NVPTXTargetTransformInfo.h

Go to the documentation of this file.

//===-- NVPTXTargetTransformInfo.h - NVPTX specific TTI ---------*- C++ -*-===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
/// \file
/// This file a TargetTransformInfoImplBase conforming object specific to the
/// NVPTX target machine. It uses the target's detailed information to
/// provide more precise answers to certain TTI queries, while letting the
/// target independent and default TTI implementations handle the rest.
///
//===----------------------------------------------------------------------===//
 
#ifndef LLVM_LIB_TARGET_NVPTX_NVPTXTARGETTRANSFORMINFO_H
#define LLVM_LIB_TARGET_NVPTX_NVPTXTARGETTRANSFORMINFO_H
 
#include "MCTargetDesc/NVPTXBaseInfo.h"
#include "NVPTXTargetMachine.h"
#include "NVPTXUtilities.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/CodeGen/BasicTTIImpl.h"
#include "llvm/CodeGen/TargetLowering.h"
#include <optional>
 
namespace llvm {
 
class NVPTXTTIImpl final : public BasicTTIImplBase<NVPTXTTIImpl> {
  typedef BasicTTIImplBase<NVPTXTTIImpl> BaseT;
  typedef TargetTransformInfo TTI;
  friend BaseT;
 
  const NVPTXSubtarget *ST;
  const NVPTXTargetLowering *TLI;
 
  const NVPTXSubtarget *getST() const { return ST; };
  const NVPTXTargetLowering *getTLI() const { return TLI; };
 
  /// \returns true if the result of the value could potentially be
  /// different across threads in a warp.
  bool isSourceOfDivergence(const Value *V) const;
 
public:
  explicit NVPTXTTIImpl(const NVPTXTargetMachine *TM, const Function &F)
      : BaseT(TM, F.getDataLayout()), ST(TM->getSubtargetImpl()),
        TLI(ST->getTargetLowering()) {}
 
  bool hasBranchDivergence(const Function *F = nullptr) const override {
    return true;
  }
 
  unsigned getFlatAddressSpace() const override {
    return AddressSpace::ADDRESS_SPACE_GENERIC;
  }
 
  unsigned getAddressSpaceJoin(unsigned AS1, unsigned AS2) const override {
    if ((AS1 == AddressSpace::ADDRESS_SPACE_SHARED &&
         AS2 == AddressSpace::ADDRESS_SPACE_SHARED_CLUSTER) ||
        (AS2 == AddressSpace::ADDRESS_SPACE_SHARED &&
         AS1 == AddressSpace::ADDRESS_SPACE_SHARED_CLUSTER))
      return AddressSpace::ADDRESS_SPACE_SHARED_CLUSTER;
    return AddressSpace::ADDRESS_SPACE_GENERIC;
  }
 
  bool
  canHaveNonUndefGlobalInitializerInAddressSpace(unsigned AS) const override {
    return AS != AddressSpace::ADDRESS_SPACE_SHARED &&
           AS != AddressSpace::ADDRESS_SPACE_LOCAL &&
           AS != AddressSpace::ADDRESS_SPACE_ENTRY_PARAM;
  }
 
  std::optional<Instruction *>
  instCombineIntrinsic(InstCombiner &IC, IntrinsicInst &II) const override;
 
  // Loads and stores can be vectorized if the alignment is at least as big as
  // the load/store we want to vectorize.
  bool isLegalToVectorizeLoadChain(unsigned ChainSizeInBytes, Align Alignment,
                                   unsigned AddrSpace) const override {
    return Alignment >= ChainSizeInBytes;
  }
  bool isLegalToVectorizeStoreChain(unsigned ChainSizeInBytes, Align Alignment,
                                    unsigned AddrSpace) const override {
    return isLegalToVectorizeLoadChain(ChainSizeInBytes, Alignment, AddrSpace);
  }
 
  // NVPTX has infinite registers of all kinds, but the actual machine doesn't.
  // We conservatively return 1 here which is just enough to enable the
  // vectorizers but disables heuristics based on the number of registers.
  // FIXME: Return a more reasonable number, while keeping an eye on
  // LoopVectorizer's unrolling heuristics.
  unsigned getNumberOfRegisters(unsigned ClassID) const override { return 1; }
 
  // Only <2 x half> should be vectorized, so always return 32 for the vector
  // register size.
  TypeSize
  getRegisterBitWidth(TargetTransformInfo::RegisterKind K) const override {
    return TypeSize::getFixed(32);
  }
  unsigned getMinVectorRegisterBitWidth() const override { return 32; }
 
  bool shouldExpandReduction(const IntrinsicInst *II) const override {
    // Turn off ExpandReductions pass for NVPTX, which doesn't have advanced
    // swizzling operations. Our backend/Selection DAG can expand these
    // reductions with less movs.
    return false;
  }
 
  // We don't want to prevent inlining because of target-cpu and -features
  // attributes that were added to newer versions of LLVM/Clang: There are
  // no incompatible functions in PTX, ptxas will throw errors in such cases.
  bool areInlineCompatible(const Function *Caller,
                           const Function *Callee) const override {
    return true;
  }
 
  // Increase the inlining cost threshold by a factor of 11, reflecting that
  // calls are particularly expensive in NVPTX.
  unsigned getInliningThresholdMultiplier() const override { return 11; }
 
  InstructionCost
  getInstructionCost(const User *U, ArrayRef<const Value *> Operands,
                     TTI::TargetCostKind CostKind) const override;
 
