shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/llvm/lib/Target/AMDGPU/GCNRegPressure.cpp
Juan Manuel Martinez Caamaño c30c2f4f3e
[AMDGPU] Rematerialize VGPR candidates when SGPR spills results in VGPR Excess (#168079) 
Before, when selecting candidates to rematerialize, we would only
consider SGPR candidates when there was an excess of SGPR registers.

Failing to eliminate the excess would result in spills to VGPRs.
This is normally not an issue, unless spilling to VGPRs results in
excess VGPRs.

This patch does 2 things:
* It relaxes the GCNRPTarget success criteria: now we accept regions
  where we spill SGPRs to VGPRs, as long as this does not end up in
  excess VGPRs.
* It changes isSaveBeneficial to consider the excess VGPRs (which
  includes the SGPRs that would be spilled to VGPR).

With these changes, the compiler rematerializes VGPRs when the excess
SGPRs would result in VGPR excess.

This has some unaddressed flaws: we should attempt to rematerialize
SGPRs
first in order to eliminate the SGPR excess that results in VGPR excess.

Related to SWDEV-549940
2026-01-16 09:08:55 +01:00

1123 lines
39 KiB
C++
Raw Blame History

//===- GCNRegPressure.cpp -------------------------------------------------===//
//
// 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 implements the GCNRegPressure class.
///
//===----------------------------------------------------------------------===//
#include "GCNRegPressure.h"
#include "AMDGPU.h"
#include "SIMachineFunctionInfo.h"
#include "llvm/CodeGen/MachineLoopInfo.h"
#include "llvm/CodeGen/RegisterPressure.h"
using namespace llvm;
#define DEBUG_TYPE "machine-scheduler"
bool llvm::isEqual(const GCNRPTracker::LiveRegSet &S1,
const GCNRPTracker::LiveRegSet &S2) {
if (S1.size() != S2.size())
return false;
for (const auto &P : S1) {
auto I = S2.find(P.first);
if (I == S2.end() || I->second != P.second)
return false;
}
return true;
}
///////////////////////////////////////////////////////////////////////////////
// GCNRegPressure
unsigned GCNRegPressure::getRegKind(const TargetRegisterClass *RC,
const SIRegisterInfo *STI) {
return STI->isSGPRClass(RC)
? SGPR
: (STI->isAGPRClass(RC)
? AGPR
: (STI->isVectorSuperClass(RC) ? AVGPR : VGPR));
}
void GCNRegPressure::inc(unsigned Reg,
LaneBitmask PrevMask,
LaneBitmask NewMask,
const MachineRegisterInfo &MRI) {
unsigned NewNumCoveredRegs = SIRegisterInfo::getNumCoveredRegs(NewMask);
unsigned PrevNumCoveredRegs = SIRegisterInfo::getNumCoveredRegs(PrevMask);
if (NewNumCoveredRegs == PrevNumCoveredRegs)
return;
int Sign = 1;
if (NewMask < PrevMask) {
std::swap(NewMask, PrevMask);
std::swap(NewNumCoveredRegs, PrevNumCoveredRegs);
Sign = -1;
}
assert(PrevMask < NewMask && PrevNumCoveredRegs < NewNumCoveredRegs &&
"prev mask should always be lesser than new");
const TargetRegisterClass *RC = MRI.getRegClass(Reg);
const TargetRegisterInfo *TRI = MRI.getTargetRegisterInfo();
const SIRegisterInfo *STI = static_cast<const SIRegisterInfo *>(TRI);
unsigned RegKind = getRegKind(RC, STI);
if (TRI->getRegSizeInBits(*RC) != 32) {
// Reg is from a tuple register class.
if (PrevMask.none()) {
unsigned TupleIdx = TOTAL_KINDS + RegKind;
Value[TupleIdx] += Sign * TRI->getRegClassWeight(RC).RegWeight;
}
// Pressure scales with number of new registers covered by the new mask.
// Note when true16 is enabled, we can no longer safely use the following
// approach to calculate the difference in the number of 32-bit registers
// between two masks:
//
// Sign *= SIRegisterInfo::getNumCoveredRegs(~PrevMask & NewMask);
//
// The issue is that the mask calculation `~PrevMask & NewMask` doesn't
// properly account for partial usage of a 32-bit register when dealing with
// 16-bit registers.
//
// Consider this example:
// Assume PrevMask = 0b0010 and NewMask = 0b1111. Here, the correct register
// usage difference should be 1, because even though PrevMask uses only half
// of a 32-bit register, it should still be counted as a full register use.
// However, the mask calculation yields `~PrevMask & NewMask = 0b1101`, and
// calling `getNumCoveredRegs` returns 2 instead of 1. This incorrect
// calculation can lead to integer overflow when Sign = -1.
