[llvm] [WIP][AMDGPU] Support the RF inline asm constraint (PR #195123)

Jun Wang via llvm-commits llvm-commits at lists.llvm.org
Fri Jul 17 15:50:23 PDT 2026


================
@@ -10502,8 +10502,15 @@ static bool prepareDAGLevelOperands(ConstraintDecisionInfo &Info,
         assert((OpInfo.isIndirect ||
                 OpInfo.ConstraintType != TargetLowering::C_Memory) &&
                "Operand must be indirect to be a mem!");
+        unsigned AddrSpace = 0;
+        const Value *Ptr = OpInfo.CallOperandVal;
+        if (Ptr) {
+          if (auto *Pt = dyn_cast<PointerType>(Ptr->getType()))
+            AddrSpace = Pt->getAddressSpace();
+        }
+
         assert(InOperandVal.getValueType() ==
-                   TLI.getPointerTy(DAG.getDataLayout()) &&
+                   TLI.getPointerTy(DAG.getDataLayout(), AddrSpace) &&
                "Memory operands expect pointer values");
----------------
jwanggit86 wrote:

Fixed as suggested.

https://github.com/llvm/llvm-project/pull/195123


More information about the llvm-commits mailing list