Add __sync_fetch_and_nand (again)

Prior to GCC 4.4, __sync_fetch_and_nand was implemented as:

  { tmp = *ptr; *ptr = ~tmp & value; return tmp; }

but this was changed in GCC 4.4 to be:

  { tmp = *ptr; *ptr = ~(tmp & value); return tmp; }

in response to this change, support for sync_fetch_and_nand (and
sync_nand_and_fetch) was removed in r99522 in order to avoid miscompiling code
depending on the old semantics. However, at this point:

  1. Many years have passed, and the amount of code relying on the old
     semantics is likely smaller.

  2. Through the work of many contributors, all LLVM backends have been updated
     such that "atomicrmw nand" provides the newer GCC 4.4+ semantics (this process
     was complete July of 2014 (added to the release notes in r212635).

  3. The lack of this intrinsic is now a needless impediment to porting codes
     from GCC to Clang (I've now seen several examples of this).

It is true, however, that we still set GNUC_MINOR to 2 (corresponding to GCC
4.2). To compensate for this, and to address the original concern regarding
code relying on the old semantics, I've added a warning that specifically
details the fact that the semantics have changed and that we provide the newer
semantics.

Fixes PR8842.

llvm-svn: 218905

7 files changed, 144 insertions, 13 deletions

diff --git a/clang/include/clang/Basic/Builtins.def b/clang/include/clang/Basic/Builtins.def
index bb6975c4946..72843f1fb6a 100644
--- a/clang/include/clang/Basic/Builtins.def
+++ b/clang/include/clang/Basic/Builtins.def
@@ -541,6 +541,12 @@ BUILTIN(__sync_fetch_and_xor_4, "iiD*i.", "tn")
 BUILTIN(__sync_fetch_and_xor_8, "LLiLLiD*LLi.", "tn")
 BUILTIN(__sync_fetch_and_xor_16, "LLLiLLLiD*LLLi.", "tn")
 
+BUILTIN(__sync_fetch_and_nand, "v.", "t")
+BUILTIN(__sync_fetch_and_nand_1, "ccD*c.", "tn")
+BUILTIN(__sync_fetch_and_nand_2, "ssD*s.", "tn")
+BUILTIN(__sync_fetch_and_nand_4, "iiD*i.", "tn")
+BUILTIN(__sync_fetch_and_nand_8, "LLiLLiD*LLi.", "tn")
+BUILTIN(__sync_fetch_and_nand_16, "LLLiLLLiD*LLLi.", "tn")
 
 BUILTIN(__sync_add_and_fetch, "v.", "t")
 BUILTIN(__sync_add_and_fetch_1, "ccD*c.", "tn")
@@ -577,6 +583,13 @@ BUILTIN(__sync_xor_and_fetch_4, "iiD*i.", "tn")
 BUILTIN(__sync_xor_and_fetch_8, "LLiLLiD*LLi.", "tn")
 BUILTIN(__sync_xor_and_fetch_16, "LLLiLLLiD*LLLi.", "tn")
 
+BUILTIN(__sync_nand_and_fetch, "v.", "t")
+BUILTIN(__sync_nand_and_fetch_1, "ccD*c.", "tn")
+BUILTIN(__sync_nand_and_fetch_2, "ssD*s.", "tn")
+BUILTIN(__sync_nand_and_fetch_4, "iiD*i.", "tn")
+BUILTIN(__sync_nand_and_fetch_8, "LLiLLiD*LLi.", "tn")
+BUILTIN(__sync_nand_and_fetch_16, "LLLiLLLiD*LLLi.", "tn")
+
 BUILTIN(__sync_bool_compare_and_swap, "v.", "t")
 BUILTIN(__sync_bool_compare_and_swap_1, "bcD*cc.", "tn")
 BUILTIN(__sync_bool_compare_and_swap_2, "bsD*ss.", "tn")
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index 06a22045ea8..c811db33bc8 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -6898,6 +6898,11 @@ def warn_duplicate_attribute : Warning<
   "attribute %0 is already applied with different parameters">,
   InGroup<IgnoredAttributes>;
 
