Type *RetTy = I->getType();
if (canTruncateToMinimalBitwidth(I, VF))
RetTy = IntegerType::get(RetTy->getContext(), MinBWs[I]);
- VectorTy = isScalarAfterVectorization(I, VF) ? RetTy : ToVectorTy(RetTy, VF);
auto SE = PSE.getSE();
TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput;
+ auto hasSingleCopyAfterVectorization = [this](Instruction *I,
+ ElementCount VF) -> bool {
+ if (VF.isScalar())
+ return true;
+
+ auto Scalarized = InstsToScalarize.find(VF);
+ assert(Scalarized != InstsToScalarize.end() &&
+ "VF not yet analyzed for scalarization profitability");
+ return !Scalarized->second.count(I) &&
+ llvm::all_of(I->users(), [&](User *U) {
+ auto *UI = cast<Instruction>(U);
+ return !Scalarized->second.count(UI);
+ });
+ };
+
+ if (isScalarAfterVectorization(I, VF)) {
+ // With the exception of GEPs and PHIs, after scalarization there should
+ // only be one copy of the instruction generated in the loop. This is
+ // because the VF is either 1, or any instructions that need scalarizing
+ // have already been dealt with by the the time we get here. As a result,
+ // it means we don't have to multiply the instruction cost by VF.
+ assert(I->getOpcode() == Instruction::GetElementPtr ||
+ I->getOpcode() == Instruction::PHI ||
+ (I->getOpcode() == Instruction::BitCast &&
+ I->getType()->isPointerTy()) ||
+ hasSingleCopyAfterVectorization(I, VF));
+ VectorTy = RetTy;
+ } else
+ VectorTy = ToVectorTy(RetTy, VF);
+
// TODO: We need to estimate the cost of intrinsic calls.
switch (I->getOpcode()) {
case Instruction::GetElementPtr:
Op2VK = TargetTransformInfo::OK_UniformValue;
SmallVector<const Value *, 4> Operands(I->operand_values());
- unsigned N = isScalarAfterVectorization(I, VF) ? VF.getKnownMinValue() : 1;
- return N * TTI.getArithmeticInstrCost(
- I->getOpcode(), VectorTy, CostKind,
- TargetTransformInfo::OK_AnyValue,
- Op2VK, TargetTransformInfo::OP_None, Op2VP, Operands, I);
+ return TTI.getArithmeticInstrCost(
+ I->getOpcode(), VectorTy, CostKind, TargetTransformInfo::OK_AnyValue,
+ Op2VK, TargetTransformInfo::OP_None, Op2VP, Operands, I);
}
case Instruction::FNeg: {
assert(!VF.isScalable() && "VF is assumed to be non scalable.");
- unsigned N = isScalarAfterVectorization(I, VF) ? VF.getKnownMinValue() : 1;
- return N * TTI.getArithmeticInstrCost(
- I->getOpcode(), VectorTy, CostKind,
- TargetTransformInfo::OK_AnyValue,
- TargetTransformInfo::OK_AnyValue,
- TargetTransformInfo::OP_None, TargetTransformInfo::OP_None,
- I->getOperand(0), I);
+ return TTI.getArithmeticInstrCost(
+ I->getOpcode(), VectorTy, CostKind, TargetTransformInfo::OK_AnyValue,
+ TargetTransformInfo::OK_AnyValue, TargetTransformInfo::OP_None,
+ TargetTransformInfo::OP_None, I->getOperand(0), I);
}
case Instruction::Select: {
SelectInst *SI = cast<SelectInst>(I);
VectorTy = ToVectorTy(getMemInstValueType(I), Width);
return getMemoryInstructionCost(I, VF);
}
+ case Instruction::BitCast:
+ if (I->getType()->isPointerTy())
+ return 0;
+ LLVM_FALLTHROUGH;
case Instruction::ZExt:
case Instruction::SExt:
case Instruction::FPToUI:
case Instruction::SIToFP:
case Instruction::UIToFP:
case Instruction::Trunc:
- case Instruction::FPTrunc:
- case Instruction::BitCast: {
+ case Instruction::FPTrunc: {
// Computes the CastContextHint from a Load/Store instruction.
auto ComputeCCH = [&](Instruction *I) -> TTI::CastContextHint {
assert((isa<LoadInst>(I) || isa<StoreInst>(I)) &&
}
}
- unsigned N;
- if (isScalarAfterVectorization(I, VF)) {
- assert(!VF.isScalable() && "VF is assumed to be non scalable");
- N = VF.getKnownMinValue();
- } else
- N = 1;
- return N *
- TTI.getCastInstrCost(Opcode, VectorTy, SrcVecTy, CCH, CostKind, I);
+ return TTI.getCastInstrCost(Opcode, VectorTy, SrcVecTy, CCH, CostKind, I);
}
case Instruction::Call: {
bool NeedToScalarize;
case Instruction::ExtractValue:
return TTI.getInstructionCost(I, TTI::TCK_RecipThroughput);
default:
- // The cost of executing VF copies of the scalar instruction. This opcode
- // is unknown. Assume that it is the same as 'mul'.
