[CIR][NFC] Rename AtomicFence to AtomicFenceOp

clang/include/clang/CIR/Dialect/IR/CIROps.td

@@ -5493,7 +5493,7 @@ def CIR_AtomicClearOp : CIR_Op<"atomic.clear"> {
   }];
 }
 
-def CIR_AtomicFence : CIR_Op<"atomic.fence"> {
+def CIR_AtomicFenceOp : CIR_Op<"atomic.fence"> {
   let summary = "Atomic thread fence";
   let description = [{
     C/C++ Atomic thread fence synchronization primitive. Implements the builtin

clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp

@@ -60,9 +60,8 @@ static RValue emitBuiltinBitOp(CIRGenFunction &cgf, const CallExpr *e,
   return RValue::get(result);
 }
 
-static mlir::Value makeAtomicFenceValue(CIRGenFunction &cgf,
-                                        const CallExpr *expr,
-                                        cir::SyncScopeKind syncScope) {
+static void emitAtomicFenceOp(CIRGenFunction &cgf, const CallExpr *expr,
+                              cir::SyncScopeKind syncScope) {
   CIRGenBuilderTy &builder = cgf.getBuilder();
   mlir::Value orderingVal = cgf.emitScalarExpr(expr->getArg(0));
 
@@ -72,19 +71,19 @@ static mlir::Value makeAtomicFenceValue(CIRGenFunction &cgf,
     // TODO(cir): Emit code to switch on `orderingVal`,
     //            and creating the fence op for valid values.
     cgf.cgm.errorNYI("Variable atomic fence ordering");
-    return {};
+    return;
   }
 
   auto constOrderingAttr = constOrdering.getValueAttr<cir::IntAttr>();
   assert(constOrderingAttr && "Expected integer constant for ordering");
 
   auto ordering = static_cast<cir::MemOrder>(constOrderingAttr.getUInt());
 
-  cir::AtomicFence::create(
+  cir::AtomicFenceOp::create(
       builder, cgf.getLoc(expr->getSourceRange()), ordering,
       cir::SyncScopeKindAttr::get(&cgf.getMLIRContext(), syncScope));
 
-  return {};
+  return;
 }
 
 namespace {
@@ -1010,12 +1009,14 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID,
   case Builtin::BI__atomic_test_and_set:
   case Builtin::BI__atomic_clear:
     return errorBuiltinNYI(*this, e, builtinID);
-  case Builtin::BI__atomic_thread_fence:
-    return RValue::get(
-        makeAtomicFenceValue(*this, e, cir::SyncScopeKind::System));
-  case Builtin::BI__atomic_signal_fence:
-    return RValue::get(
-        makeAtomicFenceValue(*this, e, cir::SyncScopeKind::SingleThread));
+  case Builtin::BI__atomic_thread_fence: {
+    emitAtomicFenceOp(*this, e, cir::SyncScopeKind::System);
+    return RValue::get(nullptr);
+  }
+  case Builtin::BI__atomic_signal_fence: {
+    emitAtomicFenceOp(*this, e, cir::SyncScopeKind::SingleThread);
+    return RValue::get(nullptr);
+  }
   case Builtin::BI__c11_atomic_thread_fence:
   case Builtin::BI__c11_atomic_signal_fence:
   case Builtin::BI__scoped_atomic_thread_fence:

clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp

@@ -860,8 +860,8 @@ mlir::LogicalResult CIRToLLVMAtomicClearOpLowering::matchAndRewrite(
   return mlir::success();
 }
 
-mlir::LogicalResult CIRToLLVMAtomicFenceLowering::matchAndRewrite(
-    cir::AtomicFence op, OpAdaptor adaptor,
+mlir::LogicalResult CIRToLLVMAtomicFenceOpLowering::matchAndRewrite(
+    cir::AtomicFenceOp op, OpAdaptor adaptor,
     mlir::ConversionPatternRewriter &rewriter) const {
   mlir::LLVM::AtomicOrdering llvmOrder = getLLVMMemOrder(adaptor.getOrdering());