Sign *= NewNumCoveredRegs - PrevNumCoveredRegs;
}
Value[RegKind] += Sign;
}
namespace {
struct RegExcess {
unsigned SGPR = 0;
unsigned VGPR = 0;
unsigned ArchVGPR = 0;
unsigned AGPR = 0;
bool anyExcess() const { return SGPR || VGPR || ArchVGPR || AGPR; }
bool hasVectorRegisterExcess() const { return VGPR || ArchVGPR || AGPR; }
RegExcess(const MachineFunction &MF, const GCNRegPressure &RP)
: RegExcess(MF, RP, GCNRPTarget(MF, RP)) {}
RegExcess(const MachineFunction &MF, const GCNRegPressure &RP,
const GCNRPTarget &Target) {
unsigned MaxSGPRs = Target.getMaxSGPRs();
unsigned MaxVGPRs = Target.getMaxVGPRs();
const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
SGPR = std::max(static_cast<int>(RP.getSGPRNum() - MaxSGPRs), 0);
// The number of virtual VGPRs required to handle excess SGPR
unsigned WaveSize = ST.getWavefrontSize();
unsigned VGPRForSGPRSpills = divideCeil(SGPR, WaveSize);
unsigned MaxArchVGPRs = ST.getAddressableNumArchVGPRs();
// Unified excess pressure conditions, accounting for VGPRs used for SGPR
// spills
VGPR = std::max(static_cast<int>(RP.getVGPRNum(ST.hasGFX90AInsts()) +
VGPRForSGPRSpills - MaxVGPRs),
0);
unsigned ArchVGPRLimit = ST.hasGFX90AInsts() ? MaxArchVGPRs : MaxVGPRs;
// Arch VGPR excess pressure conditions, accounting for VGPRs used for SGPR
// spills
ArchVGPR = std::max(static_cast<int>(RP.getArchVGPRNum() +
VGPRForSGPRSpills - ArchVGPRLimit),
0);
// AGPR excess pressure conditions
AGPR = std::max(static_cast<int>(RP.getAGPRNum() - ArchVGPRLimit), 0);
}
};
} // namespace
bool GCNRegPressure::less(const MachineFunction &MF, const GCNRegPressure &O,
unsigned MaxOccupancy) const {
const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
unsigned DynamicVGPRBlockSize =
MF.getInfo<SIMachineFunctionInfo>()->getDynamicVGPRBlockSize();
const auto SGPROcc = std::min(MaxOccupancy,
ST.getOccupancyWithNumSGPRs(getSGPRNum()));
const auto VGPROcc = std::min(
MaxOccupancy, ST.getOccupancyWithNumVGPRs(getVGPRNum(ST.hasGFX90AInsts()),
DynamicVGPRBlockSize));
const auto OtherSGPROcc = std::min(MaxOccupancy,
ST.getOccupancyWithNumSGPRs(O.getSGPRNum()));
const auto OtherVGPROcc =
std::min(MaxOccupancy,
ST.getOccupancyWithNumVGPRs(O.getVGPRNum(ST.hasGFX90AInsts()),
DynamicVGPRBlockSize));
const auto Occ = std::min(SGPROcc, VGPROcc);
const auto OtherOcc = std::min(OtherSGPROcc, OtherVGPROcc);
// Give first precedence to the better occupancy.
if (Occ != OtherOcc)
return Occ > OtherOcc;
unsigned MaxVGPRs = ST.getMaxNumVGPRs(MF);
RegExcess Excess(MF, *this);
RegExcess OtherExcess(MF, O);
unsigned MaxArchVGPRs = ST.getAddressableNumArchVGPRs();
bool ExcessRP = Excess.anyExcess();
bool OtherExcessRP = OtherExcess.anyExcess();
// Give second precedence to the reduced number of spills to hold the register
// pressure.
if (ExcessRP || OtherExcessRP) {
// The difference in excess VGPR pressure, after including VGPRs used for
// SGPR spills
int VGPRDiff =
((OtherExcess.VGPR + OtherExcess.ArchVGPR + OtherExcess.AGPR) -
(Excess.VGPR + Excess.ArchVGPR + Excess.AGPR));
int SGPRDiff = OtherExcess.SGPR - Excess.SGPR;
if (VGPRDiff != 0)
return VGPRDiff > 0;
if (SGPRDiff != 0) {
unsigned PureExcessVGPR =
std::max(static_cast<int>(getVGPRNum(ST.hasGFX90AInsts()) - MaxVGPRs),
0) +
std::max(static_cast<int>(getVGPRNum(false) - MaxArchVGPRs), 0);
unsigned OtherPureExcessVGPR =
std::max(
static_cast<int>(O.getVGPRNum(ST.hasGFX90AInsts()) - MaxVGPRs),
0) +
std::max(static_cast<int>(O.getVGPRNum(false) - MaxArchVGPRs), 0);
// If we have a special case where there is a tie in excess VGPR, but one
// of the pressures has VGPR usage from SGPR spills, prefer the pressure
// with SGPR spills.
if (PureExcessVGPR != OtherPureExcessVGPR)
return SGPRDiff < 0;
// If both pressures have the same excess pressure before and after
// accounting for SGPR spills, prefer fewer SGPR spills.
return SGPRDiff > 0;
}
}
bool SGPRImportant = SGPROcc < VGPROcc;
const bool OtherSGPRImportant = OtherSGPROcc < OtherVGPROcc;
// If both pressures disagree on what is more important compare vgprs.
if (SGPRImportant != OtherSGPRImportant) {
SGPRImportant = false;
}
// Give third precedence to lower register tuple pressure.
bool SGPRFirst = SGPRImportant;
for (int I = 2; I > 0; --I, SGPRFirst = !SGPRFirst) {
if (SGPRFirst) {
auto SW = getSGPRTuplesWeight();
auto OtherSW = O.getSGPRTuplesWeight();
if (SW != OtherSW)
return SW < OtherSW;
} else {
auto VW = getVGPRTuplesWeight();
auto OtherVW = O.getVGPRTuplesWeight();
if (VW != OtherVW)
return VW < OtherVW;
}
}
// Give final precedence to lower general RP.
return SGPRImportant ? (getSGPRNum() < O.getSGPRNum()):
(getVGPRNum(ST.hasGFX90AInsts()) <
O.getVGPRNum(ST.hasGFX90AInsts()));
}
Printable llvm::print(const GCNRegPressure &RP, const GCNSubtarget *ST,
unsigned DynamicVGPRBlockSize) {
return Printable([&RP, ST, DynamicVGPRBlockSize](raw_ostream &OS) {
OS << "VGPRs: " << RP.getArchVGPRNum() << ' '
<< "AGPRs: " << RP.getAGPRNum();
if (ST)
OS << "(O"
<< ST->getOccupancyWithNumVGPRs(RP.getVGPRNum(ST->hasGFX90AInsts()),
DynamicVGPRBlockSize)
<< ')';
OS << ", SGPRs: " << RP.getSGPRNum();
if (ST)
OS << "(O" << ST->getOccupancyWithNumSGPRs(RP.getSGPRNum()) << ')';
OS << ", LVGPR WT: " << RP.getVGPRTuplesWeight()
<< ", LSGPR WT: " << RP.getSGPRTuplesWeight();
if (ST)
OS << " -> Occ: " << RP.getOccupancy(*ST, DynamicVGPRBlockSize);
OS << '\n';
});
}
static LaneBitmask getDefRegMask(const MachineOperand &MO,
const MachineRegisterInfo &MRI) {
assert(MO.isDef() && MO.isReg() && MO.getReg().isVirtual());
// We don't rely on read-undef flag because in case of tentative schedule
// tracking it isn't set correctly yet. This works correctly however since
// use mask has been tracked before using LIS.
return MO.getSubReg() == 0 ?
MRI.getMaxLaneMaskForVReg(MO.getReg()) :
MRI.getTargetRegisterInfo()->getSubRegIndexLaneMask(MO.getSubReg());
}
static void
collectVirtualRegUses(SmallVectorImpl<VRegMaskOrUnit> &VRegMaskOrUnits,
const MachineInstr &MI, const LiveIntervals &LIS,
const MachineRegisterInfo &MRI) {
auto &TRI = *MRI.getTargetRegisterInfo();
for (const auto &MO : MI.operands()) {
if (!MO.isReg() || !MO.getReg().isVirtual())
continue;
if (!MO.isUse() || !MO.readsReg())
continue;
Register Reg = MO.getReg();
auto I = llvm::find_if(VRegMaskOrUnits, [Reg](const VRegMaskOrUnit &RM) {
return RM.VRegOrUnit.asVirtualReg() == Reg;
});
auto &P = I == VRegMaskOrUnits.end()
? VRegMaskOrUnits.emplace_back(VirtRegOrUnit(Reg),
LaneBitmask::getNone())
: *I;
P.LaneMask |= MO.getSubReg() ? TRI.getSubRegIndexLaneMask(MO.getSubReg())
: MRI.getMaxLaneMaskForVReg(Reg);
}
SlotIndex InstrSI;
for (auto &P : VRegMaskOrUnits) {
auto &LI = LIS.getInterval(P.VRegOrUnit.asVirtualReg());
if (!LI.hasSubRanges())
continue;
// For a tentative schedule LIS isn't updated yet but livemask should
// remain the same on any schedule. Subreg defs can be reordered but they
// all must dominate uses anyway.
if (!InstrSI)
InstrSI = LIS.getInstructionIndex(MI).getBaseIndex();
P.LaneMask = getLiveLaneMask(LI, InstrSI, MRI, P.LaneMask);
}
}
/// Mostly copy/paste from CodeGen/RegisterPressure.cpp
static LaneBitmask getLanesWithProperty(
const LiveIntervals &LIS, const MachineRegisterInfo &MRI,
bool TrackLaneMasks, Register Reg, SlotIndex Pos,
function_ref<bool(const LiveRange &LR, SlotIndex Pos)> Property) {
assert(Reg.isVirtual());
const LiveInterval &LI = LIS.getInterval(Reg);
LaneBitmask Result;
if (TrackLaneMasks && LI.hasSubRanges()) {
for (const LiveInterval::SubRange &SR : LI.subranges()) {
if (Property(SR, Pos))
Result |= SR.LaneMask;
}
} else if (Property(LI, Pos)) {
Result =
TrackLaneMasks ? MRI.getMaxLaneMaskForVReg(Reg) : LaneBitmask::getAll();
}
return Result;
}
/// Mostly copy/paste from CodeGen/RegisterPressure.cpp
/// Helper to find a vreg use between two indices {PriorUseIdx, NextUseIdx}.
/// The query starts with a lane bitmask which gets lanes/bits removed for every
/// use we find.
static LaneBitmask findUseBetween(unsigned Reg, LaneBitmask LastUseMask,
SlotIndex PriorUseIdx, SlotIndex NextUseIdx,
const MachineRegisterInfo &MRI,
const SIRegisterInfo *TRI,
const LiveIntervals *LIS,
bool Upward = false) {
for (const MachineOperand &MO : MRI.use_nodbg_operands(Reg)) {
if (MO.isUndef())
continue;
const MachineInstr *MI = MO.getParent();
SlotIndex InstSlot = LIS->getInstructionIndex(*MI).getRegSlot();
bool InRange = Upward ? (InstSlot > PriorUseIdx && InstSlot <= NextUseIdx)
: (InstSlot >= PriorUseIdx && InstSlot < NextUseIdx);
if (!InRange)
continue;
unsigned SubRegIdx = MO.getSubReg();
LaneBitmask UseMask = TRI->getSubRegIndexLaneMask(SubRegIdx);
LastUseMask &= ~UseMask;
if (LastUseMask.none())
return LaneBitmask::getNone();
}
return LastUseMask;
}
////////////////////////////////////////////////////////////////////////////////
// GCNRPTarget
GCNRPTarget::GCNRPTarget(const MachineFunction &MF, const GCNRegPressure &RP)
: GCNRPTarget(RP, MF) {
const Function &F = MF.getFunction();
const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
setTarget(ST.getMaxNumSGPRs(F), ST.getMaxNumVGPRs(F));
}
GCNRPTarget::GCNRPTarget(unsigned NumSGPRs, unsigned NumVGPRs,
const MachineFunction &MF, const GCNRegPressure &RP)
: GCNRPTarget(RP, MF) {
setTarget(NumSGPRs, NumVGPRs);
}
GCNRPTarget::GCNRPTarget(unsigned Occupancy, const MachineFunction &MF,
const GCNRegPressure &RP)
: GCNRPTarget(RP, MF) {
const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
unsigned DynamicVGPRBlockSize =
MF.getInfo<SIMachineFunctionInfo>()->getDynamicVGPRBlockSize();
setTarget(ST.getMaxNumSGPRs(Occupancy, /*Addressable=*/false),
ST.getMaxNumVGPRs(Occupancy, DynamicVGPRBlockSize));
}
void GCNRPTarget::setTarget(unsigned NumSGPRs, unsigned NumVGPRs) {
const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
MaxSGPRs = std::min(ST.getAddressableNumSGPRs(), NumSGPRs);
MaxVGPRs = std::min(ST.getAddressableNumArchVGPRs(), NumVGPRs);
if (UnifiedRF) {
unsigned DynamicVGPRBlockSize =
MF.getInfo<SIMachineFunctionInfo>()->getDynamicVGPRBlockSize();
MaxUnifiedVGPRs =
