[COFF, ARM64] Implement InterlockedOr*_* builtins

[mgrang]

Summary

This is fourth in a series of patches to move intrinsic defintions out of intrin.h.

[efriedma]

LGTM

include/clang/Basic/BuiltinsAArch64.def

 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange16_rel, "ssD*ss",         "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_acq,   "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_nf,    "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_rel,   "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_acq, "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_nf,  "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_rel, "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 
+TARGET_HEADER_BUILTIN(_InterlockedOr8_acq,  "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr8_nf,   "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr8_rel,  "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_acq, "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_nf,  "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_rel, "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_acq,   "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_nf,    "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_rel,   "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_acq, "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_nf,  "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_rel, "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+
 TARGET_HEADER_BUILTIN(_ReadWriteBarrier, "v", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(__getReg, "ULLii", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_ReadStatusReg,  "ii",  "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_WriteStatusReg, "vii", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_AddressOfReturnAddress, "v*", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 
 #undef BUILTIN
 #undef LANGBUILTIN
 #undef TARGET_HEADER_BUILTIN

include/clang/Basic/BuiltinsARM.def

 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange16_rel, "ssD*ss",         "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_acq,   "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_nf,    "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange_rel,   "LiLiD*LiLi",     "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_acq, "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_nf,  "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 TARGET_HEADER_BUILTIN(_InterlockedCompareExchange64_rel, "LLiLLiD*LLiLLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
 
+TARGET_HEADER_BUILTIN(_InterlockedOr8_acq,  "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr8_nf,   "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr8_rel,  "ccD*c",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_acq, "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_nf,  "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr16_rel, "ssD*s",       "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_acq,   "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_nf,    "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr_rel,   "LiLiD*Li",    "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_acq, "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_nf,  "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+TARGET_HEADER_BUILTIN(_InterlockedOr64_rel, "LLiLLiD*LLi", "nh", "intrin.h", ALL_MS_LANGUAGES, "")
+
 #undef BUILTIN
 #undef LANGBUILTIN
 #undef TARGET_HEADER_BUILTIN

lib/CodeGen/CGBuiltin.cpp

enum class CodeGenFunction::MSVCIntrin {
   _InterlockedExchangeAdd_rel,
   _InterlockedExchangeAdd_nf,
   _InterlockedExchange_acq,
   _InterlockedExchange_rel,
   _InterlockedExchange_nf,
   _InterlockedCompareExchange_acq,
   _InterlockedCompareExchange_rel,
   _InterlockedCompareExchange_nf,
+  _InterlockedOr_acq,
+  _InterlockedOr_rel,
+  _InterlockedOr_nf,
   __fastfail,
 };
 
 Value *CodeGenFunction::EmitMSVCBuiltinExpr(MSVCIntrin BuiltinID,
                                             const CallExpr *E) {
   switch (BuiltinID) {
   case MSVCIntrin::_BitScanForward:
   case MSVCIntrin::_BitScanReverse: {
case MSVCIntrin::_InterlockedExchange_nf:
     return MakeBinaryAtomicValue(*this, AtomicRMWInst::Xchg, E,
                                  AtomicOrdering::Monotonic);
   case MSVCIntrin::_InterlockedCompareExchange_acq:
     return EmitAtomicCmpXchgValue(*this, E, AtomicOrdering::Acquire);
   case MSVCIntrin::_InterlockedCompareExchange_rel:
     return EmitAtomicCmpXchgValue(*this, E, AtomicOrdering::Release);
   case MSVCIntrin::_InterlockedCompareExchange_nf:
     return EmitAtomicCmpXchgValue(*this, E, AtomicOrdering::Monotonic);
+  case MSVCIntrin::_InterlockedOr_acq:
+    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Or, E,
+                                 AtomicOrdering::Acquire);
+  case MSVCIntrin::_InterlockedOr_rel:
+    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Or, E,
+                                 AtomicOrdering::Release);
+  case MSVCIntrin::_InterlockedOr_nf:
+    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Or, E,
+                                 AtomicOrdering::Monotonic);
 
