blob: ea8e30269ece25fd2cf85c3d264bd1bc93fb06f9 [file] [log] [blame]
//===-- AArch64A57FPLoadBalancing.cpp - Balance FP ops statically on A57---===//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// For best-case performance on Cortex-A57, we should try to use a balanced
// mix of odd and even D-registers when performing a critical sequence of
// independent, non-quadword FP/ASIMD floating-point multiply or
// multiply-accumulate operations.
// This pass attempts to detect situations where the register allocation may
// adversely affect this load balancing and to change the registers used so as
// to better utilize the CPU.
// Ideally we'd just take each multiply or multiply-accumulate in turn and
// allocate it alternating even or odd registers. However, multiply-accumulates
// are most efficiently performed in the same functional unit as their
// accumulation operand. Therefore this pass tries to find maximal sequences
// ("Chains") of multiply-accumulates linked via their accumulation operand,
// and assign them all the same "color" (oddness/evenness).
// This optimization affects S-register and D-register floating point
// multiplies and FMADD/FMAs, as well as vector (floating point only) muls and
// FMADD/FMA. Q register instructions (and 128-bit vector instructions) are
// not affected.
#include "AArch64.h"
#include "AArch64InstrInfo.h"
#include "AArch64Subtarget.h"
#include "llvm/ADT/BitVector.h"
#include "llvm/ADT/EquivalenceClasses.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstr.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/RegisterClassInfo.h"
#include "llvm/CodeGen/RegisterScavenging.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/raw_ostream.h"
using namespace llvm;
#define DEBUG_TYPE "aarch64-a57-fp-load-balancing"
// Enforce the algorithm to use the scavenged register even when the original
// destination register is the correct color. Used for testing.
static cl::opt<bool>
cl::desc("Always modify dest registers regardless of color"),
cl::init(false), cl::Hidden);
// Never use the balance information obtained from chains - return a specific
// color always. Used for testing.
static cl::opt<unsigned>
cl::desc("Ignore balance information, always return "
"(1: Even, 2: Odd)."),
cl::init(0), cl::Hidden);
// Helper functions
// Is the instruction a type of multiply on 64-bit (or 32-bit) FPRs?
static bool isMul(MachineInstr *MI) {
switch (MI->getOpcode()) {
case AArch64::FMULSrr:
case AArch64::FNMULSrr:
case AArch64::FMULDrr:
case AArch64::FNMULDrr:
return true;
return false;
// Is the instruction a type of FP multiply-accumulate on 64-bit (or 32-bit) FPRs?
static bool isMla(MachineInstr *MI) {
switch (MI->getOpcode()) {
case AArch64::FMSUBSrrr:
case AArch64::FMADDSrrr:
case AArch64::FNMSUBSrrr:
case AArch64::FNMADDSrrr:
case AArch64::FMSUBDrrr:
case AArch64::FMADDDrrr:
case AArch64::FNMSUBDrrr:
case AArch64::FNMADDDrrr:
return true;
return false;
namespace {
/// A "color", which is either even or odd. Yes, these aren't really colors
/// but the algorithm is conceptually doing two-color graph coloring.
enum class Color { Even, Odd };
#ifndef NDEBUG
static const char *ColorNames[2] = { "Even", "Odd" };
class Chain;
class AArch64A57FPLoadBalancing : public MachineFunctionPass {
MachineRegisterInfo *MRI;
const TargetRegisterInfo *TRI;
RegisterClassInfo RCI;
static char ID;
explicit AArch64A57FPLoadBalancing() : MachineFunctionPass(ID) {
bool runOnMachineFunction(MachineFunction &F) override;
MachineFunctionProperties getRequiredProperties() const override {
return MachineFunctionProperties().set(
StringRef getPassName() const override {
return "A57 FP Anti-dependency breaker";
void getAnalysisUsage(AnalysisUsage &AU) const override {
bool runOnBasicBlock(MachineBasicBlock &MBB);
bool colorChainSet(std::vector<Chain*> GV, MachineBasicBlock &MBB,
int &Balance);
bool colorChain(Chain *G, Color C, MachineBasicBlock &MBB);
int scavengeRegister(Chain *G, Color C, MachineBasicBlock &MBB);
void scanInstruction(MachineInstr *MI, unsigned Idx,
std::map<unsigned, Chain*> &Active,
std::vector<std::unique_ptr<Chain>> &AllChains);
void maybeKillChain(MachineOperand &MO, unsigned Idx,
std::map<unsigned, Chain*> &RegChains);
Color getColor(unsigned Register);
Chain *getAndEraseNext(Color PreferredColor, std::vector<Chain*> &L);
char AArch64A57FPLoadBalancing::ID = 0;
"AArch64 A57 FP Load-Balancing", false, false)
"AArch64 A57 FP Load-Balancing", false, false)
namespace {
/// A Chain is a sequence of instructions that are linked together by
/// an accumulation operand. For example:
/// fmul def d0, ?
/// fmla def d1, ?, ?, killed d0
/// fmla def d2, ?, ?, killed d1
/// There may be other instructions interleaved in the sequence that
/// do not belong to the chain. These other instructions must not use
/// the "chain" register at any point.
/// We currently only support chains where the "chain" operand is killed
/// at each link in the chain for simplicity.
/// A chain has three important instructions - Start, Last and Kill.
/// * The start instruction is the first instruction in the chain.
/// * Last is the final instruction in the chain.
/// * Kill may or may not be defined. If defined, Kill is the instruction
/// where the outgoing value of the Last instruction is killed.
/// This information is important as if we know the outgoing value is
/// killed with no intervening uses, we can safely change its register.
/// Without a kill instruction, we must assume the outgoing value escapes
/// beyond our model and either must not change its register or must
/// create a fixup FMOV to keep the old register value consistent.
class Chain {
/// The important (marker) instructions.
MachineInstr *StartInst, *LastInst, *KillInst;
/// The index, from the start of the basic block, that each marker
/// appears. These are stored so we can do quick interval tests.
unsigned StartInstIdx, LastInstIdx, KillInstIdx;
/// All instructions in the chain.
std::set<MachineInstr*> Insts;
/// True if KillInst cannot be modified. If this is true,
/// we cannot change LastInst's outgoing register.
/// This will be true for tied values and regmasks.
bool KillIsImmutable;
/// The "color" of LastInst. This will be the preferred chain color,
/// as changing intermediate nodes is easy but changing the last
/// instruction can be more tricky.
Color LastColor;
Chain(MachineInstr *MI, unsigned Idx, Color C)
: StartInst(MI), LastInst(MI), KillInst(nullptr),
StartInstIdx(Idx), LastInstIdx(Idx), KillInstIdx(0),
LastColor(C) {
/// Add a new instruction into the chain. The instruction's dest operand
/// has the given color.
void add(MachineInstr *MI, unsigned Idx, Color C) {
LastInst = MI;
LastInstIdx = Idx;
LastColor = C;
assert((KillInstIdx == 0 || LastInstIdx < KillInstIdx) &&
"Chain: broken invariant. A Chain can only be killed after its last "
/// Return true if MI is a member of the chain.
bool contains(MachineInstr &MI) { return Insts.count(&MI) > 0; }
/// Return the number of instructions in the chain.
unsigned size() const {
return Insts.size();
/// Inform the chain that its last active register (the dest register of
/// LastInst) is killed by MI with no intervening uses or defs.
void setKill(MachineInstr *MI, unsigned Idx, bool Immutable) {
KillInst = MI;
KillInstIdx = Idx;
KillIsImmutable = Immutable;
assert((KillInstIdx == 0 || LastInstIdx < KillInstIdx) &&
"Chain: broken invariant. A Chain can only be killed after its last "
/// Return the first instruction in the chain.
MachineInstr *getStart() const { return StartInst; }
/// Return the last instruction in the chain.
MachineInstr *getLast() const { return LastInst; }
/// Return the "kill" instruction (as set with setKill()) or NULL.
