Index: lib/Transforms/Scalar/MemCpyOptimizer.cpp
===================================================================
--- lib/Transforms/Scalar/MemCpyOptimizer.cpp
+++ lib/Transforms/Scalar/MemCpyOptimizer.cpp
@@ -884,22 +884,40 @@
   //   c) memcpy from freshly alloca'd space or space that has just started its
   //      lifetime copies undefined data, and we can therefore eliminate the
   //      memcpy in favor of the data that was already at the destination.
-  MemDepResult DepInfo = MD->getDependency(M);
+  AliasAnalysis::Location SrcLoc = AliasAnalysis::getLocationForSource(M);
+  MemDepResult SrcDepInfo = MD->getPointerDependencyFrom(SrcLoc, true,
+                                                         M, M->getParent());
+  AliasAnalysis::Location DstLoc = AliasAnalysis::getLocationForDest(M);
+  MemDepResult DstDepInfo = MD->getPointerDependencyFrom(DstLoc, true,
+                                                         M, M->getParent());
+
+  // If the destination just started its lifetime, pretend that the lifetime
+  // start isn't present, try the callslot optimization and if it succeeds,
+  // extend the lifetime accordingly
+  IntrinsicInst *LT = dyn_cast_or_null<IntrinsicInst>(DstDepInfo.getInst());
+  if (LT && LT->getIntrinsicID() != Intrinsic::lifetime_start)
+    LT = nullptr;
+
+  MemDepResult DepInfo = LT ? SrcDepInfo : MD->getDependency(M);
   if (DepInfo.isClobber()) {
     if (CallInst *C = dyn_cast<CallInst>(DepInfo.getInst())) {
       if (performCallSlotOptzn(M, M->getDest(), M->getSource(),
                                CopySize->getZExtValue(), M->getAlignment(),
                                C)) {
+        if (LT) {
+          IRBuilder<> Builder(SrcDepInfo.getInst());
+          Builder.CreateLifetimeStart(M->getDest(),
+                                      cast<ConstantInt>(LT->getArgOperand(0)));
+          MD->removeInstruction(LT);
+          LT->eraseFromParent();
+        }
+
         MD->removeInstruction(M);
         M->eraseFromParent();
         return true;
       }
     }
   }
-
-  AliasAnalysis::Location SrcLoc = AliasAnalysis::getLocationForSource(M);
-  MemDepResult SrcDepInfo = MD->getPointerDependencyFrom(SrcLoc, true,
-                                                         M, M->getParent());
   if (SrcDepInfo.isClobber()) {
     if (MemCpyInst *MDep = dyn_cast<MemCpyInst>(SrcDepInfo.getInst()))
       return processMemCpyMemCpyDependence(M, MDep, CopySize->getZExtValue());
Index: test/Transforms/MemCpyOpt/extend-lifetime.ll
===================================================================
--- /dev/null
+++ test/Transforms/MemCpyOpt/extend-lifetime.ll
@@ -0,0 +1,29 @@
+; RUN: opt < %s -basicaa -memcpyopt -S | FileCheck %s
+target datalayout = "e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+
+declare void @foo([100000 x i32]*)
+
+declare void @llvm.lifetime.start(i64, i8* nocapture)
+declare void @llvm.lifetime.end(i64, i8* nocapture)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
+
+; Function Attrs: nounwind uwtable
+define void @testfunc() {
+  %src = alloca [100000 x i32], align 4
+  %dst = alloca [100000 x i32], align 4
+  %1 = bitcast [100000 x i32]* %src to i8*
+  call void @llvm.memset.p0i8.i64(i8* %1, i8 0, i64 400000, i32 4, i1 false)
+  %2 = bitcast [100000 x i32]* %dst to i8*
+  call void @llvm.lifetime.start(i64 400000, i8* %2)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %1, i64 400000, i32 4, i1 false)
+  call void @foo([100000 x i32]* %dst)
+  call void @llvm.lifetime.end(i64 400000, i8* %2)
+  ret void
+; CHECK-LABEL: @testfunc
+; CHECK-NOT: memcpy
+; CHECK: lifetime.start
+; CHECK: memset
+}