   case MSVCIntrin::_InterlockedDecrement: {
     llvm::Type *IntTy = ConvertType(E->getType());
     AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
       AtomicRMWInst::Sub,
       EmitScalarExpr(E->getArg(0)),
       ConstantInt::get(IntTy, 1),
       llvm::AtomicOrdering::SequentiallyConsistent);
Value *CodeGenFunction::EmitARMBuiltinExpr(unsigned BuiltinID,
   case ARM::BI_InterlockedCompareExchange_rel:
   case ARM::BI_InterlockedCompareExchange64_rel:
     return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_rel, E);
   case ARM::BI_InterlockedCompareExchange8_nf:
   case ARM::BI_InterlockedCompareExchange16_nf:
   case ARM::BI_InterlockedCompareExchange_nf:
   case ARM::BI_InterlockedCompareExchange64_nf:
     return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_nf, E);
+  case ARM::BI_InterlockedOr8_acq:
+  case ARM::BI_InterlockedOr16_acq:
+  case ARM::BI_InterlockedOr_acq:
+  case ARM::BI_InterlockedOr64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_acq, E);
+  case ARM::BI_InterlockedOr8_rel:
+  case ARM::BI_InterlockedOr16_rel:
+  case ARM::BI_InterlockedOr_rel:
+  case ARM::BI_InterlockedOr64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_rel, E);
+  case ARM::BI_InterlockedOr8_nf:
+  case ARM::BI_InterlockedOr16_nf:
+  case ARM::BI_InterlockedOr_nf:
+  case ARM::BI_InterlockedOr64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_nf, E);
   }
 
   // Get the last argument, which specifies the vector type.
   assert(HasExtraArg);
   llvm::APSInt Result;
   const Expr *Arg = E->getArg(E->getNumArgs()-1);
   if (!Arg->isIntegerConstantExpr(Result, getContext()))
     return nullptr;

   case AArch64::BI_InterlockedCompareExchange_rel:
   case AArch64::BI_InterlockedCompareExchange64_rel:
     return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_rel, E);
   case AArch64::BI_InterlockedCompareExchange8_nf:
   case AArch64::BI_InterlockedCompareExchange16_nf:
   case AArch64::BI_InterlockedCompareExchange_nf:
   case AArch64::BI_InterlockedCompareExchange64_nf:
     return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_nf, E);
+  case AArch64::BI_InterlockedOr8_acq:
+  case AArch64::BI_InterlockedOr16_acq:
+  case AArch64::BI_InterlockedOr_acq:
+  case AArch64::BI_InterlockedOr64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_acq, E);
+  case AArch64::BI_InterlockedOr8_rel:
+  case AArch64::BI_InterlockedOr16_rel:
+  case AArch64::BI_InterlockedOr_rel:
+  case AArch64::BI_InterlockedOr64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_rel, E);
+  case AArch64::BI_InterlockedOr8_nf:
+  case AArch64::BI_InterlockedOr16_nf:
+  case AArch64::BI_InterlockedOr_nf:
+  case AArch64::BI_InterlockedOr64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_nf, E);
 
   case AArch64::BI_InterlockedAdd: {
     Value *Arg0 = EmitScalarExpr(E->getArg(0));
     Value *Arg1 = EmitScalarExpr(E->getArg(1));
     AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
       AtomicRMWInst::Add, Arg0, Arg1,
       llvm::AtomicOrdering::SequentiallyConsistent);
     return Builder.CreateAdd(RMWI, Arg1);

