populate instcombine's initial worklist more carefully, causing
it to visit instructions from the start of the function to the
end of the function in the first path. This greatly speeds up
some pathological cases (e.g. PR5150).
Try #3, this time with some unneeded debug info stuff removed
which was causing dead pointers to be added to the worklist.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@83818 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Transforms/Scalar/InstructionCombining.cpp b/lib/Transforms/Scalar/InstructionCombining.cpp
index 0c790f6..3b6f5d4 100644
--- a/lib/Transforms/Scalar/InstructionCombining.cpp
+++ b/lib/Transforms/Scalar/InstructionCombining.cpp
@@ -101,6 +101,20 @@
Add(I);
}
+ /// AddInitialGroup - Add the specified batch of stuff in reverse order.
+ /// which should only be done when the worklist is empty and when the group
+ /// has no duplicates.
+ void AddInitialGroup(Instruction *const *List, unsigned NumEntries) {
+ assert(Worklist.empty() && "Worklist must be empty to add initial group");
+ Worklist.reserve(NumEntries+16);
+ DEBUG(errs() << "IC: ADDING: " << NumEntries << " instrs to worklist\n");
+ for (; NumEntries; --NumEntries) {
+ Instruction *I = List[NumEntries-1];
+ WorklistMap.insert(std::make_pair(I, Worklist.size()));
+ Worklist.push_back(I);
+ }
+ }
+
// Remove - remove I from the worklist if it exists.
void Remove(Instruction *I) {
DenseMap<Instruction*, unsigned>::iterator It = WorklistMap.find(I);
@@ -12663,6 +12677,9 @@
const TargetData *TD) {
SmallVector<BasicBlock*, 256> Worklist;
Worklist.push_back(BB);
+
+ std::vector<Instruction*> InstrsForInstCombineWorklist;
+ InstrsForInstCombineWorklist.reserve(128);
while (!Worklist.empty()) {
BB = Worklist.back();
@@ -12671,7 +12688,6 @@
// We have now visited this block! If we've already been here, ignore it.
if (!Visited.insert(BB)) continue;
- DbgInfoIntrinsic *DBI_Prev = NULL;
for (BasicBlock::iterator BBI = BB->begin(), E = BB->end(); BBI != E; ) {
Instruction *Inst = BBI++;
@@ -12692,24 +12708,8 @@
Inst->eraseFromParent();
continue;
}
-
- // If there are two consecutive llvm.dbg.stoppoint calls then
- // it is likely that the optimizer deleted code in between these
- // two intrinsics.
- DbgInfoIntrinsic *DBI_Next = dyn_cast<DbgInfoIntrinsic>(Inst);
- if (DBI_Next) {
- if (DBI_Prev
- && DBI_Prev->getIntrinsicID() == llvm::Intrinsic::dbg_stoppoint
- && DBI_Next->getIntrinsicID() == llvm::Intrinsic::dbg_stoppoint) {
- IC.Worklist.Remove(DBI_Prev);
- DBI_Prev->eraseFromParent();
- }
- DBI_Prev = DBI_Next;
- } else {
- DBI_Prev = 0;
- }
- IC.Worklist.Add(Inst);
+ InstrsForInstCombineWorklist.push_back(Inst);
}
// Recursively visit successors. If this is a branch or switch on a
@@ -12741,6 +12741,14 @@
for (unsigned i = 0, e = TI->getNumSuccessors(); i != e; ++i)
Worklist.push_back(TI->getSuccessor(i));
}
+
+ // Once we've found all of the instructions to add to instcombine's worklist,
+ // add them in reverse order. This way instcombine will visit from the top
+ // of the function down. This jives well with the way that it adds all uses
+ // of instructions to the worklist after doing a transformation, thus avoiding
+ // some N^2 behavior in pathological cases.
+ IC.Worklist.AddInitialGroup(&InstrsForInstCombineWorklist[0],
+ InstrsForInstCombineWorklist.size());
}
bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {