At the moment MachineLoopInfo has a very simple way to determine if a use of a physical register will be invariant: if it is not a constant value or if it's not an ignorable use, then it's not considered invariant. From a compile-time performance perspective this makes a lot of sense, but it limits code that uses implicit physical registers from being hoisted until the later MachineLICM pass (after register allocation), which has a lot fewer opportunities to hoist. For AArch64 SME we use an implicit physical register ($vg) to avoid rematerialization beyond certain instructions. Doing this led to regressions because simple expressions were no longer hoisted by Early MachineLICM. This patch adds some extra checks to 'isLoopInvariant' to see if any of the defs are found in the loop. If not, we can considered it loop invariant. We expect the impact on compile-time to be negligible because there is an incentive for users to reduce the need for the smstart/smstop instructions that define $vg. In either case, we've put the functionality under a target interface to limit this only to specific registers.
283 lines
9.3 KiB
C++
283 lines
9.3 KiB
C++
//===- MachineLoopInfo.cpp - Natural Loop Calculator ----------------------===//
|
|
//
|
|
// 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
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
//
|
|
// This file defines the MachineLoopInfo class that is used to identify natural
|
|
// loops and determine the loop depth of various nodes of the CFG. Note that
|
|
// the loops identified may actually be several natural loops that share the
|
|
// same header node... not just a single natural loop.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#include "llvm/CodeGen/MachineLoopInfo.h"
|
|
#include "llvm/CodeGen/MachineDominators.h"
|
|
#include "llvm/CodeGen/MachineRegisterInfo.h"
|
|
#include "llvm/CodeGen/TargetInstrInfo.h"
|
|
#include "llvm/CodeGen/TargetSubtargetInfo.h"
|
|
#include "llvm/Config/llvm-config.h"
|
|
#include "llvm/InitializePasses.h"
|
|
#include "llvm/Pass.h"
|
|
#include "llvm/PassRegistry.h"
|
|
#include "llvm/Support/GenericLoopInfoImpl.h"
|
|
|
|
using namespace llvm;
|
|
|
|
// Explicitly instantiate methods in LoopInfoImpl.h for MI-level Loops.
|
|
template class llvm::LoopBase<MachineBasicBlock, MachineLoop>;
|
|
template class llvm::LoopInfoBase<MachineBasicBlock, MachineLoop>;
|
|
|
|
char MachineLoopInfo::ID = 0;
|
|
MachineLoopInfo::MachineLoopInfo() : MachineFunctionPass(ID) {
|
|
initializeMachineLoopInfoPass(*PassRegistry::getPassRegistry());
|
|
}
|
|
INITIALIZE_PASS_BEGIN(MachineLoopInfo, "machine-loops",
|
|
"Machine Natural Loop Construction", true, true)
|
|
INITIALIZE_PASS_DEPENDENCY(MachineDominatorTree)
|
|
INITIALIZE_PASS_END(MachineLoopInfo, "machine-loops",
|
|
"Machine Natural Loop Construction", true, true)
|
|
|
|
char &llvm::MachineLoopInfoID = MachineLoopInfo::ID;
|
|
|
|
bool MachineLoopInfo::runOnMachineFunction(MachineFunction &) {
|
|
calculate(getAnalysis<MachineDominatorTree>());
|
|
return false;
|
|
}
|
|
|
|
void MachineLoopInfo::calculate(MachineDominatorTree &MDT) {
|
|
releaseMemory();
|
|
LI.analyze(MDT.getBase());
|
|
}
|
|
|
|
void MachineLoopInfo::getAnalysisUsage(AnalysisUsage &AU) const {
|
|
AU.setPreservesAll();
|
|
AU.addRequired<MachineDominatorTree>();
|
|
MachineFunctionPass::getAnalysisUsage(AU);
|
|
}
|
|
|
|
MachineBasicBlock *MachineLoop::getTopBlock() {
|
|
MachineBasicBlock *TopMBB = getHeader();
|
|
MachineFunction::iterator Begin = TopMBB->getParent()->begin();
|
|
if (TopMBB->getIterator() != Begin) {
|
|
MachineBasicBlock *PriorMBB = &*std::prev(TopMBB->getIterator());
|
|
while (contains(PriorMBB)) {
|
|
TopMBB = PriorMBB;
|
|
if (TopMBB->getIterator() == Begin)
|
|
break;
|
|
PriorMBB = &*std::prev(TopMBB->getIterator());
|
|
}
|
|
}
|
|
return TopMBB;
|
|
}
|
|
|
|
MachineBasicBlock *MachineLoop::getBottomBlock() {
|
|
MachineBasicBlock *BotMBB = getHeader();
|
|
MachineFunction::iterator End = BotMBB->getParent()->end();
|
|
if (BotMBB->getIterator() != std::prev(End)) {
|
|
MachineBasicBlock *NextMBB = &*std::next(BotMBB->getIterator());
|
|
while (contains(NextMBB)) {
|
|
BotMBB = NextMBB;
|
|
if (BotMBB == &*std::next(BotMBB->getIterator()))
|
|
break;
|
|
NextMBB = &*std::next(BotMBB->getIterator());
|
|
}
|
|
}
|
|
return BotMBB;
|
|
}
|
|
|
|
MachineBasicBlock *MachineLoop::findLoopControlBlock() const {
|
|
if (MachineBasicBlock *Latch = getLoopLatch()) {
|
|
if (isLoopExiting(Latch))
|
|
return Latch;
|
|
else
|
|
return getExitingBlock();
|
|
}
|
|
return nullptr;
|
|
}
|
|
|
|
DebugLoc MachineLoop::getStartLoc() const {
|
|
// Try the pre-header first.
