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
index 4f42462f655e260..0df66e1ffc519f2 100644
--- a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
+++ b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
@@ -999,7 +999,7 @@ def int_amdgcn_make_buffer_rsrc : DefaultAttrsIntrinsic <
llvm_i32_ty, // NumRecords / extent
llvm_i32_ty], // flags
// Attributes lifted from ptrmask + some extra argument attributes.
- [IntrNoMem, NoCapture<ArgIndex<0>>, ReadNone<ArgIndex<0>>,
+ [IntrNoMem, ReadNone<ArgIndex<0>>,
IntrSpeculatable, IntrWillReturn]>;
defset list<AMDGPURsrcIntrinsic> AMDGPUBufferIntrinsics = {
@@ -1928,7 +1928,7 @@ def int_amdgcn_inverse_ballot :
[IntrNoMem, IntrWillReturn, IntrNoCallback, IntrNoFree]>;
class AMDGPUWaveReduce<LLVMType data_ty = llvm_anyint_ty> : Intrinsic<
- [data_ty],
+ [data_ty],
[
LLVMMatchType<0>, // llvm value to reduce (SGPR/VGPR)
llvm_i32_ty // Reduction Strategy Switch for lowering ( 0: Default,
More information about the llvm-commits
mailing list