lib/Headers/intrin.h

unsigned char _interlockedbittestandreset_nf(long volatile *_BitBase,
                                              long _BitPos);
 unsigned char _interlockedbittestandreset_rel(long volatile *_BitBase,
                                               long _BitPos);
 #endif
 /*----------------------------------------------------------------------------*\
 |* Interlocked Or
 \*----------------------------------------------------------------------------*/
 #if defined(__arm__) || defined(__aarch64__)
-static __inline__ char __DEFAULT_FN_ATTRS
-_InterlockedOr8_acq(char volatile *_Value, char _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_ACQUIRE);
-}
-static __inline__ char __DEFAULT_FN_ATTRS
-_InterlockedOr8_nf(char volatile *_Value, char _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELAXED);
-}
-static __inline__ char __DEFAULT_FN_ATTRS
-_InterlockedOr8_rel(char volatile *_Value, char _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELEASE);
-}
-static __inline__ short __DEFAULT_FN_ATTRS
-_InterlockedOr16_acq(short volatile *_Value, short _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_ACQUIRE);
-}
-static __inline__ short __DEFAULT_FN_ATTRS
-_InterlockedOr16_nf(short volatile *_Value, short _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELAXED);
-}
-static __inline__ short __DEFAULT_FN_ATTRS
-_InterlockedOr16_rel(short volatile *_Value, short _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELEASE);
-}
-static __inline__ long __DEFAULT_FN_ATTRS
-_InterlockedOr_acq(long volatile *_Value, long _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_ACQUIRE);
-}
-static __inline__ long __DEFAULT_FN_ATTRS
-_InterlockedOr_nf(long volatile *_Value, long _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELAXED);
-}
-static __inline__ long __DEFAULT_FN_ATTRS
-_InterlockedOr_rel(long volatile *_Value, long _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELEASE);
-}
-static __inline__ __int64 __DEFAULT_FN_ATTRS
-_InterlockedOr64_acq(__int64 volatile *_Value, __int64 _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_ACQUIRE);
-}
-static __inline__ __int64 __DEFAULT_FN_ATTRS
-_InterlockedOr64_nf(__int64 volatile *_Value, __int64 _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELAXED);
-}
-static __inline__ __int64 __DEFAULT_FN_ATTRS
-_InterlockedOr64_rel(__int64 volatile *_Value, __int64 _Mask) {
-  return __atomic_fetch_or(_Value, _Mask, __ATOMIC_RELEASE);
-}
+char _InterlockedOr8_acq(char volatile *_Value, char _Mask);
+char _InterlockedOr8_nf(char volatile *_Value, char _Mask);
+char _InterlockedOr8_rel(char volatile *_Value, char _Mask);
+short _InterlockedOr16_acq(short volatile *_Value, short _Mask);
+short _InterlockedOr16_nf(short volatile *_Value, short _Mask);
+short _InterlockedOr16_rel(short volatile *_Value, short _Mask);
+long _InterlockedOr_acq(long volatile *_Value, long _Mask);
+long _InterlockedOr_nf(long volatile *_Value, long _Mask);
+long _InterlockedOr_rel(long volatile *_Value, long _Mask);
+__int64 _InterlockedOr64_acq(__int64 volatile *_Value, __int64 _Mask);
+__int64 _InterlockedOr64_nf(__int64 volatile *_Value, __int64 _Mask);
+__int64 _InterlockedOr64_rel(__int64 volatile *_Value, __int64 _Mask);
 #endif
 /*----------------------------------------------------------------------------*\
 |* Interlocked Xor
 \*----------------------------------------------------------------------------*/
 #if defined(__arm__) || defined(__aarch64__)
 static __inline__ char __DEFAULT_FN_ATTRS
 _InterlockedXor8_acq(char volatile *_Value, char _Mask) {
   return __atomic_fetch_xor(_Value, _Mask, __ATOMIC_ACQUIRE);

