[ARM] Allow signed icmps in ARMCodeGenPrepare
Originally committed in r339755 which was reverted in r339806 due to
an asan issue. The issue was caused by my assumption that operands to
a CallInst mapped to the FunctionType Params. CallInsts are now
handled by iterating over their ArgOperands instead of Operands.
Original Message:
Treat signed icmps as 'sinks', allowing them to be in the use-def
tree, enabling more promotions to be performed. As a sink, any
promoted incoming values need to be truncated before being used by
the signed icmp.
Differential Revision: https://reviews.llvm.org/D50067
llvm-svn: 339858
diff --git a/llvm/lib/Target/ARM/ARMCodeGenPrepare.cpp b/llvm/lib/Target/ARM/ARMCodeGenPrepare.cpp
index 99c89eb..164b255 100644
--- a/llvm/lib/Target/ARM/ARMCodeGenPrepare.cpp
+++ b/llvm/lib/Target/ARM/ARMCodeGenPrepare.cpp
@@ -181,6 +181,8 @@
return UsesNarrowValue(Return->getReturnValue());
if (auto *Trunc = dyn_cast<TruncInst>(V))
return UsesNarrowValue(Trunc->getOperand(0));
+ if (auto *ICmp = dyn_cast<ICmpInst>(V))
+ return ICmp->isSigned();
return isa<CallInst>(V);
}
@@ -294,6 +296,11 @@
LLVM_DEBUG(dbgs() << "ARM CGP: Promoting use-def chains to from "
<< ARMCodeGenPrepare::TypeSize << " to 32-bits\n");
+ // Cache original types.
+ DenseMap<Value*, Type*> TruncTysMap;
+ for (auto *V : Visited)
+ TruncTysMap[V] = V->getType();
+
auto ReplaceAllUsersOfWith = [&](Value *From, Value *To) {
SmallVector<Instruction*, 4> Users;
Instruction *InstTo = dyn_cast<Instruction>(To);
@@ -337,6 +344,7 @@
ReplaceAllUsersOfWith(I, Call);
InstsToRemove.push_back(I);
NewInsts.insert(Call);
+ TruncTysMap[Call] = OrigTy;
};
auto InsertZExt = [&](Value *V, Instruction *InsertPt) {
@@ -351,6 +359,7 @@
ZExt->moveAfter(InsertPt);
ReplaceAllUsersOfWith(V, ZExt);
NewInsts.insert(ZExt);
+ TruncTysMap[ZExt] = TruncTysMap[V];
};
// First, insert extending instructions between the leaves and their users.
@@ -409,42 +418,48 @@
InsertDSPIntrinsic(cast<Instruction>(V));
}
+ auto InsertTrunc = [&](Value *V) -> Instruction* {
+ if (!isa<Instruction>(V) || !isa<IntegerType>(V->getType()))
+ return nullptr;
+
+ if ((!Promoted.count(V) && !NewInsts.count(V)) || !TruncTysMap.count(V))
+ return nullptr;
+
+ Type *TruncTy = TruncTysMap[V];
+ if (TruncTy == ExtTy)
+ return nullptr;
+
+ LLVM_DEBUG(dbgs() << "ARM CGP: Creating " << *TruncTy << " Trunc for "
+ << *V << "\n");
+ Builder.SetInsertPoint(cast<Instruction>(V));
+ auto *Trunc = cast<Instruction>(Builder.CreateTrunc(V, TruncTy));
+ NewInsts.insert(Trunc);
+ return Trunc;
+ };
+
LLVM_DEBUG(dbgs() << "ARM CGP: Fixing up the roots:\n");
// Fix up any stores or returns that use the results of the promoted
// chain.
for (auto I : Roots) {
LLVM_DEBUG(dbgs() << " - " << *I << "\n");
- Type *TruncTy = OrigTy;
- if (auto *Store = dyn_cast<StoreInst>(I)) {
- auto *PtrTy = cast<PointerType>(Store->getPointerOperandType());
- TruncTy = PtrTy->getElementType();
- } else if (isa<ReturnInst>(I)) {
- Function *F = I->getParent()->getParent();
- TruncTy = F->getFunctionType()->getReturnType();
+
+ // Handle calls separately as we need to iterate over arg operands.
+ if (auto *Call = dyn_cast<CallInst>(I)) {
+ for (unsigned i = 0; i < Call->getNumArgOperands(); ++i) {
+ Value *Arg = Call->getArgOperand(i);
+ if (Instruction *Trunc = InsertTrunc(Arg)) {
+ Trunc->moveBefore(Call);
+ Call->setArgOperand(i, Trunc);
+ }
+ }
+ continue;
}
+ // Now handle the others.
for (unsigned i = 0; i < I->getNumOperands(); ++i) {
- Value *V = I->getOperand(i);
- if (!isa<IntegerType>(V->getType()))
- continue;
-
- if (Promoted.count(V) || NewInsts.count(V)) {
- if (auto *Op = dyn_cast<Instruction>(V)) {
-
- if (auto *Call = dyn_cast<CallInst>(I))
- TruncTy = Call->getFunctionType()->getParamType(i);
-
- if (TruncTy == ExtTy)
- continue;
-
- LLVM_DEBUG(dbgs() << "ARM CGP: Creating " << *TruncTy
- << " Trunc for " << *Op << "\n");
- Builder.SetInsertPoint(Op);
- auto *Trunc = cast<Instruction>(Builder.CreateTrunc(Op, TruncTy));
- Trunc->moveBefore(I);
- I->setOperand(i, Trunc);
- NewInsts.insert(Trunc);
- }
+ if (Instruction *Trunc = InsertTrunc(I->getOperand(i))) {
+ Trunc->moveBefore(I);
+ I->setOperand(i, Trunc);
}
}
}
@@ -458,8 +473,8 @@
bool ARMCodeGenPrepare::isSupportedValue(Value *V) {
LLVM_DEBUG(dbgs() << "ARM CGP: Is " << *V << " supported?\n");
- if (auto *ICmp = dyn_cast<ICmpInst>(V))
- return ICmp->isEquality() || !ICmp->isSigned();
+ if (isa<ICmpInst>(V))
+ return true;
// Memory instructions
if (isa<StoreInst>(V) || isa<GetElementPtrInst>(V))