Fix a FIXME about the format and add a test.

[oota-llvm.git] / lib / Transforms / Scalar / MemCpyOptimizer.cpp

diff --git a/lib/Transforms/Scalar/MemCpyOptimizer.cpp b/lib/Transforms/Scalar/MemCpyOptimizer.cpp
index 97fff9edd68f77233c9278ae47528db861c0207e..8f61ffd84bfcb541369ccb88fddd3eff6c0f3ca1 100644 (file)
--- a/lib/Transforms/Scalar/MemCpyOptimizer.cpp
+++ b/lib/Transforms/Scalar/MemCpyOptimizer.cpp
@@ -14,20 +14,20 @@
 
 #define DEBUG_TYPE "memcpyopt"
 #include "llvm/Transforms/Scalar.h"
-#include "llvm/GlobalVariable.h"
-#include "llvm/IRBuilder.h"
-#include "llvm/Instructions.h"
-#include "llvm/IntrinsicInst.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/Analysis/AliasAnalysis.h"
 #include "llvm/Analysis/Dominators.h"
 #include "llvm/Analysis/MemoryDependenceAnalysis.h"
 #include "llvm/Analysis/ValueTracking.h"
+#include "llvm/IR/DataLayout.h"
+#include "llvm/IR/GlobalVariable.h"
+#include "llvm/IR/IRBuilder.h"
+#include "llvm/IR/Instructions.h"
+#include "llvm/IR/IntrinsicInst.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/GetElementPtrTypeIterator.h"
 #include "llvm/Support/raw_ostream.h"
-#include "llvm/DataLayout.h"
 #include "llvm/Target/TargetLibraryInfo.h"
 #include "llvm/Transforms/Utils/Local.h"
 #include <list>
@@ -174,11 +174,10 @@ bool MemsetRange::isProfitableToUseMemset(const DataLayout &TD) const {
   // this width can be stored.  If so, check to see whether we will end up
   // actually reducing the number of stores used.
   unsigned Bytes = unsigned(End-Start);
-  unsigned AS = cast<StoreInst>(TheStores[0])->getPointerAddressSpace();
-  unsigned NumPointerStores = Bytes/TD.getPointerSize(AS);
+  unsigned NumPointerStores = Bytes/TD.getPointerSize();
 
   // Assume the remaining bytes if any are done a byte at a time.
-  unsigned NumByteStores = Bytes - NumPointerStores*TD.getPointerSize(AS);
+  unsigned NumByteStores = Bytes - NumPointerStores*TD.getPointerSize();
 
   // If we will reduce the # stores (according to this heuristic), do the
   // transformation.  This encourages merging 4 x i8 -> i32 and 2 x i16 -> i32
@@ -466,7 +465,7 @@ Instruction *MemCpyOpt::tryMergingIntoMemset(Instruction *StartInst,
       AMemSet->setDebugLoc(Range.TheStores[0]->getDebugLoc());
 
     // Zap all the stores.
-    for (SmallVector<Instruction*, 16>::const_iterator
+    for (SmallVectorImpl<Instruction *>::const_iterator
          SI = Range.TheStores.begin(),
          SE = Range.TheStores.end(); SI != SE; ++SI) {
       MD->removeInstruction(*SI);
@@ -627,8 +626,14 @@ bool MemCpyOpt::performCallSlotOptzn(Instruction *cpy,
       return false;
 
     Type *StructTy = cast<PointerType>(A->getType())->getElementType();
-    uint64_t destSize = TD->getTypeAllocSize(StructTy);
+    if (!StructTy->isSized()) {
+      // The call may never return and hence the copy-instruction may never
+      // be executed, and therefore it's not safe to say "the destination
+      // has at least <cpyLen> bytes, as implied by the copy-instruction",
+      return false;
+    }
 
+    uint64_t destSize = TD->getTypeAllocSize(StructTy);
     if (destSize < srcSize)
       return false;
   } else {