MachineInstr *getKill() const { return KillInst; }
/// Return an instruction that can be used as an iterator for the end
/// of the chain. This is the maximum of KillInst (if set) and LastInst.
MachineBasicBlock::iterator end() const {
return ++MachineBasicBlock::iterator(KillInst ? KillInst : LastInst);
MachineBasicBlock::iterator begin() const { return getStart(); }
/// Can the Kill instruction (assuming one exists) be modified?
bool isKillImmutable() const { return KillIsImmutable; }
/// Return the preferred color of this chain.
Color getPreferredColor() {
if (OverrideBalance != 0)
return OverrideBalance == 1 ? Color::Even : Color::Odd;
return LastColor;
/// Return true if this chain (StartInst..KillInst) overlaps with Other.
bool rangeOverlapsWith(const Chain &Other) const {
unsigned End = KillInst ? KillInstIdx : LastInstIdx;
unsigned OtherEnd = Other.KillInst ?
Other.KillInstIdx : Other.LastInstIdx;
return StartInstIdx <= OtherEnd && Other.StartInstIdx <= End;
/// Return true if this chain starts before Other.
bool startsBefore(const Chain *Other) const {
return StartInstIdx < Other->StartInstIdx;
/// Return true if the group will require a fixup MOV at the end.
bool requiresFixup() const {
return (getKill() && isKillImmutable()) || !getKill();
/// Return a simple string representation of the chain.
std::string str() const {
std::string S;
raw_string_ostream OS(S);
OS << "{";
StartInst->print(OS, /* SkipOpers= */true);
OS << " -> ";
LastInst->print(OS, /* SkipOpers= */true);
if (KillInst) {
OS << " (kill @ ";
KillInst->print(OS, /* SkipOpers= */true);
OS << ")";
OS << "}";
return OS.str();
} // end anonymous namespace
bool AArch64A57FPLoadBalancing::runOnMachineFunction(MachineFunction &F) {
if (skipFunction(F.getFunction()))
return false;
if (!F.getSubtarget<AArch64Subtarget>().balanceFPOps())
return false;
bool Changed = false;
LLVM_DEBUG(dbgs() << "***** AArch64A57FPLoadBalancing *****\n");
MRI = &F.getRegInfo();
TRI = F.getRegInfo().getTargetRegisterInfo();
for (auto &MBB : F) {
Changed |= runOnBasicBlock(MBB);
return Changed;
bool AArch64A57FPLoadBalancing::runOnBasicBlock(MachineBasicBlock &MBB) {
bool Changed = false;
LLVM_DEBUG(dbgs() << "Running on MBB: " << MBB
<< " - scanning instructions...\n");
// First, scan the basic block producing a set of chains.
// The currently "active" chains - chains that can be added to and haven't
// been killed yet. This is keyed by register - all chains can only have one
// "link" register between each inst in the chain.
std::map<unsigned, Chain*> ActiveChains;
std::vector<std::unique_ptr<Chain>> AllChains;
unsigned Idx = 0;
for (auto &MI : MBB)
scanInstruction(&MI, Idx++, ActiveChains, AllChains);
LLVM_DEBUG(dbgs() << "Scan complete, " << AllChains.size()
<< " chains created.\n");
// Group the chains into disjoint sets based on their liveness range. This is
// a poor-man's version of graph coloring. Ideally we'd create an interference
// graph and perform full-on graph coloring on that, but;
// (a) That's rather heavyweight for only two colors.
// (b) We expect multiple disjoint interference regions - in practice the live
// range of chains is quite small and they are clustered between loads
// and stores.
EquivalenceClasses<Chain*> EC;
for (auto &I : AllChains)
for (auto &I : AllChains)
for (auto &J : AllChains)
if (I != J && I->rangeOverlapsWith(*J))
EC.unionSets(I.get(), J.get());
LLVM_DEBUG(dbgs() << "Created " << EC.getNumClasses() << " disjoint sets.\n");
// Now we assume that every member of an equivalence class interferes
// with every other member of that class, and with no members of other classes.