  InstructionCost getArithmeticInstrCost(
      unsigned Opcode, Type *Ty, TTI::TargetCostKind CostKind,
      TTI::OperandValueInfo Op1Info = {TTI::OK_AnyValue, TTI::OP_None},
      TTI::OperandValueInfo Op2Info = {TTI::OK_AnyValue, TTI::OP_None},
      ArrayRef<const Value *> Args = {},
      const Instruction *CxtI = nullptr) const override;
 
  InstructionCost
  getScalarizationOverhead(VectorType *InTy, const APInt &DemandedElts,
                           bool Insert, bool Extract,
                           TTI::TargetCostKind CostKind,
                           bool ForPoisonSrc = true, ArrayRef<Value *> VL = {},
                           TTI::VectorInstrContext VIC =
                               TTI::VectorInstrContext::None) const override {
    if (!InTy->getElementCount().isFixed())
      return InstructionCost::getInvalid();
 
    auto VT = getTLI()->getValueType(DL, InTy);
    auto NumElements = InTy->getElementCount().getFixedValue();
    InstructionCost Cost = 0;
    if (Insert && !VL.empty()) {
      bool AllConstant = all_of(seq(NumElements), [&](int Idx) {
        return !DemandedElts[Idx] || isa<Constant>(VL[Idx]);
      });
      if (AllConstant) {
        Cost += TTI::TCC_Free;
        Insert = false;
      }
    }
    if (Insert && NVPTX::isPackedVectorTy(VT) && VT.is32BitVector()) {
      // Can be built in a single 32-bit mov (64-bit regs are emulated in SASS
      // with 2x 32-bit regs)
      Cost += 1;
      Insert = false;
    }
    if (Insert && VT == MVT::v4i8) {
      InstructionCost Cost = 3; // 3 x PRMT
      for (auto Idx : seq(NumElements))
        if (DemandedElts[Idx])
          Cost += 1; // zext operand to i32
      Insert = false;
    }
    return Cost + BaseT::getScalarizationOverhead(InTy, DemandedElts, Insert,
                                                  Extract, CostKind,
                                                  ForPoisonSrc, VL);
  }
 
  void getUnrollingPreferences(Loop *L, ScalarEvolution &SE,
                               TTI::UnrollingPreferences &UP,
                               OptimizationRemarkEmitter *ORE) const override;
 
  void getPeelingPreferences(Loop *L, ScalarEvolution &SE,
                             TTI::PeelingPreferences &PP) const override;
 
  bool hasVolatileVariant(Instruction *I, unsigned AddrSpace) const override {
    // Volatile loads/stores are only supported for shared and global address
    // spaces, or for generic AS that maps to them.
    if (!(AddrSpace == llvm::ADDRESS_SPACE_GENERIC ||
          AddrSpace == llvm::ADDRESS_SPACE_GLOBAL ||
          AddrSpace == llvm::ADDRESS_SPACE_SHARED))
      return false;
 
    switch(I->getOpcode()){
    default:
      return false;
    case Instruction::Load:
    case Instruction::Store:
      return true;
    }
  }
 
  APInt getAddrSpaceCastPreservedPtrMask(unsigned SrcAS,
                                         unsigned DstAS) const override {
    if (SrcAS != llvm::ADDRESS_SPACE_GENERIC)
      return BaseT::getAddrSpaceCastPreservedPtrMask(SrcAS, DstAS);
    if (DstAS != llvm::ADDRESS_SPACE_GLOBAL &&
        DstAS != llvm::ADDRESS_SPACE_SHARED)
      return BaseT::getAddrSpaceCastPreservedPtrMask(SrcAS, DstAS);
 
    // Address change within 4K size does not change the original address space
    // and is safe to perform address cast form SrcAS to DstAS.
    APInt PtrMask(DL.getPointerSizeInBits(llvm::ADDRESS_SPACE_GENERIC), 0xfff);
    return PtrMask;
  }
 
  bool collectFlatAddressOperands(SmallVectorImpl<int> &OpIndexes,
                                  Intrinsic::ID IID) const override;
 
  bool isLegalMaskedStore(Type *DataType, Align Alignment, unsigned AddrSpace,
                          TTI::MaskKind MaskKind) const override;
 
  bool isLegalMaskedLoad(Type *DataType, Align Alignment, unsigned AddrSpace,
                         TTI::MaskKind MaskKind) const override;
 
  unsigned getLoadStoreVecRegBitWidth(unsigned AddrSpace) const override;
 
  Value *rewriteIntrinsicWithAddressSpace(IntrinsicInst *II, Value *OldV,
                                          Value *NewV) const override;
  unsigned getAssumedAddrSpace(const Value *V) const override;
 
  void collectKernelLaunchBounds(
      const Function &F,
      SmallVectorImpl<std::pair<StringRef, int64_t>> &LB) const override;
 
  bool shouldBuildRelLookupTables() const override {
    // Self-referential globals are not supported.
    return false;
  }
 
  InstructionCost getPartialReductionCost(
      unsigned Opcode, Type *InputTypeA, Type *InputTypeB, Type *AccumType,
      ElementCount VF, TTI::PartialReductionExtendKind OpAExtend,
      TTI::PartialReductionExtendKind OpBExtend, std::optional<unsigned> BinOp,
      TTI::TargetCostKind CostKind,
      std::optional<FastMathFlags> FMF) const override {
    return InstructionCost::getInvalid();
  }
 
  ValueUniformity getValueUniformity(const Value *V) const override;
};
 
} // end namespace llvm
 
#endif