std::min(ST.getAddressableNumVGPRs(DynamicVGPRBlockSize), NumVGPRs);
} else {
MaxUnifiedVGPRs = 0;
}
}
bool GCNRPTarget::isSaveBeneficial(Register Reg) const {
const MachineRegisterInfo &MRI = MF.getRegInfo();
const TargetRegisterClass *RC = MRI.getRegClass(Reg);
const TargetRegisterInfo *TRI = MRI.getTargetRegisterInfo();
const SIRegisterInfo *SRI = static_cast<const SIRegisterInfo *>(TRI);
RegExcess Excess(MF, RP, *this);
if (SRI->isSGPRClass(RC))
return Excess.SGPR;
if (SRI->isAGPRClass(RC))
return (UnifiedRF && Excess.VGPR) || Excess.AGPR;
return (UnifiedRF && Excess.VGPR) || Excess.ArchVGPR;
}
bool GCNRPTarget::satisfied() const {
if (RP.getSGPRNum() > MaxSGPRs || RP.getVGPRNum(false) > MaxVGPRs)
return false;
if (UnifiedRF && RP.getVGPRNum(true) > MaxUnifiedVGPRs)
return false;
return true;
}
bool GCNRPTarget::hasVectorRegisterExcess() const {
RegExcess Excess(MF, RP, *this);
return Excess.hasVectorRegisterExcess();
}
///////////////////////////////////////////////////////////////////////////////
// GCNRPTracker
LaneBitmask llvm::getLiveLaneMask(unsigned Reg, SlotIndex SI,
const LiveIntervals &LIS,
const MachineRegisterInfo &MRI,
LaneBitmask LaneMaskFilter) {
return getLiveLaneMask(LIS.getInterval(Reg), SI, MRI, LaneMaskFilter);
}
LaneBitmask llvm::getLiveLaneMask(const LiveInterval &LI, SlotIndex SI,
const MachineRegisterInfo &MRI,
LaneBitmask LaneMaskFilter) {
LaneBitmask LiveMask;
if (LI.hasSubRanges()) {
for (const auto &S : LI.subranges())
if ((S.LaneMask & LaneMaskFilter).any() && S.liveAt(SI)) {
LiveMask |= S.LaneMask;
assert(LiveMask == (LiveMask & MRI.getMaxLaneMaskForVReg(LI.reg())));
}
} else if (LI.liveAt(SI)) {
LiveMask = MRI.getMaxLaneMaskForVReg(LI.reg());
}
LiveMask &= LaneMaskFilter;
return LiveMask;
}
GCNRPTracker::LiveRegSet llvm::getLiveRegs(SlotIndex SI,
const LiveIntervals &LIS,
const MachineRegisterInfo &MRI,
GCNRegPressure::RegKind RegKind) {
GCNRPTracker::LiveRegSet LiveRegs;
for (unsigned I = 0, E = MRI.getNumVirtRegs(); I != E; ++I) {
auto Reg = Register::index2VirtReg(I);
if (RegKind != GCNRegPressure::TOTAL_KINDS &&
GCNRegPressure::getRegKind(Reg, MRI) != RegKind)
continue;
if (!LIS.hasInterval(Reg))
continue;
auto LiveMask = getLiveLaneMask(Reg, SI, LIS, MRI);
if (LiveMask.any())
LiveRegs[Reg] = LiveMask;
}
return LiveRegs;
}
void GCNRPTracker::reset(const MachineInstr &MI,
const LiveRegSet *LiveRegsCopy,
bool After) {
const MachineFunction &MF = *MI.getMF();
MRI = &MF.getRegInfo();
if (LiveRegsCopy) {
if (&LiveRegs != LiveRegsCopy)
LiveRegs = *LiveRegsCopy;
} else {
LiveRegs = After ? getLiveRegsAfter(MI, LIS)
: getLiveRegsBefore(MI, LIS);
}
MaxPressure = CurPressure = getRegPressure(*MRI, LiveRegs);
}
void GCNRPTracker::reset(const MachineRegisterInfo &MRI_,
const LiveRegSet &LiveRegs_) {
MRI = &MRI_;
LiveRegs = LiveRegs_;
LastTrackedMI = nullptr;
MaxPressure = CurPressure = getRegPressure(MRI_, LiveRegs_);
}
/// Mostly copy/paste from CodeGen/RegisterPressure.cpp
LaneBitmask GCNRPTracker::getLastUsedLanes(Register Reg, SlotIndex Pos) const {
return getLanesWithProperty(
LIS, *MRI, true, Reg, Pos.getBaseIndex(),
[](const LiveRange &LR, SlotIndex Pos) {
const LiveRange::Segment *S = LR.getSegmentContaining(Pos);
return S != nullptr && S->end == Pos.getRegSlot();
});
}
////////////////////////////////////////////////////////////////////////////////
// GCNUpwardRPTracker
void GCNUpwardRPTracker::recede(const MachineInstr &MI) {
assert(MRI && "call reset first");
LastTrackedMI = &MI;
if (MI.isDebugInstr())
return;
// Kill all defs.
GCNRegPressure DefPressure, ECDefPressure;
bool HasECDefs = false;
for (const MachineOperand &MO : MI.all_defs()) {
if (!MO.getReg().isVirtual())
continue;
Register Reg = MO.getReg();
LaneBitmask DefMask = getDefRegMask(MO, *MRI);
// Treat a def as fully live at the moment of definition: keep a record.
if (MO.isEarlyClobber()) {
ECDefPressure.inc(Reg, LaneBitmask::getNone(), DefMask, *MRI);
HasECDefs = true;
} else
DefPressure.inc(Reg, LaneBitmask::getNone(), DefMask, *MRI);
auto I = LiveRegs.find(Reg);
if (I == LiveRegs.end())
continue;
LaneBitmask &LiveMask = I->second;
LaneBitmask PrevMask = LiveMask;
LiveMask &= ~DefMask;
CurPressure.inc(Reg, PrevMask, LiveMask, *MRI);
if (LiveMask.none())
LiveRegs.erase(I);
}
// Update MaxPressure with defs pressure.
