e830f5427b
The cost for scalarized operations is computed as N * (scalar operation cost + 1 extractelement + 1 insertelement). This partially fixes inflating the cost of scalarized operations since every operation is scalarized and free. I don't think we want any cost asociated with scalarization, but for now insertelement is still counted. I'm not sure if we should pretend that insertelement is also free, or add a way to compute a custom scalarization cost. llvm-svn: 254438
70 lines
2.4 KiB
C++
70 lines
2.4 KiB
C++
//===-- AMDGPUTargetTransformInfo.h - AMDGPU specific TTI -------*- C++ -*-===//
|
|
//
|
|
// The LLVM Compiler Infrastructure
|
|
//
|
|
// This file is distributed under the University of Illinois Open Source
|
|
// License. See LICENSE.TXT for details.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
/// \file
|
|
/// This file a TargetTransformInfo::Concept conforming object specific to the
|
|
/// AMDGPU 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_R600_AMDGPUTARGETTRANSFORMINFO_H
|
|
#define LLVM_LIB_TARGET_R600_AMDGPUTARGETTRANSFORMINFO_H
|
|
|
|
#include "AMDGPU.h"
|
|
#include "AMDGPUTargetMachine.h"
|
|
#include "llvm/Analysis/TargetTransformInfo.h"
|
|
#include "llvm/CodeGen/BasicTTIImpl.h"
|
|
#include "llvm/Target/TargetLowering.h"
|
|
|
|
namespace llvm {
|
|
|
|
class AMDGPUTTIImpl : public BasicTTIImplBase<AMDGPUTTIImpl> {
|
|
typedef BasicTTIImplBase<AMDGPUTTIImpl> BaseT;
|
|
typedef TargetTransformInfo TTI;
|
|
friend BaseT;
|
|
|
|
const AMDGPUSubtarget *ST;
|
|
const AMDGPUTargetLowering *TLI;
|
|
|
|
const AMDGPUSubtarget *getST() const { return ST; }
|
|
const AMDGPUTargetLowering *getTLI() const { return TLI; }
|
|
|
|
public:
|
|
explicit AMDGPUTTIImpl(const AMDGPUTargetMachine *TM, const DataLayout &DL)
|
|
: BaseT(TM, DL), ST(TM->getSubtargetImpl()),
|
|
TLI(ST->getTargetLowering()) {}
|
|
|
|
// Provide value semantics. MSVC requires that we spell all of these out.
|
|
AMDGPUTTIImpl(const AMDGPUTTIImpl &Arg)
|
|
: BaseT(static_cast<const BaseT &>(Arg)), ST(Arg.ST), TLI(Arg.TLI) {}
|
|
AMDGPUTTIImpl(AMDGPUTTIImpl &&Arg)
|
|
: BaseT(std::move(static_cast<BaseT &>(Arg))), ST(std::move(Arg.ST)),
|
|
TLI(std::move(Arg.TLI)) {}
|
|
|
|
bool hasBranchDivergence() { return true; }
|
|
|
|
void getUnrollingPreferences(Loop *L, TTI::UnrollingPreferences &UP);
|
|
|
|
TTI::PopcntSupportKind getPopcntSupport(unsigned TyWidth) {
|
|
assert(isPowerOf2_32(TyWidth) && "Ty width must be power of 2");
|
|
return ST->hasBCNT(TyWidth) ? TTI::PSK_FastHardware : TTI::PSK_Software;
|
|
}
|
|
|
|
unsigned getNumberOfRegisters(bool Vector);
|
|
unsigned getRegisterBitWidth(bool Vector);
|
|
unsigned getMaxInterleaveFactor(unsigned VF);
|
|
|
|
int getVectorInstrCost(unsigned Opcode, Type *ValTy, unsigned Index);
|
|
};
|
|
|
|
} // end namespace llvm
|
|
|
|
#endif
|