test/CodeGen/ms-intrinsics.c

 __int64 test_InterlockedCompareExchange64_nf(__int64 volatile *Destination, __int64 Exchange, __int64 Comperand) {
   return _InterlockedCompareExchange64_nf(Destination, Exchange, Comperand);
 }
 // CHECK-ARM-ARM64: define{{.*}}i64 @test_InterlockedCompareExchange64_nf(i64*{{[a-z_ ]*}}%Destination, i64{{[a-z_ ]*}}%Exchange, i64{{[a-z_ ]*}}%Comperand){{.*}}{
 // CHECK-ARM-ARM64: [[TMP:%[0-9]+]] = cmpxchg volatile i64* %Destination, i64 %Comperand, i64 %Exchange monotonic monotonic
 // CHECK-ARM-ARM64: [[RESULT:%[0-9]+]] = extractvalue { i64, i1 } [[TMP]], 0
 // CHECK-ARM-ARM64: ret i64 [[RESULT]]
 // CHECK-ARM-ARM64: }
+
+char test_InterlockedOr8_acq(char volatile *value, char mask) {
+  return _InterlockedOr8_acq(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i8 @test_InterlockedOr8_acq(i8*{{[a-z_ ]*}}%value, i8{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i8* %value, i8 %mask acquire
+// CHECK-ARM-ARM64:   ret i8 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+char test_InterlockedOr8_rel(char volatile *value, char mask) {
+  return _InterlockedOr8_rel(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i8 @test_InterlockedOr8_rel(i8*{{[a-z_ ]*}}%value, i8{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i8* %value, i8 %mask release
+// CHECK-ARM-ARM64:   ret i8 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+char test_InterlockedOr8_nf(char volatile *value, char mask) {
+  return _InterlockedOr8_nf(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i8 @test_InterlockedOr8_nf(i8*{{[a-z_ ]*}}%value, i8{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i8* %value, i8 %mask monotonic
+// CHECK-ARM-ARM64:   ret i8 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+short test_InterlockedOr16_acq(short volatile *value, short mask) {
+  return _InterlockedOr16_acq(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i16 @test_InterlockedOr16_acq(i16*{{[a-z_ ]*}}%value, i16{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i16* %value, i16 %mask acquire
+// CHECK-ARM-ARM64:   ret i16 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+short test_InterlockedOr16_rel(short volatile *value, short mask) {
+  return _InterlockedOr16_rel(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i16 @test_InterlockedOr16_rel(i16*{{[a-z_ ]*}}%value, i16{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i16* %value, i16 %mask release
+// CHECK-ARM-ARM64:   ret i16 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+short test_InterlockedOr16_nf(short volatile *value, short mask) {
+  return _InterlockedOr16_nf(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i16 @test_InterlockedOr16_nf(i16*{{[a-z_ ]*}}%value, i16{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i16* %value, i16 %mask monotonic
+// CHECK-ARM-ARM64:   ret i16 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+long test_InterlockedOr_acq(long volatile *value, long mask) {
+  return _InterlockedOr_acq(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i32 @test_InterlockedOr_acq(i32*{{[a-z_ ]*}}%value, i32{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i32* %value, i32 %mask acquire
+// CHECK-ARM-ARM64:   ret i32 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+long test_InterlockedOr_rel(long volatile *value, long mask) {
+  return _InterlockedOr_rel(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i32 @test_InterlockedOr_rel(i32*{{[a-z_ ]*}}%value, i32{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i32* %value, i32 %mask release
+// CHECK-ARM-ARM64:   ret i32 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+long test_InterlockedOr_nf(long volatile *value, long mask) {
+  return _InterlockedOr_nf(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i32 @test_InterlockedOr_nf(i32*{{[a-z_ ]*}}%value, i32{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i32* %value, i32 %mask monotonic
+// CHECK-ARM-ARM64:   ret i32 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+__int64 test_InterlockedOr64_acq(__int64 volatile *value, __int64 mask) {
+  return _InterlockedOr64_acq(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i64 @test_InterlockedOr64_acq(i64*{{[a-z_ ]*}}%value, i64{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i64* %value, i64 %mask acquire
+// CHECK-ARM-ARM64:   ret i64 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+__int64 test_InterlockedOr64_rel(__int64 volatile *value, __int64 mask) {
+  return _InterlockedOr64_rel(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i64 @test_InterlockedOr64_rel(i64*{{[a-z_ ]*}}%value, i64{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i64* %value, i64 %mask release
+// CHECK-ARM-ARM64:   ret i64 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
+
+__int64 test_InterlockedOr64_nf(__int64 volatile *value, __int64 mask) {
+  return _InterlockedOr64_nf(value, mask);
+}
+// CHECK-ARM-ARM64: define{{.*}}i64 @test_InterlockedOr64_nf(i64*{{[a-z_ ]*}}%value, i64{{[a-z_ ]*}}%mask){{.*}}{
+// CHECK-ARM-ARM64:   [[RESULT:%[0-9]+]] = atomicrmw or i64* %value, i64 %mask monotonic
+// CHECK-ARM-ARM64:   ret i64 [[RESULT:%[0-9]+]]
+// CHECK-ARM-ARM64: }
 #endif
 
 #if !defined(__aarch64__)
 void test__fastfail() {
   __fastfail(42);
 }
 // CHECK-LABEL: define{{.*}} void @test__fastfail()
 // CHECK-ARM: call void asm sideeffect "udf #251", "{r0}"(i32 42) #[[NORETURN:[0-9]+]]
 // CHECK-INTEL: call void asm sideeffect "int $$0x29", "{cx}"(i32 42) #[[NORETURN]]
 
 // Attributes come last.
 
 // CHECK: attributes #[[NORETURN]] = { noreturn{{.*}} }
 #endif