DefPressure += CurPressure;
if (HasECDefs)
DefPressure += ECDefPressure;
MaxPressure = max(DefPressure, MaxPressure);
// Make uses alive.
SmallVector<VRegMaskOrUnit, 8> RegUses;
collectVirtualRegUses(RegUses, MI, LIS, *MRI);
for (const VRegMaskOrUnit &U : RegUses) {
LaneBitmask &LiveMask = LiveRegs[U.VRegOrUnit.asVirtualReg()];
LaneBitmask PrevMask = LiveMask;
LiveMask |= U.LaneMask;
CurPressure.inc(U.VRegOrUnit.asVirtualReg(), PrevMask, LiveMask, *MRI);
}
// Update MaxPressure with uses plus early-clobber defs pressure.
MaxPressure = HasECDefs ? max(CurPressure + ECDefPressure, MaxPressure)
: max(CurPressure, MaxPressure);
assert(CurPressure == getRegPressure(*MRI, LiveRegs));
}
////////////////////////////////////////////////////////////////////////////////
// GCNDownwardRPTracker
bool GCNDownwardRPTracker::reset(const MachineInstr &MI,
const LiveRegSet *LiveRegsCopy) {
MRI = &MI.getMF()->getRegInfo();
LastTrackedMI = nullptr;
MBBEnd = MI.getParent()->end();
NextMI = &MI;
NextMI = skipDebugInstructionsForward(NextMI, MBBEnd);
if (NextMI == MBBEnd)
return false;
GCNRPTracker::reset(*NextMI, LiveRegsCopy, false);
return true;
}
bool GCNDownwardRPTracker::advanceBeforeNext(MachineInstr *MI,
bool UseInternalIterator) {
assert(MRI && "call reset first");
SlotIndex SI;
const MachineInstr *CurrMI;
if (UseInternalIterator) {
if (!LastTrackedMI)
return NextMI == MBBEnd;
assert(NextMI == MBBEnd || !NextMI->isDebugInstr());
CurrMI = LastTrackedMI;
SI = NextMI == MBBEnd
? LIS.getInstructionIndex(*LastTrackedMI).getDeadSlot()
: LIS.getInstructionIndex(*NextMI).getBaseIndex();
} else { //! UseInternalIterator
SI = LIS.getInstructionIndex(*MI).getBaseIndex();
CurrMI = MI;
}
assert(SI.isValid());
// Remove dead registers or mask bits.
SmallSet<Register, 8> SeenRegs;
for (auto &MO : CurrMI->operands()) {
if (!MO.isReg() || !MO.getReg().isVirtual())
continue;
if (MO.isUse() && !MO.readsReg())
continue;
if (!UseInternalIterator && MO.isDef())
continue;
if (!SeenRegs.insert(MO.getReg()).second)
continue;
const LiveInterval &LI = LIS.getInterval(MO.getReg());
if (LI.hasSubRanges()) {
auto It = LiveRegs.end();
for (const auto &S : LI.subranges()) {
if (!S.liveAt(SI)) {
if (It == LiveRegs.end()) {
It = LiveRegs.find(MO.getReg());
if (It == LiveRegs.end())
llvm_unreachable("register isn't live");
}
auto PrevMask = It->second;
It->second &= ~S.LaneMask;
CurPressure.inc(MO.getReg(), PrevMask, It->second, *MRI);
}
}
if (It != LiveRegs.end() && It->second.none())
LiveRegs.erase(It);
} else if (!LI.liveAt(SI)) {
auto It = LiveRegs.find(MO.getReg());
if (It == LiveRegs.end())
llvm_unreachable("register isn't live");
CurPressure.inc(MO.getReg(), It->second, LaneBitmask::getNone(), *MRI);
LiveRegs.erase(It);
}
}
MaxPressure = max(MaxPressure, CurPressure);
LastTrackedMI = nullptr;
return UseInternalIterator && (NextMI == MBBEnd);
}
void GCNDownwardRPTracker::advanceToNext(MachineInstr *MI,
bool UseInternalIterator) {
if (UseInternalIterator) {
LastTrackedMI = &*NextMI++;
NextMI = skipDebugInstructionsForward(NextMI, MBBEnd);
} else {
LastTrackedMI = MI;
}
const MachineInstr *CurrMI = LastTrackedMI;
// Add new registers or mask bits.
