Check alignment of loads when deciding whether it is safe to execute them

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

diff --git a/lib/Transforms/Scalar/JumpThreading.cpp b/lib/Transforms/Scalar/JumpThreading.cpp
index 953131155181c0c35c26993572ffa496db1c9094..3eff3d8d23d4d775ff1ce068ffac104de326718c 100644 (file)
--- a/lib/Transforms/Scalar/JumpThreading.cpp
+++ b/lib/Transforms/Scalar/JumpThreading.cpp
@@ -451,6 +451,12 @@ static unsigned GetBestDestForJumpOnUndef(BasicBlock *BB) {
 /// ProcessBlock - If there are any predecessors whose control can be threaded
 /// through to a successor, transform them now.
 bool JumpThreading::ProcessBlock(BasicBlock *BB) {
+  // If the block is trivially dead, just return and let the caller nuke it.
+  // This simplifies other transformations.
+  if (pred_begin(BB) == pred_end(BB) &&
+      BB != &BB->getParent()->getEntryBlock())
+    return false;
+  
   // If this block has a single predecessor, and if that pred has a single
   // successor, merge the blocks.  This encourages recursive jump threading
   // because now the condition in this block can be threaded through
@@ -1117,6 +1123,11 @@ bool JumpThreading::ProcessBranchOnXOR(BinaryOperator *BO) {
       isa<ConstantInt>(BO->getOperand(1)))
     return false;
   
+  // If the first instruction in BB isn't a phi, we won't be able to infer
+  // anything special about any particular predecessor.
+  if (!isa<PHINode>(BB->front()))
+    return false;
+  
   // If we have a xor as the branch input to this block, and we know that the
   // LHS or RHS of the xor in any predecessor is true/false, then we can clone
   // the condition into the predecessor and fix that value to true, saving some
@@ -1174,6 +1185,26 @@ bool JumpThreading::ProcessBranchOnXOR(BinaryOperator *BO) {
     BlocksToFoldInto.push_back(XorOpValues[i].second);
   }
   
+  // If we inferred a value for all of the predecessors, then duplication won't
+  // help us.  However, we can just replace the LHS or RHS with the constant.
+  if (BlocksToFoldInto.size() ==
+      cast<PHINode>(BB->front()).getNumIncomingValues()) {
+    if (SplitVal == 0) {
+      // If all preds provide undef, just nuke the xor, because it is undef too.
+      BO->replaceAllUsesWith(UndefValue::get(BO->getType()));
+      BO->eraseFromParent();
+    } else if (SplitVal->isZero()) {
+      // If all preds provide 0, replace the xor with the other input.
+      BO->replaceAllUsesWith(BO->getOperand(isLHS));
+      BO->eraseFromParent();
+    } else {
+      // If all preds provide 1, set the computed value to 1.
+      BO->setOperand(!isLHS, SplitVal);
+    }
+    
+    return true;
+  }
+  
   // Try to duplicate BB into PredBB.
   return DuplicateCondBranchOnPHIIntoPred(BB, BlocksToFoldInto);
 }
@@ -1393,9 +1424,9 @@ bool JumpThreading::DuplicateCondBranchOnPHIIntoPred(BasicBlock *BB,
   
   // Unless PredBB ends with an unconditional branch, split the edge so that we
   // can just clone the bits from BB into the end of the new PredBB.
-  BranchInst *OldPredBranch = cast<BranchInst>(PredBB->getTerminator());
+  BranchInst *OldPredBranch = dyn_cast<BranchInst>(PredBB->getTerminator());
   
-  if (!OldPredBranch->isUnconditional()) {
+  if (OldPredBranch == 0 || !OldPredBranch->isUnconditional()) {
     PredBB = SplitEdge(PredBB, BB, this);
     OldPredBranch = cast<BranchInst>(PredBB->getTerminator());
   }