|
|
if (MachineBasicBlock *PHeadMBB = getLoopPreheader())
|
|
if (const BasicBlock *PHeadBB = PHeadMBB->getBasicBlock())
|
|
if (DebugLoc DL = PHeadBB->getTerminator()->getDebugLoc())
|
|
return DL;
|
|
|
|
// If we have no pre-header or there are no instructions with debug
|
|
// info in it, try the header.
|
|
if (MachineBasicBlock *HeadMBB = getHeader())
|
|
if (const BasicBlock *HeadBB = HeadMBB->getBasicBlock())
|
|
return HeadBB->getTerminator()->getDebugLoc();
|
|
|
|
return DebugLoc();
|
|
}
|
|
|
|
MachineBasicBlock *
|
|
MachineLoopInfo::findLoopPreheader(MachineLoop *L, bool SpeculativePreheader,
|
|
bool FindMultiLoopPreheader) const {
|
|
if (MachineBasicBlock *PB = L->getLoopPreheader())
|
|
return PB;
|
|
|
|
if (!SpeculativePreheader)
|
|
return nullptr;
|
|
|
|
MachineBasicBlock *HB = L->getHeader(), *LB = L->getLoopLatch();
|
|
if (HB->pred_size() != 2 || HB->hasAddressTaken())
|
|
return nullptr;
|
|
// Find the predecessor of the header that is not the latch block.
|
|
MachineBasicBlock *Preheader = nullptr;
|
|
for (MachineBasicBlock *P : HB->predecessors()) {
|
|
if (P == LB)
|
|
continue;
|
|
// Sanity.
|
|
if (Preheader)
|
|
return nullptr;
|
|
Preheader = P;
|
|
}
|
|
|
|
// Check if the preheader candidate is a successor of any other loop
|
|
// headers. We want to avoid having two loop setups in the same block.
|
|
if (!FindMultiLoopPreheader) {
|
|
for (MachineBasicBlock *S : Preheader->successors()) {
|
|
if (S == HB)
|
|
continue;
|
|
MachineLoop *T = getLoopFor(S);
|
|
if (T && T->getHeader() == S)
|
|
return nullptr;
|
|
}
|
|
}
|
|
return Preheader;
|
|
}
|
|
|
|
MDNode *MachineLoop::getLoopID() const {
|
|
MDNode *LoopID = nullptr;
|
|
if (const auto *MBB = findLoopControlBlock()) {
|
|
// If there is a single latch block, then the metadata
|
|
// node is attached to its terminating instruction.
|
|
const auto *BB = MBB->getBasicBlock();
|
|
if (!BB)
|
|
return nullptr;
|
|
if (const auto *TI = BB->getTerminator())
|
|
LoopID = TI->getMetadata(LLVMContext::MD_loop);
|
|
} else if (const auto *MBB = getHeader()) {
|
|
// There seem to be multiple latch blocks, so we have to
|
|
// visit all predecessors of the loop header and check
|
|
// their terminating instructions for the metadata.
|
|
if (const auto *Header = MBB->getBasicBlock()) {
|
|
// Walk over all blocks in the loop.