// Convert the EquivalenceClasses to a simpler set of sets.
std::vector<std::vector<Chain*> > V;
for (auto I = EC.begin(), E = EC.end(); I != E; ++I) {
std::vector<Chain*> Cs(EC.member_begin(I), EC.member_end());
if (Cs.empty()) continue;
// Now we have a set of sets, order them by start address so
// we can iterate over them sequentially.
[](const std::vector<Chain *> &A, const std::vector<Chain *> &B) {
return A.front()->startsBefore(B.front());
// As we only have two colors, we can track the global (BB-level) balance of
// odds versus evens. We aim to keep this near zero to keep both execution
// units fed.
// Positive means we're even-heavy, negative we're odd-heavy.
// FIXME: If chains have interdependencies, for example:
// mul r0, r1, r2
// mul r3, r0, r1
// We do not model this and may color each one differently, assuming we'll
// get ILP when we obviously can't. This hasn't been seen to be a problem
// in practice so far, so we simplify the algorithm by ignoring it.
int Parity = 0;
for (auto &I : V)
Changed |= colorChainSet(std::move(I), MBB, Parity);
return Changed;
Chain *AArch64A57FPLoadBalancing::getAndEraseNext(Color PreferredColor,
std::vector<Chain*> &L) {
if (L.empty())
return nullptr;
// We try and get the best candidate from L to color next, given that our
// preferred color is "PreferredColor". L is ordered from larger to smaller
// chains. It is beneficial to color the large chains before the small chains,
// but if we can't find a chain of the maximum length with the preferred color,
// we fuzz the size and look for slightly smaller chains before giving up and
// returning a chain that must be recolored.
// FIXME: Does this need to be configurable?
const unsigned SizeFuzz = 1;
unsigned MinSize = L.front()->size() - SizeFuzz;
for (auto I = L.begin(), E = L.end(); I != E; ++I) {
if ((*I)->size() <= MinSize) {
// We've gone past the size limit. Return the previous item.
Chain *Ch = *--I;
return Ch;
if ((*I)->getPreferredColor() == PreferredColor) {
Chain *Ch = *I;
return Ch;
// Bailout case - just return the first item.
Chain *Ch = L.front();
return Ch;
bool AArch64A57FPLoadBalancing::colorChainSet(std::vector<Chain*> GV,
MachineBasicBlock &MBB,
int &Parity) {
bool Changed = false;
LLVM_DEBUG(dbgs() << "colorChainSet(): #sets=" << GV.size() << "\n");
// Sort by descending size order so that we allocate the most important
// sets first.
// Tie-break equivalent sizes by sorting chains requiring fixups before
// those without fixups. The logic here is that we should look at the
// chains that we cannot change before we look at those we can,
// so the parity counter is updated and we know what color we should
// change them to!
// Final tie-break with instruction order so pass output is stable (i.e. not
// dependent on malloc'd pointer values).
llvm::sort(GV, [](const Chain *G1, const Chain *G2) {
if (G1->size() != G2->size())
return G1->size() > G2->size();
if (G1->requiresFixup() != G2->requiresFixup())
return G1->requiresFixup() > G2->requiresFixup();
// Make sure startsBefore() produces a stable final order.
assert((G1 == G2 || (G1->startsBefore(G2) ^ G2->startsBefore(G1))) &&
"Starts before not total order!");
return G1->startsBefore(G2);
Color PreferredColor = Parity < 0 ? Color::Even : Color::Odd;
while (Chain *G = getAndEraseNext(PreferredColor, GV)) {
// Start off by assuming we'll color to our own preferred color.
Color C = PreferredColor;
if (Parity == 0)
// But if we really don't care, use the chain's preferred color.