for (const auto &MO : CurrMI->all_defs()) {
Register Reg = MO.getReg();
if (!Reg.isVirtual())
continue;
auto &LiveMask = LiveRegs[Reg];
auto PrevMask = LiveMask;
LiveMask |= getDefRegMask(MO, *MRI);
CurPressure.inc(Reg, PrevMask, LiveMask, *MRI);
}
MaxPressure = max(MaxPressure, CurPressure);
}
bool GCNDownwardRPTracker::advance(MachineInstr *MI, bool UseInternalIterator) {
if (UseInternalIterator && NextMI == MBBEnd)
return false;
advanceBeforeNext(MI, UseInternalIterator);
advanceToNext(MI, UseInternalIterator);
if (!UseInternalIterator) {
// We must remove any dead def lanes from the current RP
advanceBeforeNext(MI, true);
}
return true;
}
bool GCNDownwardRPTracker::advance(MachineBasicBlock::const_iterator End) {
while (NextMI != End)
if (!advance()) return false;
return true;
}
bool GCNDownwardRPTracker::advance(MachineBasicBlock::const_iterator Begin,
MachineBasicBlock::const_iterator End,
const LiveRegSet *LiveRegsCopy) {
reset(*Begin, LiveRegsCopy);
return advance(End);
}
Printable llvm::reportMismatch(const GCNRPTracker::LiveRegSet &LISLR,
const GCNRPTracker::LiveRegSet &TrackedLR,
const TargetRegisterInfo *TRI, StringRef Pfx) {
return Printable([&LISLR, &TrackedLR, TRI, Pfx](raw_ostream &OS) {
for (auto const &P : TrackedLR) {
auto I = LISLR.find(P.first);
if (I == LISLR.end()) {
OS << Pfx << printReg(P.first, TRI) << ":L" << PrintLaneMask(P.second)
<< " isn't found in LIS reported set\n";
} else if (I->second != P.second) {
OS << Pfx << printReg(P.first, TRI)
<< " masks doesn't match: LIS reported " << PrintLaneMask(I->second)
<< ", tracked " << PrintLaneMask(P.second) << '\n';
}
}
for (auto const &P : LISLR) {
auto I = TrackedLR.find(P.first);
if (I == TrackedLR.end()) {
OS << Pfx << printReg(P.first, TRI) << ":L" << PrintLaneMask(P.second)
<< " isn't found in tracked set\n";
}
}
});
}
GCNRegPressure
GCNDownwardRPTracker::bumpDownwardPressure(const MachineInstr *MI,
const SIRegisterInfo *TRI) const {
assert(!MI->isDebugOrPseudoInstr() && "Expect a nondebug instruction.");
SlotIndex SlotIdx;
SlotIdx = LIS.getInstructionIndex(*MI).getRegSlot();
// Account for register pressure similar to RegPressureTracker::recede().
RegisterOperands RegOpers;
RegOpers.collect(*MI, *TRI, *MRI, true, /*IgnoreDead=*/false);
RegOpers.adjustLaneLiveness(LIS, *MRI, SlotIdx);
GCNRegPressure TempPressure = CurPressure;
for (const VRegMaskOrUnit &Use : RegOpers.Uses) {
if (!Use.VRegOrUnit.isVirtualReg())
continue;
Register Reg = Use.VRegOrUnit.asVirtualReg();
LaneBitmask LastUseMask = getLastUsedLanes(Reg, SlotIdx);
if (LastUseMask.none())
continue;
// The LastUseMask is queried from the liveness information of instruction
// which may be further down the schedule. Some lanes may actually not be
// last uses for the current position.
// FIXME: allow the caller to pass in the list of vreg uses that remain
// to be bottom-scheduled to avoid searching uses at each query.
SlotIndex CurrIdx;
const MachineBasicBlock *MBB = MI->getParent();
MachineBasicBlock::const_iterator IdxPos = skipDebugInstructionsForward(
LastTrackedMI ? LastTrackedMI : MBB->begin(), MBB->end());
if (IdxPos == MBB->end()) {
CurrIdx = LIS.getMBBEndIdx(MBB);
} else {
CurrIdx = LIS.getInstructionIndex(*IdxPos).getRegSlot();
}
LastUseMask =
findUseBetween(Reg, LastUseMask, CurrIdx, SlotIdx, *MRI, TRI, &LIS);
if (LastUseMask.none())
continue;
auto It = LiveRegs.find(Reg);
LaneBitmask LiveMask = It != LiveRegs.end() ? It->second : LaneBitmask(0);
LaneBitmask NewMask = LiveMask & ~LastUseMask;
TempPressure.inc(Reg, LiveMask, NewMask, *MRI);
}
// Generate liveness for defs.
for (const VRegMaskOrUnit &Def : RegOpers.Defs) {
if (!Def.VRegOrUnit.isVirtualReg())
continue;
Register Reg = Def.VRegOrUnit.asVirtualReg();
auto It = LiveRegs.find(Reg);
LaneBitmask LiveMask = It != LiveRegs.end() ? It->second : LaneBitmask(0);
LaneBitmask NewMask = LiveMask | Def.LaneMask;
TempPressure.inc(Reg, LiveMask, NewMask, *MRI);
}
return TempPressure;
}
bool GCNUpwardRPTracker::isValid() const {
const auto &SI = LIS.getInstructionIndex(*LastTrackedMI).getBaseIndex();
const auto LISLR = llvm::getLiveRegs(SI, LIS, *MRI);
const auto &TrackedLR = LiveRegs;
if (!isEqual(LISLR, TrackedLR)) {
dbgs() << "\nGCNUpwardRPTracker error: Tracked and"
" LIS reported livesets mismatch:\n"
<< print(LISLR, *MRI);
reportMismatch(LISLR, TrackedLR, MRI->getTargetRegisterInfo());
return false;
}
auto LISPressure = getRegPressure(*MRI, LISLR);
if (LISPressure != CurPressure) {
dbgs() << "GCNUpwardRPTracker error: Pressure sets different\nTracked: "
<< print(CurPressure) << "LIS rpt: " << print(LISPressure);
return false;
}
return true;
}
Printable llvm::print(const GCNRPTracker::LiveRegSet &LiveRegs,
const MachineRegisterInfo &MRI) {
return Printable([&LiveRegs, &MRI](raw_ostream &OS) {
const TargetRegisterInfo *TRI = MRI.getTargetRegisterInfo();
for (unsigned I = 0, E = MRI.getNumVirtRegs(); I != E; ++I) {
Register Reg = Register::index2VirtReg(I);
auto It = LiveRegs.find(Reg);
if (It != LiveRegs.end() && It->second.any())
OS << ' ' << printReg(Reg, TRI) << ':' << PrintLaneMask(It->second);
}
OS << '\n';
});
}
void GCNRegPressure::dump() const { dbgs() << print(*this); }
static cl::opt<bool> UseDownwardTracker(
"amdgpu-print-rp-downward",
cl::desc("Use GCNDownwardRPTracker for GCNRegPressurePrinter pass"),
cl::init(false), cl::Hidden);
char llvm::GCNRegPressurePrinter::ID = 0;
char &llvm::GCNRegPressurePrinterID = GCNRegPressurePrinter::ID;
INITIALIZE_PASS(GCNRegPressurePrinter, "amdgpu-print-rp", "", true, true)
// Return lanemask of Reg's subregs that are live-through at [Begin, End] and
// are fully covered by Mask.