- return VF.getKnownMinValue() * TTI.getArithmeticInstrCost(
- Instruction::Mul, VectorTy, CostKind) +
- getScalarizationOverhead(I, VF);
+ // This opcode is unknown. Assume that it is the same as 'mul'.
+ return TTI.getArithmeticInstrCost(Instruction::Mul, VectorTy, CostKind);
} // end of switch.
}
ret void
}
+; CHECK-LABEL: predicated_store_phi
+;
+; Same as predicate_store except we use a pointer PHI to maintain the address
+;
+; CHECK: Found new scalar instruction: %addr = phi i32* [ %a, %entry ], [ %addr.next, %for.inc ]
+; CHECK: Found new scalar instruction: %addr.next = getelementptr inbounds i32, i32* %addr, i64 1
+; CHECK: Scalarizing and predicating: store i32 %tmp2, i32* %addr, align 4
+; CHECK: Found an estimated cost of 0 for VF 2 For instruction: %addr = phi i32* [ %a, %entry ], [ %addr.next, %for.inc ]
+; CHECK: Found an estimated cost of 3 for VF 2 For instruction: store i32 %tmp2, i32* %addr, align 4
+;
+define void @predicated_store_phi(i32* %a, i1 %c, i32 %x, i64 %n) {
+entry:
+ br label %for.body
+
+for.body:
+ %i = phi i64 [ 0, %entry ], [ %i.next, %for.inc ]
+ %addr = phi i32 * [ %a, %entry ], [ %addr.next, %for.inc ]
+ %tmp1 = load i32, i32* %addr, align 4
+ %tmp2 = add nsw i32 %tmp1, %x
+ br i1 %c, label %if.then, label %for.inc
+
+if.then:
+ store i32 %tmp2, i32* %addr, align 4
+ br label %for.inc
+
+for.inc:
+ %i.next = add nuw nsw i64 %i, 1
+ %cond = icmp slt i64 %i.next, %n
+ %addr.next = getelementptr inbounds i32, i32* %addr, i64 1
+ br i1 %cond, label %for.body, label %for.end
+
+for.end:
+ ret void
+}
+
; CHECK-LABEL: predicated_udiv_scalarized_operand
;
; This test checks that we correctly compute the cost of the predicated udiv
--- /dev/null
+; REQUIRES: asserts
+; RUN: opt -loop-vectorize -force-vector-width=2 -debug-only=loop-vectorize -S -o - < %s 2>&1 | FileCheck %s
+
+%struct.foo = type { i32, i64 }
+
+; CHECK: LV: Found an estimated cost of 0 for VF 2 For instruction: %0 = bitcast i64* %b to i32*
+
+; The bitcast below will be scalarized due to the predication in the loop. Bitcasts
+; between pointer types should be treated as free, despite the scalarization.
+define void @foo(%struct.foo* noalias nocapture %in, i32* noalias nocapture readnone %out, i64 %n) {
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %if.end
+ %i.012 = phi i64 [ %inc, %if.end ], [ 0, %entry ]
+ %b = getelementptr inbounds %struct.foo, %struct.foo* %in, i64 %i.012, i32 1
+ %0 = bitcast i64* %b to i32*
+ %a = getelementptr inbounds %struct.foo, %struct.foo* %in, i64 %i.012, i32 0
+ %1 = load i32, i32* %a, align 8
+ %tobool.not = icmp eq i32 %1, 0
+ br i1 %tobool.not, label %if.end, label %land.lhs.true
+
+land.lhs.true: ; preds = %for.body
+ %2 = load i32, i32* %0, align 4
+ %cmp2 = icmp sgt i32 %2, 0
+ br i1 %cmp2, label %if.then, label %if.end
+
+if.then: ; preds = %land.lhs.true
+ %sub = add nsw i32 %2, -1
+ store i32 %sub, i32* %0, align 4
+ br label %if.end
+
+if.end: ; preds = %if.then, %land.lhs.true, %for.body
+ %inc = add nuw nsw i64 %i.012, 1
+ %exitcond.not = icmp eq i64 %inc, %n
+ br i1 %exitcond.not, label %for.end, label %for.body
+
+for.end: ; preds = %if.end
+ ret void
+}