[llvm-branch-commits] [llvm] [AMDGPU] Return two MMOs for load-to-lds and store-from-lds intrinsics (PR #175845)

[Kevin Choi via llvm-branch-commits]

================
@@ -1626,35 +1642,64 @@ void SITargetLowering::getTgtMemIntrinsic(SmallVectorImpl<IntrinsicInfo> &Infos,
   case Intrinsic::amdgcn_cluster_load_async_to_lds_b32:
   case Intrinsic::amdgcn_cluster_load_async_to_lds_b64:
   case Intrinsic::amdgcn_cluster_load_async_to_lds_b128: {
+    // Entry 0: Load from source (global/flat).
     Info.opc = ISD::INTRINSIC_VOID;
     Info.memVT = EVT::getIntegerVT(CI.getContext(), getIntrMemWidth(IntrID));
-    Info.ptrVal = CI.getArgOperand(1);
-    Info.flags |= MachineMemOperand::MOLoad | MachineMemOperand::MOStore;
+    Info.ptrVal = CI.getArgOperand(0); // Global pointer
+    Info.offset = cast<ConstantInt>(CI.getArgOperand(2))->getSExtValue();
+    Info.flags |= MachineMemOperand::MOLoad;
----------------
choikwa wrote:

Info.flags &= ~MachineMemOperand::MOStore;  

or is it unnecessary?

https://github.com/llvm/llvm-project/pull/175845