+def warn_sync_fetch_and_nand_semantics_change : Warning<
+  "the semantics of this intrinsic changed with GCC "
+  "version 4.4 - the newer semantics are provided here">,
+  InGroup<DiagGroup<"sync-fetch-and-nand-semantics-changed">>;
+
 // Type
 def ext_invalid_sign_spec : Extension<"'%0' cannot be signed or unsigned">;
 def warn_receiver_forward_class : Warning<
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 9238aae88cd..f59cbaa4b8a 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -113,7 +113,8 @@ static RValue EmitBinaryAtomic(CodeGenFunction &CGF,
 static RValue EmitBinaryAtomicPost(CodeGenFunction &CGF,
                                    llvm::AtomicRMWInst::BinOp Kind,
                                    const CallExpr *E,
-                                   Instruction::BinaryOps Op) {
+                                   Instruction::BinaryOps Op,
+                                   bool Invert = false) {
   QualType T = E->getType();
   assert(E->getArg(0)->getType()->isPointerType());
   assert(CGF.getContext().hasSameUnqualifiedType(T,
@@ -138,6 +139,9 @@ static RValue EmitBinaryAtomicPost(CodeGenFunction &CGF,
       CGF.Builder.CreateAtomicRMW(Kind, Args[0], Args[1],
                                   llvm::SequentiallyConsistent);
   Result = CGF.Builder.CreateBinOp(Op, Result, Args[1]);
+  if (Invert)
+    Result = CGF.Builder.CreateBinOp(llvm::Instruction::Xor, Result,
+                                     llvm::ConstantInt::get(IntType, -1));
   Result = EmitFromInt(CGF, Result, T, ValueType);
   return RValue::get(Result);
 }
@@ -879,11 +883,13 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   case Builtin::BI__sync_fetch_and_or:
   case Builtin::BI__sync_fetch_and_and:
   case Builtin::BI__sync_fetch_and_xor:
+  case Builtin::BI__sync_fetch_and_nand:
   case Builtin::BI__sync_add_and_fetch:
   case Builtin::BI__sync_sub_and_fetch:
   case Builtin::BI__sync_and_and_fetch:
   case Builtin::BI__sync_or_and_fetch:
   case Builtin::BI__sync_xor_and_fetch:
+  case Builtin::BI__sync_nand_and_fetch:
   case Builtin::BI__sync_val_compare_and_swap:
   case Builtin::BI__sync_bool_compare_and_swap:
   case Builtin::BI__sync_lock_test_and_set:
@@ -920,6 +926,12 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   case Builtin::BI__sync_fetch_and_xor_8:
   case Builtin::BI__sync_fetch_and_xor_16:
     return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Xor, E);
+  case Builtin::BI__sync_fetch_and_nand_1:
+  case Builtin::BI__sync_fetch_and_nand_2:
+  case Builtin::BI__sync_fetch_and_nand_4:
+  case Builtin::BI__sync_fetch_and_nand_8:
+  case Builtin::BI__sync_fetch_and_nand_16:
+    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Nand, E);
 
   // Clang extensions: not overloaded yet.
   case Builtin::BI__sync_fetch_and_min:
@@ -966,6 +978,13 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   case Builtin::BI__sync_xor_and_fetch_16:
     return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Xor, E,
                                 llvm::Instruction::Xor);
+  case Builtin::BI__sync_nand_and_fetch_1:
+  case Builtin::BI__sync_nand_and_fetch_2:
+  case Builtin::BI__sync_nand_and_fetch_4:
+  case Builtin::BI__sync_nand_and_fetch_8:
+  case Builtin::BI__sync_nand_and_fetch_16:
+    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Nand, E,
+                                llvm::Instruction::And, true);
 
   case Builtin::BI__sync_val_compare_and_swap_1:
   case Builtin::BI__sync_val_compare_and_swap_2:
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index b4f90d3b4f6..a856c98bf05 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -273,6 +273,12 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID,
   case Builtin::BI__sync_fetch_and_xor_4:
   case Builtin::BI__sync_fetch_and_xor_8:
   case Builtin::BI__sync_fetch_and_xor_16:
+  case Builtin::BI__sync_fetch_and_nand:
+  case Builtin::BI__sync_fetch_and_nand_1:
+  case Builtin::BI__sync_fetch_and_nand_2:
+  case Builtin::BI__sync_fetch_and_nand_4:
+  case Builtin::BI__sync_fetch_and_nand_8:
+  case Builtin::BI__sync_fetch_and_nand_16:
   case Builtin::BI__sync_add_and_fetch:
   case Builtin::BI__sync_add_and_fetch_1:
   case Builtin::BI__sync_add_and_fetch_2:
@@ -303,6 +309,12 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID,
   case Builtin::BI__sync_xor_and_fetch_4:
   case Builtin::BI__sync_xor_and_fetch_8:
   case Builtin::BI__sync_xor_and_fetch_16:
+  case Builtin::BI__sync_nand_and_fetch:
+  case Builtin::BI__sync_nand_and_fetch_1:
+  case Builtin::BI__sync_nand_and_fetch_2:
+  case Builtin::BI__sync_nand_and_fetch_4:
+  case Builtin::BI__sync_nand_and_fetch_8:
+  case Builtin::BI__sync_nand_and_fetch_16:
   case Builtin::BI__sync_val_compare_and_swap:
   case Builtin::BI__sync_val_compare_and_swap_1:
   case Builtin::BI__sync_val_compare_and_swap_2:
@@ -1518,12 +1530,14 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
     BUILTIN_ROW(__sync_fetch_and_or),
     BUILTIN_ROW(__sync_fetch_and_and),
     BUILTIN_ROW(__sync_fetch_and_xor),
+    BUILTIN_ROW(__sync_fetch_and_nand),
 
