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
@@ -13926,25 +13926,15 @@
   }
   case X86::BI__shiftleft128:
   case X86::BI__shiftright128: {
-    // FIXME: Once fshl/fshr no longer add an unneeded and and cmov, do this:
-    // llvm::Function *F = CGM.getIntrinsic(
-    //   BuiltinID == X86::BI__shiftleft128 ? Intrinsic::fshl : Intrinsic::fshr,
-    //   Int64Ty);
-    // Ops[2] = Builder.CreateZExt(Ops[2], Int64Ty);
-    // return Builder.CreateCall(F, Ops);
-    llvm::Type *Int128Ty = Builder.getInt128Ty();
-    Value *HighPart128 =
-        Builder.CreateShl(Builder.CreateZExt(Ops[1], Int128Ty), 64);
-    Value *LowPart128 = Builder.CreateZExt(Ops[0], Int128Ty);
-    Value *Val = Builder.CreateOr(HighPart128, LowPart128);
-    Value *Amt = Builder.CreateAnd(Builder.CreateZExt(Ops[2], Int128Ty),
-                                   llvm::ConstantInt::get(Int128Ty, 0x3f));
-    Value *Res;
-    if (BuiltinID == X86::BI__shiftleft128)
-      Res = Builder.CreateLShr(Builder.CreateShl(Val, Amt), 64);
-    else
-      Res = Builder.CreateLShr(Val, Amt);
-    return Builder.CreateTrunc(Res, Int64Ty);
+    llvm::Function *F = CGM.getIntrinsic(
+        BuiltinID == X86::BI__shiftleft128 ? Intrinsic::fshl : Intrinsic::fshr,
+        Int64Ty);
+    // Flip low/high ops and zero-extend amount to matching type.
+    // shiftleft128(Low, High, Amt) -> fshl(High, Low, Amt)
+    // shiftright128(Low, High, Amt) -> fshr(High, Low, Amt)
+    std::swap(Ops[0], Ops[1]);
+    Ops[2] = Builder.CreateZExt(Ops[2], Int64Ty);
+    return Builder.CreateCall(F, Ops);
   }
   case X86::BI_ReadWriteBarrier:
   case X86::BI_ReadBarrier:
diff --git a/clang/test/CodeGen/X86/ms-x86-intrinsics.c b/clang/test/CodeGen/X86/ms-x86-intrinsics.c
--- a/clang/test/CodeGen/X86/ms-x86-intrinsics.c
+++ b/clang/test/CodeGen/X86/ms-x86-intrinsics.c
@@ -144,14 +144,8 @@
   return __shiftleft128(l, h, d);
 }
 // CHECK-X64-LABEL: define dso_local i64 @test__shiftleft128(i64 %l, i64 %h, i8 %d)
-// CHECK-X64:  = zext i64 %{{.*}} to i128
-// CHECK-X64: = shl nuw i128 %{{.*}}, 64
-// CHECK-X64: = zext i64 %{{.*}} to i128
-// CHECK-X64: = or i128 %
-// CHECK-X64: = and i8 %{{.*}}, 63
-// CHECK-X64: = shl i128 %
-// CHECK-X64: = lshr i128 %
-// CHECK-X64: = trunc i128 %
+// CHECK-X64: = zext i8 %{{.*}} to i64
+// CHECK-X64: = tail call i64 @llvm.fshl.i64(i64 %h, i64 %l, i64 %{{.*}})
 // CHECK-X64:  ret i64 %
 
 unsigned __int64 test__shiftright128(unsigned __int64 l, unsigned __int64 h,
@@ -159,13 +153,8 @@
   return __shiftright128(l, h, d);
 }
 // CHECK-X64-LABEL: define dso_local i64 @test__shiftright128(i64 %l, i64 %h, i8 %d)
-// CHECK-X64:  = zext i64 %{{.*}} to i128
-// CHECK-X64: = shl nuw i128 %{{.*}}, 64
-// CHECK-X64: = zext i64 %{{.*}} to i128
-// CHECK-X64: = or i128 %
-// CHECK-X64: = and i8 %{{.*}}, 63
-// CHECK-X64: = lshr i128 %
-// CHECK-X64: = trunc i128 %
+// CHECK-X64: = zext i8 %{{.*}} to i64
+// CHECK-X64: = tail call i64 @llvm.fshr.i64(i64 %h, i64 %l, i64 %{{.*}})
 // CHECK-X64:  ret i64 %
 
 #endif // defined(__x86_64__)