static LaneBitmask
getRegLiveThroughMask(const MachineRegisterInfo &MRI, const LiveIntervals &LIS,
Register Reg, SlotIndex Begin, SlotIndex End,
LaneBitmask Mask = LaneBitmask::getAll()) {
auto IsInOneSegment = [Begin, End](const LiveRange &LR) -> bool {
auto *Segment = LR.getSegmentContaining(Begin);
return Segment && Segment->contains(End);
};
LaneBitmask LiveThroughMask;
const LiveInterval &LI = LIS.getInterval(Reg);
if (LI.hasSubRanges()) {
for (auto &SR : LI.subranges()) {
if ((SR.LaneMask & Mask) == SR.LaneMask && IsInOneSegment(SR))
LiveThroughMask |= SR.LaneMask;
}
} else {
LaneBitmask RegMask = MRI.getMaxLaneMaskForVReg(Reg);
if ((RegMask & Mask) == RegMask && IsInOneSegment(LI))
LiveThroughMask = RegMask;
}
return LiveThroughMask;
}
bool GCNRegPressurePrinter::runOnMachineFunction(MachineFunction &MF) {
const MachineRegisterInfo &MRI = MF.getRegInfo();
const TargetRegisterInfo *TRI = MRI.getTargetRegisterInfo();
const LiveIntervals &LIS = getAnalysis<LiveIntervalsWrapperPass>().getLIS();
auto &OS = dbgs();
// Leading spaces are important for YAML syntax.
#define PFX " "
OS << "---\nname: " << MF.getName() << "\nbody: |\n";
auto printRP = [](const GCNRegPressure &RP) {
return Printable([&RP](raw_ostream &OS) {
OS << format(PFX " %-5d", RP.getSGPRNum())
<< format(" %-5d", RP.getVGPRNum(false));
});
};
auto ReportLISMismatchIfAny = [&](const GCNRPTracker::LiveRegSet &TrackedLR,
const GCNRPTracker::LiveRegSet &LISLR) {
if (LISLR != TrackedLR) {
OS << PFX " mis LIS: " << llvm::print(LISLR, MRI)
<< reportMismatch(LISLR, TrackedLR, TRI, PFX " ");
}
};
// Register pressure before and at an instruction (in program order).
SmallVector<std::pair<GCNRegPressure, GCNRegPressure>, 16> RP;
for (auto &MBB : MF) {
RP.clear();
RP.reserve(MBB.size());
OS << PFX;
MBB.printName(OS);
OS << ":\n";
SlotIndex MBBStartSlot = LIS.getSlotIndexes()->getMBBStartIdx(&MBB);
SlotIndex MBBLastSlot = LIS.getSlotIndexes()->getMBBLastIdx(&MBB);
GCNRPTracker::LiveRegSet LiveIn, LiveOut;
GCNRegPressure RPAtMBBEnd;
if (UseDownwardTracker) {
if (MBB.empty()) {
LiveIn = LiveOut = getLiveRegs(MBBStartSlot, LIS, MRI);
RPAtMBBEnd = getRegPressure(MRI, LiveIn);
} else {
GCNDownwardRPTracker RPT(LIS);
RPT.reset(MBB.front());
LiveIn = RPT.getLiveRegs();
while (!RPT.advanceBeforeNext()) {
GCNRegPressure RPBeforeMI = RPT.getPressure();
RPT.advanceToNext();
RP.emplace_back(RPBeforeMI, RPT.getPressure());
}
LiveOut = RPT.getLiveRegs();
RPAtMBBEnd = RPT.getPressure();
}
} else {
GCNUpwardRPTracker RPT(LIS);
RPT.reset(MRI, MBBLastSlot);
LiveOut = RPT.getLiveRegs();
RPAtMBBEnd = RPT.getPressure();
for (auto &MI : reverse(MBB)) {
RPT.resetMaxPressure();
RPT.recede(MI);
if (!MI.isDebugInstr())
RP.emplace_back(RPT.getPressure(), RPT.getMaxPressure());
}
LiveIn = RPT.getLiveRegs();
}
OS << PFX " Live-in: " << llvm::print(LiveIn, MRI);
if (!UseDownwardTracker)
ReportLISMismatchIfAny(LiveIn, getLiveRegs(MBBStartSlot, LIS, MRI));
OS << PFX " SGPR VGPR\n";
int I = 0;
for (auto &MI : MBB) {
if (!MI.isDebugInstr()) {
auto &[RPBeforeInstr, RPAtInstr] =
RP[UseDownwardTracker ? I : (RP.size() - 1 - I)];
++I;
OS << printRP(RPBeforeInstr) << '\n' << printRP(RPAtInstr) << " ";
} else
OS << PFX " ";
MI.print(OS);
}
OS << printRP(RPAtMBBEnd) << '\n';
OS << PFX " Live-out:" << llvm::print(LiveOut, MRI);
if (UseDownwardTracker)
ReportLISMismatchIfAny(LiveOut, getLiveRegs(MBBLastSlot, LIS, MRI));
GCNRPTracker::LiveRegSet LiveThrough;
for (auto [Reg, Mask] : LiveIn) {
LaneBitmask MaskIntersection = Mask & LiveOut.lookup(Reg);
if (MaskIntersection.any()) {
LaneBitmask LTMask = getRegLiveThroughMask(
MRI, LIS, Reg, MBBStartSlot, MBBLastSlot, MaskIntersection);
if (LTMask.any())
LiveThrough[Reg] = LTMask;
}
}
OS << PFX " Live-thr:" << llvm::print(LiveThrough, MRI);
OS << printRP(getRegPressure(MRI, LiveThrough)) << '\n';
}
OS << "...\n";
return false;
#undef PFX
}
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
LLVM_DUMP_METHOD void llvm::dumpMaxRegPressure(MachineFunction &MF,
GCNRegPressure::RegKind Kind,
LiveIntervals &LIS,
const MachineLoopInfo *MLI) {
const MachineRegisterInfo &MRI = MF.getRegInfo();
const TargetRegisterInfo *TRI = MRI.getTargetRegisterInfo();
auto &OS = dbgs();
const char *RegName = GCNRegPressure::getName(Kind);
unsigned MaxNumRegs = 0;
const MachineInstr *MaxPressureMI = nullptr;
GCNUpwardRPTracker RPT(LIS);
for (const MachineBasicBlock &MBB : MF) {
RPT.reset(MRI, LIS.getSlotIndexes()->getMBBEndIdx(&MBB).getPrevSlot());
for (const MachineInstr &MI : reverse(MBB)) {
RPT.recede(MI);
unsigned NumRegs = RPT.getMaxPressure().getNumRegs(Kind);
if (NumRegs > MaxNumRegs) {
MaxNumRegs = NumRegs;
MaxPressureMI = &MI;
}
}
}
SlotIndex MISlot = LIS.getInstructionIndex(*MaxPressureMI);
// Max pressure can occur at either the early-clobber or register slot.