     BUILTIN_ROW(__sync_add_and_fetch),
     BUILTIN_ROW(__sync_sub_and_fetch),
     BUILTIN_ROW(__sync_and_and_fetch),
     BUILTIN_ROW(__sync_or_and_fetch),
     BUILTIN_ROW(__sync_xor_and_fetch),
+    BUILTIN_ROW(__sync_nand_and_fetch),
 
     BUILTIN_ROW(__sync_val_compare_and_swap),
     BUILTIN_ROW(__sync_bool_compare_and_swap),
@@ -1553,6 +1567,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   // as the number of fixed args.
   unsigned BuiltinID = FDecl->getBuiltinID();
   unsigned BuiltinIndex, NumFixed = 1;
+  bool WarnAboutSemanticsChange = false;
   switch (BuiltinID) {
   default: llvm_unreachable("Unknown overloaded atomic builtin!");
   case Builtin::BI__sync_fetch_and_add: 
@@ -1600,13 +1615,23 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
     BuiltinIndex = 4; 
     break;
 
+  case Builtin::BI__sync_fetch_and_nand: 
+  case Builtin::BI__sync_fetch_and_nand_1:
+  case Builtin::BI__sync_fetch_and_nand_2:
+  case Builtin::BI__sync_fetch_and_nand_4:
+  case Builtin::BI__sync_fetch_and_nand_8:
+  case Builtin::BI__sync_fetch_and_nand_16:
+    BuiltinIndex = 5;
+    WarnAboutSemanticsChange = true;
+    break;
+
   case Builtin::BI__sync_add_and_fetch: 
   case Builtin::BI__sync_add_and_fetch_1:
   case Builtin::BI__sync_add_and_fetch_2:
   case Builtin::BI__sync_add_and_fetch_4:
   case Builtin::BI__sync_add_and_fetch_8:
   case Builtin::BI__sync_add_and_fetch_16:
-    BuiltinIndex = 5; 
+    BuiltinIndex = 6; 
     break;
       
   case Builtin::BI__sync_sub_and_fetch: 
@@ -1615,7 +1640,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_sub_and_fetch_4:
   case Builtin::BI__sync_sub_and_fetch_8:
   case Builtin::BI__sync_sub_and_fetch_16:
-    BuiltinIndex = 6; 
+    BuiltinIndex = 7; 
     break;
       
   case Builtin::BI__sync_and_and_fetch: 
@@ -1624,7 +1649,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_and_and_fetch_4:
   case Builtin::BI__sync_and_and_fetch_8:
   case Builtin::BI__sync_and_and_fetch_16:
-    BuiltinIndex = 7; 
+    BuiltinIndex = 8; 
     break;
       
   case Builtin::BI__sync_or_and_fetch:  
@@ -1633,7 +1658,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_or_and_fetch_4:
   case Builtin::BI__sync_or_and_fetch_8:
   case Builtin::BI__sync_or_and_fetch_16:
-    BuiltinIndex = 8; 
+    BuiltinIndex = 9; 
     break;
       
   case Builtin::BI__sync_xor_and_fetch: 
@@ -1642,7 +1667,17 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_xor_and_fetch_4:
   case Builtin::BI__sync_xor_and_fetch_8:
   case Builtin::BI__sync_xor_and_fetch_16:
-    BuiltinIndex = 9; 
+    BuiltinIndex = 10;
+    break;
+
+  case Builtin::BI__sync_nand_and_fetch: 
+  case Builtin::BI__sync_nand_and_fetch_1:
+  case Builtin::BI__sync_nand_and_fetch_2:
+  case Builtin::BI__sync_nand_and_fetch_4:
+  case Builtin::BI__sync_nand_and_fetch_8:
+  case Builtin::BI__sync_nand_and_fetch_16:
+    BuiltinIndex = 11;
+    WarnAboutSemanticsChange = true;
     break;
 
   case Builtin::BI__sync_val_compare_and_swap:
@@ -1651,7 +1686,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_val_compare_and_swap_4:
   case Builtin::BI__sync_val_compare_and_swap_8:
   case Builtin::BI__sync_val_compare_and_swap_16:
-    BuiltinIndex = 10;
+    BuiltinIndex = 12;
     NumFixed = 2;
     break;
       
