[llvm] [AMDGPU] Fix `HasFP32Denormals` check in FDIV32 lowering (PR #66212)

via llvm-commits llvm-commits at lists.llvm.org
Wed Sep 13 07:13:41 PDT 2023


llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu
            
<details>
<summary>Changes</summary>
Fixes SWDEV-403219
--
Full diff: https://github.com/llvm/llvm-project/pull/66212.diff

1 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/SIISelLowering.cpp (+1-1) 


<pre>
diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
index 777fe76df1151cd..998904bf08820c0 100644
--- a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
+++ b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -9589,7 +9589,7 @@ SDValue SITargetLowering::LowerFDIV32(SDValue Op, SelectionDAG &DAG) const {
   const SIMachineFunctionInfo *Info = MF.getInfo<SIMachineFunctionInfo>();
   const DenormalMode DenormMode = Info->getMode().FP32Denormals;
 
-  const bool HasFP32Denormals = DenormMode == DenormalMode::getIEEE();
+  const bool HasFP32Denormals = DenormMode != DenormalMode::getPreserveSign();
 
   if (!HasFP32Denormals) {
     // Note we can&#x27;t use the STRICT_FMA/STRICT_FMUL for the non-strict FDIV
</pre>
</details>


https://github.com/llvm/llvm-project/pull/66212


More information about the llvm-commits mailing list