C = G->getPreferredColor();
LLVM_DEBUG(dbgs() << " - Parity=" << Parity
<< ", Color=" << ColorNames[(int)C] << "\n");
// If we'll need a fixup FMOV, don't bother. Testing has shown that this
// happens infrequently and when it does it has at least a 50% chance of
// slowing code down instead of speeding it up.
if (G->requiresFixup() && C != G->getPreferredColor()) {
C = G->getPreferredColor();
LLVM_DEBUG(dbgs() << " - " << G->str()
<< " - not worthwhile changing; "
"color remains "
<< ColorNames[(int)C] << "\n");
Changed |= colorChain(G, C, MBB);
Parity += (C == Color::Even) ? G->size() : -G->size();
PreferredColor = Parity < 0 ? Color::Even : Color::Odd;
return Changed;
int AArch64A57FPLoadBalancing::scavengeRegister(Chain *G, Color C,
MachineBasicBlock &MBB) {
// Can we find an appropriate register that is available throughout the life
// of the chain? Simulate liveness backwards until the end of the chain.
LiveRegUnits Units(*TRI);
MachineBasicBlock::iterator I = MBB.end();
MachineBasicBlock::iterator ChainEnd = G->end();
while (I != ChainEnd) {
// Check which register units are alive throughout the chain.
MachineBasicBlock::iterator ChainBegin = G->begin();
assert(ChainBegin != ChainEnd && "Chain should contain instructions");
do {
} while (I != ChainBegin);
// Make sure we allocate in-order, to get the cheapest registers first.
unsigned RegClassID = ChainBegin->getDesc().operands()[0].RegClass;
auto Ord = RCI.getOrder(TRI->getRegClass(RegClassID));
for (auto Reg : Ord) {
if (!Units.available(Reg))
if (C == getColor(Reg))
return Reg;
return -1;
bool AArch64A57FPLoadBalancing::colorChain(Chain *G, Color C,
MachineBasicBlock &MBB) {
bool Changed = false;
LLVM_DEBUG(dbgs() << " - colorChain(" << G->str() << ", "
<< ColorNames[(int)C] << ")\n");
// Try and obtain a free register of the right class. Without a register
// to play with we cannot continue.
int Reg = scavengeRegister(G, C, MBB);
if (Reg == -1) {
LLVM_DEBUG(dbgs() << "Scavenging (thus coloring) failed!\n");
return false;
LLVM_DEBUG(dbgs() << " - Scavenged register: " << printReg(Reg, TRI) << "\n");
std::map<unsigned, unsigned> Substs;
for (MachineInstr &I : *G) {
if (!G->contains(I) && (&I != G->getKill() || G->isKillImmutable()))
// I is a member of G, or I is a mutable instruction that kills G.
std::vector<unsigned> ToErase;
for (auto &U : I.operands()) {
if (U.isReg() && U.isUse() && Substs.find(U.getReg()) != Substs.end()) {
Register OrigReg = U.getReg();
if (U.isKill())
// Don't erase straight away, because there may be other operands
// that also reference this substitution!
} else if (U.isRegMask()) {
for (auto J : Substs) {
if (U.clobbersPhysReg(J.first))
// Now it's safe to remove the substs identified earlier.
for (auto J : ToErase)
// Only change the def if this isn't the last instruction.
if (&I != G->getKill()) {
MachineOperand &MO = I.getOperand(0);
bool Change = TransformAll || getColor(MO.getReg()) != C;
if (G->requiresFixup() && &I == G->getLast())
Change = false;
if (Change) {
Substs[MO.getReg()] = Reg;
Changed = true;
assert(Substs.size() == 0 && "No substitutions should be left active!");
if (G->getKill()) {
LLVM_DEBUG(dbgs() << " - Kill instruction seen.\n");
} else {
// We didn't have a kill instruction, but we didn't seem to need to change
// the destination register anyway.
LLVM_DEBUG(dbgs() << " - Destination register not changed.\n");
return Changed;
void AArch64A57FPLoadBalancing::scanInstruction(
MachineInstr *MI, unsigned Idx, std::map<unsigned, Chain *> &ActiveChains,
std::vector<std::unique_ptr<Chain>> &AllChains) {
// Inspect "MI", updating ActiveChains and AllChains.
if (isMul(MI)) {
for (auto &I : MI->uses())
maybeKillChain(I, Idx, ActiveChains);
for (auto &I : MI->defs())
maybeKillChain(I, Idx, ActiveChains);
// Create a new chain. Multiplies don't require forwarding so can go on any
// unit.