@@ -1661,7 +1696,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_bool_compare_and_swap_4:
   case Builtin::BI__sync_bool_compare_and_swap_8:
   case Builtin::BI__sync_bool_compare_and_swap_16:
-    BuiltinIndex = 11;
+    BuiltinIndex = 13;
     NumFixed = 2;
     ResultType = Context.BoolTy;
     break;
@@ -1672,7 +1707,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_lock_test_and_set_4:
   case Builtin::BI__sync_lock_test_and_set_8:
   case Builtin::BI__sync_lock_test_and_set_16:
-    BuiltinIndex = 12; 
+    BuiltinIndex = 14; 
     break;
       
   case Builtin::BI__sync_lock_release:
@@ -1681,7 +1716,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_lock_release_4:
   case Builtin::BI__sync_lock_release_8:
   case Builtin::BI__sync_lock_release_16:
-    BuiltinIndex = 13;
+    BuiltinIndex = 15;
     NumFixed = 0;
     ResultType = Context.VoidTy;
     break;
@@ -1692,7 +1727,7 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
   case Builtin::BI__sync_swap_4:
   case Builtin::BI__sync_swap_8:
   case Builtin::BI__sync_swap_16:
-    BuiltinIndex = 14; 
+    BuiltinIndex = 16; 
     break;
   }
 
@@ -1705,6 +1740,11 @@ Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) {
     return ExprError();
   }
 
+  if (WarnAboutSemanticsChange) {
+    Diag(TheCall->getLocEnd(), diag::warn_sync_fetch_and_nand_semantics_change)
+      << TheCall->getCallee()->getSourceRange();
+  }
+
   // Get the decl for the concrete builtin from this, we can tell what the
   // concrete integer type we should convert to is.
   unsigned NewBuiltinID = BuiltinIndices[BuiltinIndex][SizeIndex];
diff --git a/clang/test/CodeGen/Atomics.c b/clang/test/CodeGen/Atomics.c
index 684f36d4044..f957883575d 100644
--- a/clang/test/CodeGen/Atomics.c
+++ b/clang/test/CodeGen/Atomics.c
@@ -49,6 +49,15 @@ void test_op_ignore (void) // CHECK-LABEL: define void @test_op_ignore
   (void) __sync_fetch_and_xor (&sll, 1); // CHECK: atomicrmw xor i64
   (void) __sync_fetch_and_xor (&ull, 1); // CHECK: atomicrmw xor i64
 
+  (void) __sync_fetch_and_nand (&sc, 1); // CHECK: atomicrmw nand i8
+  (void) __sync_fetch_and_nand (&uc, 1); // CHECK: atomicrmw nand i8
+  (void) __sync_fetch_and_nand (&ss, 1); // CHECK: atomicrmw nand i16
+  (void) __sync_fetch_and_nand (&us, 1); // CHECK: atomicrmw nand i16
+  (void) __sync_fetch_and_nand (&si, 1); // CHECK: atomicrmw nand i32
+  (void) __sync_fetch_and_nand (&ui, 1); // CHECK: atomicrmw nand i32
+  (void) __sync_fetch_and_nand (&sll, 1); // CHECK: atomicrmw nand i64
+  (void) __sync_fetch_and_nand (&ull, 1); // CHECK: atomicrmw nand i64
+
   (void) __sync_fetch_and_and (&sc, 1); // CHECK: atomicrmw and i8
   (void) __sync_fetch_and_and (&uc, 1); // CHECK: atomicrmw and i8
   (void) __sync_fetch_and_and (&ss, 1); // CHECK: atomicrmw and i16
@@ -98,6 +107,15 @@ void test_fetch_and_op (void) // CHECK-LABEL: define void @test_fetch_and_op
   sll = __sync_fetch_and_xor (&sll, 11); // CHECK: atomicrmw xor
   ull = __sync_fetch_and_xor (&ull, 11); // CHECK: atomicrmw xor
 