|
|
for (const auto *MBB : this->blocks()) {
|
|
const auto *BB = MBB->getBasicBlock();
|
|
if (!BB)
|
|
return nullptr;
|
|
const auto *TI = BB->getTerminator();
|
|
if (!TI)
|
|
return nullptr;
|
|
MDNode *MD = nullptr;
|
|
// Check if this terminating instruction jumps to the loop header.
|
|
for (const auto *Succ : successors(TI)) {
|
|
if (Succ == Header) {
|
|
// This is a jump to the header - gather the metadata from it.
|
|
MD = TI->getMetadata(LLVMContext::MD_loop);
|
|
break;
|
|
}
|
|
}
|
|
if (!MD)
|
|
return nullptr;
|
|
if (!LoopID)
|
|
LoopID = MD;
|
|
else if (MD != LoopID)
|
|
return nullptr;
|
|
}
|
|
}
|
|
}
|
|
if (LoopID &&
|
|
(LoopID->getNumOperands() == 0 || LoopID->getOperand(0) != LoopID))
|
|
LoopID = nullptr;
|
|
return LoopID;
|
|
}
|
|
|
|
bool MachineLoop::isLoopInvariantImplicitPhysReg(Register Reg) const {
|
|
MachineFunction *MF = getHeader()->getParent();
|
|
MachineRegisterInfo *MRI = &MF->getRegInfo();
|
|
|
|
if (MRI->isConstantPhysReg(Reg))
|
|
return true;
|
|
|
|
if (!MF->getSubtarget()
|
|
.getRegisterInfo()
|
|
->shouldAnalyzePhysregInMachineLoopInfo(Reg))
|
|
return false;
|
|
|
|
return !llvm::any_of(
|
|
MRI->def_instructions(Reg),
|
|
[this, Reg](const MachineInstr &MI) { return this->contains(&MI); });
|
|
}
|
|
|
|
bool MachineLoop::isLoopInvariant(MachineInstr &I,
|
|
const Register ExcludeReg) const {
|
|
MachineFunction *MF = I.getParent()->getParent();
|
|
MachineRegisterInfo *MRI = &MF->getRegInfo();
|
|
const TargetSubtargetInfo &ST = MF->getSubtarget();
|
|
const TargetRegisterInfo *TRI = ST.getRegisterInfo();
|
|
const TargetInstrInfo *TII = ST.getInstrInfo();
|
|
|
|
// The instruction is loop invariant if all of its operands are.
|
|
for (const MachineOperand &MO : I.operands()) {
|
|
if (!MO.isReg())
|
|
continue;
|
|
|
|
Register Reg = MO.getReg();
|
|
if (Reg == 0) continue;
|
|
|
|
if (ExcludeReg == Reg)
|
|
continue;
|
|
|
|
// An instruction that uses or defines a physical register can't e.g. be
|
|
// hoisted, so mark this as not invariant.
|
|
if (Reg.isPhysical()) {
|
|
if (MO.isUse()) {
|
|
// If the physreg has no defs anywhere, it's just an ambient register
|
|
// and we can freely move its uses. Alternatively, if it's allocatable,
|
|
// it could get allocated to something with a def during allocation.
|
|
// However, if the physreg is known to always be caller saved/restored
|
|
// then this use is safe to hoist.
|
|
if (!isLoopInvariantImplicitPhysReg(Reg) &&
|
|
!(TRI->isCallerPreservedPhysReg(Reg.asMCReg(), *I.getMF())) &&
|
|
!TII->isIgnorableUse(MO))
|
|
return false;
|
|
// Otherwise it's safe to move.
|
|
continue;
|
|
} else if (!MO.isDead()) {
|
|
// A def that isn't dead can't be moved.
|
|
return false;
|
|
} else if (getHeader()->isLiveIn(Reg)) {
|
|
// If the reg is live into the loop, we can't hoist an instruction
|
|
// which would clobber it.
|
|
return false;
|
|
}
|
|
}
|
|
|
|
if (!MO.isUse())
|
|
continue;
|
|
|
|
assert(MRI->getVRegDef(Reg) &&
|
|
"Machine instr not mapped for this vreg?!");
|
|
|
|
// If the loop contains the definition of an operand, then the instruction
|
|
// isn't loop invariant.
|
|
if (contains(MRI->getVRegDef(Reg)))
|
|
return false;
|
|
}
|
|
|
|
// If we got this far, the instruction is loop invariant!
|
|
return true;
|
|
}
|
|
|
|
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
|
|
LLVM_DUMP_METHOD void MachineLoop::dump() const {
|
|
print(dbgs());
|
|
}
|
|
#endif
|