Register DestReg = MI->getOperand(0).getReg();
LLVM_DEBUG(dbgs() << "New chain started for register "
<< printReg(DestReg, TRI) << " at " << *MI);
auto G = std::make_unique<Chain>(MI, Idx, getColor(DestReg));
ActiveChains[DestReg] = G.get();
} else if (isMla(MI)) {
// It is beneficial to keep MLAs on the same functional unit as their
// accumulator operand.
Register DestReg = MI->getOperand(0).getReg();
Register AccumReg = MI->getOperand(3).getReg();
maybeKillChain(MI->getOperand(1), Idx, ActiveChains);
maybeKillChain(MI->getOperand(2), Idx, ActiveChains);
if (DestReg != AccumReg)
maybeKillChain(MI->getOperand(0), Idx, ActiveChains);
if (ActiveChains.find(AccumReg) != ActiveChains.end()) {
LLVM_DEBUG(dbgs() << "Chain found for accumulator register "
<< printReg(AccumReg, TRI) << " in MI " << *MI);
// For simplicity we only chain together sequences of MULs/MLAs where the
// accumulator register is killed on each instruction. This means we don't
// need to track other uses of the registers we want to rewrite.
// FIXME: We could extend to handle the non-kill cases for more coverage.
if (MI->getOperand(3).isKill()) {
// Add to chain.
LLVM_DEBUG(dbgs() << "Instruction was successfully added to chain.\n");
ActiveChains[AccumReg]->add(MI, Idx, getColor(DestReg));
// Handle cases where the destination is not the same as the accumulator.
if (DestReg != AccumReg) {
ActiveChains[DestReg] = ActiveChains[AccumReg];
dbgs() << "Cannot add to chain because accumulator operand wasn't "
<< "marked <kill>!\n");
maybeKillChain(MI->getOperand(3), Idx, ActiveChains);
LLVM_DEBUG(dbgs() << "Creating new chain for dest register "
<< printReg(DestReg, TRI) << "\n");
auto G = std::make_unique<Chain>(MI, Idx, getColor(DestReg));
ActiveChains[DestReg] = G.get();
} else {
// Non-MUL or MLA instruction. Invalidate any chain in the uses or defs
// lists.
for (auto &I : MI->uses())
maybeKillChain(I, Idx, ActiveChains);
for (auto &I : MI->defs())
maybeKillChain(I, Idx, ActiveChains);
void AArch64A57FPLoadBalancing::
maybeKillChain(MachineOperand &MO, unsigned Idx,
std::map<unsigned, Chain*> &ActiveChains) {
// Given an operand and the set of active chains (keyed by register),
// determine if a chain should be ended and remove from ActiveChains.
MachineInstr *MI = MO.getParent();
if (MO.isReg()) {
// If this is a KILL of a current chain, record it.
if (MO.isKill() && ActiveChains.find(MO.getReg()) != ActiveChains.end()) {
LLVM_DEBUG(dbgs() << "Kill seen for chain " << printReg(MO.getReg(), TRI)
<< "\n");
ActiveChains[MO.getReg()]->setKill(MI, Idx, /*Immutable=*/MO.isTied());
} else if (MO.isRegMask()) {
for (auto I = ActiveChains.begin(), E = ActiveChains.end();
I != E;) {
if (MO.clobbersPhysReg(I->first)) {
LLVM_DEBUG(dbgs() << "Kill (regmask) seen for chain "
<< printReg(I->first, TRI) << "\n");
I->second->setKill(MI, Idx, /*Immutable=*/true);
} else
Color AArch64A57FPLoadBalancing::getColor(unsigned Reg) {
if ((TRI->getEncodingValue(Reg) % 2) == 0)
return Color::Even;
return Color::Odd;
// Factory function used by AArch64TargetMachine to add the pass to the passmanager.
FunctionPass *llvm::createAArch64A57FPLoadBalancing() {
return new AArch64A57FPLoadBalancing();