// Choose the maximum liveset between both slots. This is ugly but this is
// diagnostic code.
SlotIndex ECSlot = MISlot.getRegSlot(true);
SlotIndex RSlot = MISlot.getRegSlot(false);
GCNRPTracker::LiveRegSet ECLiveSet = getLiveRegs(ECSlot, LIS, MRI, Kind);
GCNRPTracker::LiveRegSet RLiveSet = getLiveRegs(RSlot, LIS, MRI, Kind);
unsigned ECNumRegs = getRegPressure(MRI, ECLiveSet).getNumRegs(Kind);
unsigned RNumRegs = getRegPressure(MRI, RLiveSet).getNumRegs(Kind);
GCNRPTracker::LiveRegSet *LiveSet =
ECNumRegs > RNumRegs ? &ECLiveSet : &RLiveSet;
SlotIndex MaxPressureSlot = ECNumRegs > RNumRegs ? ECSlot : RSlot;
assert(getRegPressure(MRI, *LiveSet).getNumRegs(Kind) == MaxNumRegs);
// Split live registers into single-def and multi-def sets.
GCNRegPressure SDefPressure, MDefPressure;
SmallVector<Register, 16> SDefRegs, MDefRegs;
for (auto [Reg, LaneMask] : *LiveSet) {
assert(GCNRegPressure::getRegKind(Reg, MRI) == Kind);
LiveInterval &LI = LIS.getInterval(Reg);
if (LI.getNumValNums() == 1 ||
(LI.hasSubRanges() &&
llvm::all_of(LI.subranges(), [](const LiveInterval::SubRange &SR) {
return SR.getNumValNums() == 1;
}))) {
SDefPressure.inc(Reg, LaneBitmask::getNone(), LaneMask, MRI);
SDefRegs.push_back(Reg);
} else {
MDefPressure.inc(Reg, LaneBitmask::getNone(), LaneMask, MRI);
MDefRegs.push_back(Reg);
}
}
unsigned SDefNumRegs = SDefPressure.getNumRegs(Kind);
unsigned MDefNumRegs = MDefPressure.getNumRegs(Kind);
assert(SDefNumRegs + MDefNumRegs == MaxNumRegs);
auto printLoc = [&](const MachineBasicBlock *MBB, SlotIndex SI) {
return Printable([&, MBB, SI](raw_ostream &OS) {
OS << SI << ':' << printMBBReference(*MBB);
if (MLI)
if (const MachineLoop *ML = MLI->getLoopFor(MBB))
OS << " (LoopHdr " << printMBBReference(*ML->getHeader())
<< ", Depth " << ML->getLoopDepth() << ")";
});
};
auto PrintRegInfo = [&](Register Reg, LaneBitmask LiveMask) {
GCNRegPressure RegPressure;
RegPressure.inc(Reg, LaneBitmask::getNone(), LiveMask, MRI);
OS << " " << printReg(Reg, TRI) << ':'
<< TRI->getRegClassName(MRI.getRegClass(Reg)) << ", LiveMask "
<< PrintLaneMask(LiveMask) << " (" << RegPressure.getNumRegs(Kind) << ' '
<< RegName << "s)\n";
// Use std::map to sort def/uses by SlotIndex.
std::map<SlotIndex, const MachineInstr *> Instrs;
for (const MachineInstr &MI : MRI.reg_nodbg_instructions(Reg)) {
Instrs[LIS.getInstructionIndex(MI).getRegSlot()] = &MI;
}
for (const auto &[SI, MI] : Instrs) {
OS << " ";
if (MI->definesRegister(Reg, TRI))
OS << "def ";
if (MI->readsRegister(Reg, TRI))
OS << "use ";
OS << printLoc(MI->getParent(), SI) << ": " << *MI;
}
};
OS << "\n*** Register pressure info (" << RegName << "s) for " << MF.getName()
<< " ***\n";
OS << "Max pressure is " << MaxNumRegs << ' ' << RegName << "s at "
<< printLoc(MaxPressureMI->getParent(), MaxPressureSlot) << ": "
<< *MaxPressureMI;
OS << "\nLive registers with single definition (" << SDefNumRegs << ' '
<< RegName << "s):\n";
// Sort SDefRegs by number of uses (smallest first)
llvm::sort(SDefRegs, [&](Register A, Register B) {
return std::distance(MRI.use_nodbg_begin(A), MRI.use_nodbg_end()) <
std::distance(MRI.use_nodbg_begin(B), MRI.use_nodbg_end());
});
for (const Register Reg : SDefRegs) {
PrintRegInfo(Reg, LiveSet->lookup(Reg));
}
OS << "\nLive registers with multiple definitions (" << MDefNumRegs << ' '
<< RegName << "s):\n";
for (const Register Reg : MDefRegs) {
PrintRegInfo(Reg, LiveSet->lookup(Reg));
}
}
#endif
Reference in New Issue View Git Blame Copy Permalink