[MemCpyOpt] Don't hoist store that's not guaranteed to execute

[nikic]

Summary

MemCpyOpt can hoist stores while load+store pairs into memcpy. This hoisting can currently result in stores being executed that weren't guaranteed to execute in the original problem.

[jdoerfert]

LGTM

llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp

bool MemCpyOptPass::moveUp(StoreInst *SI, Instruction *P, const LoadInst *LI) {
   // Lifted calls.
   SmallVector<const CallBase *, 8> Calls;
 
   const MemoryLocation LoadLoc = MemoryLocation::get(LI);
 
   for (auto I = --SI->getIterator(), E = P->getIterator(); I != E; --I) {
     auto *C = &*I;
 
+    // Make sure hoisting does not perform a store that was not guaranteed to
+    // happen.
+    if (!isGuaranteedToTransferExecutionToSuccessor(C))
+      return false;
+
     bool MayAlias = isModOrRefSet(AA->getModRefInfo(C, None));
 
     bool NeedLift = false;
     if (Args.erase(C))
       NeedLift = true;
     else if (MayAlias) {
       NeedLift = llvm::any_of(MemLocs, [C, this](const MemoryLocation &ML) {
         return isModOrRefSet(AA->getModRefInfo(C, ML));

llvm/test/Transforms/MemCpyOpt/fca2memcpy.ll

;
   %dstindex = or i32 %2, 1
   %dst2 = getelementptr %S , %S* %dst, i32 %dstindex
   store %S %1, %S* %dst2
   ret void
 }
 
 define void @throwing_call(%S* noalias %src, %S* %dst) {
 ; CHECK-LABEL: @throwing_call(
-; CHECK-NEXT:    [[TMP1:%.*]] = bitcast %S* [[SRC:%.*]] to i8*
-; CHECK-NEXT:    [[TMP2:%.*]] = bitcast %S* [[DST:%.*]] to i8*
-; CHECK-NEXT:    [[TMP3:%.*]] = bitcast %S* [[SRC]] to i8*
-; CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], i64 16, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0i8.i64(i8* align 8 [[TMP1]], i8 0, i64 16, i1 false)
+; CHECK-NEXT:    [[TMP1:%.*]] = load [[S:%.*]], %S* [[SRC:%.*]], align 8
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast %S* [[SRC]] to i8*
+; CHECK-NEXT:    call void @llvm.memset.p0i8.i64(i8* align 8 [[TMP2]], i8 0, i64 16, i1 false)
 ; CHECK-NEXT:    call void @call() [[ATTR2:#.*]]
+; CHECK-NEXT:    store [[S]] [[TMP1]], %S* [[DST:%.*]], align 8
 ; CHECK-NEXT:    ret void
 ;
   %1 = load %S, %S* %src
   store %S zeroinitializer, %S* %src
   call void @call() readnone
   store %S %1, %S* %dst
   ret void
 }
 
 declare void @call()