+  sc = __sync_fetch_and_nand (&sc, 11); // CHECK: atomicrmw nand
+  uc = __sync_fetch_and_nand (&uc, 11); // CHECK: atomicrmw nand
+  ss = __sync_fetch_and_nand (&ss, 11); // CHECK: atomicrmw nand
+  us = __sync_fetch_and_nand (&us, 11); // CHECK: atomicrmw nand
+  si = __sync_fetch_and_nand (&si, 11); // CHECK: atomicrmw nand
+  ui = __sync_fetch_and_nand (&ui, 11); // CHECK: atomicrmw nand
+  sll = __sync_fetch_and_nand (&sll, 11); // CHECK: atomicrmw nand
+  ull = __sync_fetch_and_nand (&ull, 11); // CHECK: atomicrmw nand
+
   sc = __sync_fetch_and_and (&sc, 11); // CHECK: atomicrmw and
   uc = __sync_fetch_and_and (&uc, 11); // CHECK: atomicrmw and
   ss = __sync_fetch_and_and (&ss, 11); // CHECK: atomicrmw and
@@ -147,6 +165,31 @@ void test_op_and_fetch (void)
   sll = __sync_xor_and_fetch (&sll, uc); // CHECK: atomicrmw xor
   ull = __sync_xor_and_fetch (&ull, uc); // CHECK: atomicrmw xor
 
+  sc = __sync_nand_and_fetch (&sc, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  uc = __sync_nand_and_fetch (&uc, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  ss = __sync_nand_and_fetch (&ss, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  us = __sync_nand_and_fetch (&us, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  si = __sync_nand_and_fetch (&si, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  ui = __sync_nand_and_fetch (&ui, uc); // CHECK: atomicrmw nand
+                                        // CHECK: and
+                                        // CHECK: xor
+  sll = __sync_nand_and_fetch (&sll, uc); // CHECK: atomicrmw nand
+                                          // CHECK: and
+                                          // CHECK: xor
+  ull = __sync_nand_and_fetch (&ull, uc); // CHECK: atomicrmw nand
+                                          // CHECK: and
+                                          // CHECK: xor
+
   sc = __sync_and_and_fetch (&sc, uc); // CHECK: atomicrmw and
   uc = __sync_and_and_fetch (&uc, uc); // CHECK: atomicrmw and
   ss = __sync_and_and_fetch (&ss, uc); // CHECK: atomicrmw and
diff --git a/clang/test/CodeGen/atomic.c b/clang/test/CodeGen/atomic.c
index 43f5bc81ee9..4db3c8e6d69 100644
--- a/clang/test/CodeGen/atomic.c
+++ b/clang/test/CodeGen/atomic.c
@@ -50,7 +50,10 @@ int atomic(void) {
 
   old = __sync_fetch_and_xor(&val, 0xb);
   // CHECK: atomicrmw xor i32* %val, i32 11 seq_cst
-  
+ 
+  old = __sync_fetch_and_nand(&val, 0xc);
+  // CHECK: atomicrmw nand i32* %val, i32 12 seq_cst
+ 
   old = __sync_add_and_fetch(&val, 1);
   // CHECK: atomicrmw add i32* %val, i32 1 seq_cst
 
@@ -65,7 +68,10 @@ int atomic(void) {
 
   old = __sync_xor_and_fetch(&valc, 5);
   // CHECK: atomicrmw xor i8* %valc, i8 5 seq_cst  
-  
+ 
+  old = __sync_nand_and_fetch(&valc, 6);
+  // CHECK: atomicrmw nand i8* %valc, i8 6 seq_cst  
+ 
   __sync_val_compare_and_swap((void **)0, (void *)0, (void *)0);
   // CHECK: [[PAIR:%[a-z0-9_.]+]] = cmpxchg i32* null, i32 0, i32 0 seq_cst
   // CHECK: extractvalue { i32, i1 } [[PAIR]], 0
diff --git a/clang/test/Sema/builtins.c b/clang/test/Sema/builtins.c
index f8df2f85cf6..0e2925e0748 100644
--- a/clang/test/Sema/builtins.c
+++ b/clang/test/Sema/builtins.c
@@ -66,6 +66,11 @@ void test9_3(volatile int* ptr, int val) {
   __sync_fetch_and_add_4(ptr, val);
 }
 
+void test9_4(volatile int* ptr, int val) {
+  // expected-warning@+1 {{the semantics of this intrinsic changed with GCC version 4.4 - the newer semantics are provided here}}
+  __sync_fetch_and_nand(ptr, val);
+}
+
 // rdar://7236819
 void test10(void) __attribute__((noreturn));