Index: include/llvm/Transforms/Utils/SimplifyLibCalls.h
===================================================================
--- include/llvm/Transforms/Utils/SimplifyLibCalls.h
+++ include/llvm/Transforms/Utils/SimplifyLibCalls.h
@@ -168,6 +168,7 @@
   Value *optimizeBCmp(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemCmpBCmpCommon(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemCpy(CallInst *CI, IRBuilder<> &B);
+  Value *optimizeMemPCpy(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemMove(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemSet(CallInst *CI, IRBuilder<> &B);
   Value *optimizeRealloc(CallInst *CI, IRBuilder<> &B);
Index: lib/Transforms/Utils/SimplifyLibCalls.cpp
===================================================================
--- lib/Transforms/Utils/SimplifyLibCalls.cpp
+++ lib/Transforms/Utils/SimplifyLibCalls.cpp
@@ -962,6 +962,14 @@
   return CI->getArgOperand(0);
 }
 
+Value *LibCallSimplifier::optimizeMemPCpy(CallInst *CI, IRBuilder<> &B) {
+  Value *Dst = CI->getArgOperand(0);
+  Value *N = CI->getArgOperand(2);
+  // mempcpy(x, y, n) -> llvm.memcpy(align 1 x, align 1 y, n), x + n
+  B.CreateMemCpy(Dst, 1, CI->getArgOperand(1), 1, N);
+  return B.CreateInBoundsGEP(B.getInt8Ty(), Dst, N);
+}
+
 Value *LibCallSimplifier::optimizeMemMove(CallInst *CI, IRBuilder<> &B) {
   // memmove(x, y, n) -> llvm.memmove(align 1 x, align 1 y, n)
   B.CreateMemMove(CI->getArgOperand(0), 1, CI->getArgOperand(1), 1,
@@ -2579,6 +2587,8 @@
       return optimizeMemCmp(CI, Builder);
     case LibFunc_memcpy:
       return optimizeMemCpy(CI, Builder);
+    case LibFunc_mempcpy:
+      return optimizeMemPCpy(CI, Builder);
     case LibFunc_memmove:
       return optimizeMemMove(CI, Builder);
     case LibFunc_memset:
Index: test/Transforms/InstCombine/mempcpy.ll
===================================================================
--- test/Transforms/InstCombine/mempcpy.ll
+++ test/Transforms/InstCombine/mempcpy.ll
@@ -3,8 +3,9 @@
 
 define i8* @memcpy_nonconst_n(i8* %d, i8* nocapture readonly %s, i64 %n) {
 ; CHECK-LABEL: @memcpy_nonconst_n(
-; CHECK-NEXT:    [[R:%.*]] = tail call i8* @mempcpy(i8* [[D:%.*]], i8* [[S:%.*]], i64 [[N:%.*]])
-; CHECK-NEXT:    ret i8* [[R]]
+; CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 [[D:%.*]], i8* align 1 [[S:%.*]], i64 [[N:%.*]], i1 false)
+; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i8, i8* [[D]], i64 [[N]]
+; CHECK-NEXT:    ret i8* [[TMP1]]
 ;
   %r = tail call i8* @mempcpy(i8* %d, i8* %s, i64 %n)
   ret i8* %r
@@ -12,8 +13,12 @@
 
 define i8* @memcpy_small_const_n(i8* %d, i8* nocapture readonly %s) {
 ; CHECK-LABEL: @memcpy_small_const_n(
-; CHECK-NEXT:    [[R:%.*]] = tail call i8* @mempcpy(i8* [[D:%.*]], i8* [[S:%.*]], i64 8)
-; CHECK-NEXT:    ret i8* [[R]]
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast i8* [[S:%.*]] to i64*
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[D:%.*]] to i64*
+; CHECK-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]], align 1
+; CHECK-NEXT:    store i64 [[TMP3]], i64* [[TMP2]], align 1
+; CHECK-NEXT:    [[TMP4:%.*]] = getelementptr inbounds i8, i8* [[D]], i64 8
+; CHECK-NEXT:    ret i8* [[TMP4]]
 ;
   %r = tail call i8* @mempcpy(i8* %d, i8* %s, i64 8)
   ret i8* %r
@@ -21,8 +26,9 @@
 
 define i8* @memcpy_big_const_n(i8* %d, i8* nocapture readonly %s) {
 ; CHECK-LABEL: @memcpy_big_const_n(
-; CHECK-NEXT:    [[R:%.*]] = tail call i8* @mempcpy(i8* [[D:%.*]], i8* [[S:%.*]], i64 1024)
-; CHECK-NEXT:    ret i8* [[R]]
+; CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 [[D:%.*]], i8* align 1 [[S:%.*]], i64 1024, i1 false)
+; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i8, i8* [[D]], i64 1024
+; CHECK-NEXT:    ret i8* [[TMP1]]
 ;
   %r = tail call i8* @mempcpy(i8* %d, i8* %s, i64 1024)
   ret i8* %r