diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -11895,16 +11895,28 @@
   case X86::BI_mm_setcsr:
   case X86::BI__builtin_ia32_ldmxcsr: {
     Address Tmp = CreateMemTemp(E->getArg(0)->getType());
+    llvm::Value *TmpPtr = Tmp.getPointer();
+    llvm::Value *TmpSize = EmitLifetimeStart(
+        CGM.getDataLayout().getTypeAllocSize(Tmp.getElementType()), TmpPtr);
     Builder.CreateStore(Ops[0], Tmp);
-    return Builder.CreateCall(CGM.getIntrinsic(Intrinsic::x86_sse_ldmxcsr),
-                          Builder.CreateBitCast(Tmp.getPointer(), Int8PtrTy));
+    Value *Call = Builder.CreateCall(CGM.getIntrinsic(Intrinsic::x86_sse_ldmxcsr),
+                                     Builder.CreateBitCast(Tmp.getPointer(), Int8PtrTy));
+    if (TmpSize)
+      EmitLifetimeEnd(TmpSize, TmpPtr);
+    return Call;
   }
   case X86::BI_mm_getcsr:
   case X86::BI__builtin_ia32_stmxcsr: {
     Address Tmp = CreateMemTemp(E->getType());
+    llvm::Value *TmpPtr = Tmp.getPointer();
+    llvm::Value *TmpSize = EmitLifetimeStart(
+        CGM.getDataLayout().getTypeAllocSize(Tmp.getElementType()), TmpPtr);
     Builder.CreateCall(CGM.getIntrinsic(Intrinsic::x86_sse_stmxcsr),
                        Builder.CreateBitCast(Tmp.getPointer(), Int8PtrTy));
-    return Builder.CreateLoad(Tmp, "stmxcsr");
+    Value *Call = Builder.CreateLoad(Tmp, "stmxcsr");
+    if (TmpSize)
+      EmitLifetimeEnd(TmpSize, TmpPtr);
+    return Call;
   }
   case X86::BI__builtin_ia32_xsave:
   case X86::BI__builtin_ia32_xsave64: