blob: 4c4d196427e2ff5372baa200cb57f46b689251ea [file] [log] [blame]
//===- SelectionDAGISel.cpp - Implement the SelectionDAGISel class --------===//
// The LLVM Compiler Infrastructure
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
// This implements the SelectionDAGISel class.
#include "ScheduleDAGSDNodes.h"
#include "SelectionDAGBuilder.h"
#include "llvm/ADT/APInt.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/None.h"
#include "llvm/ADT/PostOrderIterator.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/BranchProbabilityInfo.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/OptimizationRemarkEmitter.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/CodeGen/FastISel.h"
#include "llvm/CodeGen/FunctionLoweringInfo.h"
#include "llvm/CodeGen/GCMetadata.h"
#include "llvm/CodeGen/ISDOpcodes.h"
#include "llvm/CodeGen/MachineBasicBlock.h"
#include "llvm/CodeGen/MachineFrameInfo.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstr.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineMemOperand.h"
#include "llvm/CodeGen/MachineOperand.h"
#include "llvm/CodeGen/MachinePassRegistry.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/MachineValueType.h"
#include "llvm/CodeGen/SchedulerRegistry.h"
#include "llvm/CodeGen/SelectionDAG.h"
#include "llvm/CodeGen/SelectionDAGISel.h"
#include "llvm/CodeGen/SelectionDAGNodes.h"
#include "llvm/CodeGen/StackProtector.h"
#include "llvm/CodeGen/ValueTypes.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DebugInfoMetadata.h"
#include "llvm/IR/DebugLoc.h"
#include "llvm/IR/DiagnosticInfo.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/InstrTypes.h"
#include "llvm/IR/Instruction.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/Metadata.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/User.h"
#include "llvm/IR/Value.h"
#include "llvm/MC/MCInstrDesc.h"
#include "llvm/MC/MCRegisterInfo.h"
#include "llvm/Pass.h"
#include "llvm/Support/BranchProbability.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/CodeGen.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Compiler.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/KnownBits.h"
#include "llvm/Support/Timer.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetInstrInfo.h"
#include "llvm/Target/TargetIntrinsicInfo.h"
#include "llvm/Target/TargetLowering.h"
#include "llvm/Target/TargetMachine.h"
#include "llvm/Target/TargetOptions.h"
#include "llvm/Target/TargetRegisterInfo.h"
#include "llvm/Target/TargetSubtargetInfo.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include <algorithm>
#include <cassert>
#include <cstdint>
#include <iterator>
#include <limits>
#include <memory>
#include <string>
#include <utility>
#include <vector>
using namespace llvm;
#define DEBUG_TYPE "isel"
STATISTIC(NumFastIselFailures, "Number of instructions fast isel failed on");
STATISTIC(NumFastIselSuccess, "Number of instructions fast isel selected");
STATISTIC(NumFastIselBlocks, "Number of blocks selected entirely by fast isel");
STATISTIC(NumDAGBlocks, "Number of blocks selected using DAG");
STATISTIC(NumDAGIselRetries,"Number of times dag isel has to try another path");
STATISTIC(NumEntryBlocks, "Number of entry blocks encountered");
"Number of entry blocks where fast isel failed to lower arguments");
static cl::opt<int> EnableFastISelAbort(
"fast-isel-abort", cl::Hidden,
cl::desc("Enable abort calls when \"fast\" instruction selection "
"fails to lower an instruction: 0 disable the abort, 1 will "
"abort but for args, calls and terminators, 2 will also "
"abort for argument lowering, and 3 will never fallback "
"to SelectionDAG."));
static cl::opt<bool> EnableFastISelFallbackReport(
"fast-isel-report-on-fallback", cl::Hidden,
cl::desc("Emit a diagnostic when \"fast\" instruction selection "
"falls back to SelectionDAG."));
static cl::opt<bool>
cl::desc("use Machine Branch Probability Info"),
cl::init(true), cl::Hidden);
#ifndef NDEBUG
static cl::opt<std::string>
FilterDAGBasicBlockName("filter-view-dags", cl::Hidden,
cl::desc("Only display the basic block whose name "
"matches this for all view-*-dags options"));
static cl::opt<bool>
ViewDAGCombine1("view-dag-combine1-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before the first "
"dag combine pass"));
static cl::opt<bool>
ViewLegalizeTypesDAGs("view-legalize-types-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before legalize types"));
static cl::opt<bool>
ViewLegalizeDAGs("view-legalize-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before legalize"));
static cl::opt<bool>
ViewDAGCombine2("view-dag-combine2-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before the second "
"dag combine pass"));
static cl::opt<bool>
ViewDAGCombineLT("view-dag-combine-lt-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before the post legalize types"
" dag combine pass"));
static cl::opt<bool>
ViewISelDAGs("view-isel-dags", cl::Hidden,
cl::desc("Pop up a window to show isel dags as they are selected"));
static cl::opt<bool>
ViewSchedDAGs("view-sched-dags", cl::Hidden,
cl::desc("Pop up a window to show sched dags as they are processed"));
static cl::opt<bool>
ViewSUnitDAGs("view-sunit-dags", cl::Hidden,
cl::desc("Pop up a window to show SUnit dags after they are processed"));
static const bool ViewDAGCombine1 = false,
ViewLegalizeTypesDAGs = false, ViewLegalizeDAGs = false,
ViewDAGCombine2 = false,
ViewDAGCombineLT = false,
ViewISelDAGs = false, ViewSchedDAGs = false,
ViewSUnitDAGs = false;
/// RegisterScheduler class - Track the registration of instruction schedulers.
MachinePassRegistry RegisterScheduler::Registry;
/// ISHeuristic command line option for instruction schedulers.
static cl::opt<RegisterScheduler::FunctionPassCtor, false,
cl::init(&createDefaultScheduler), cl::Hidden,
cl::desc("Instruction schedulers available (before register"
" allocation):"));
static RegisterScheduler
defaultListDAGScheduler("default", "Best scheduler for the target",
namespace llvm {
/// \brief This class is used by SelectionDAGISel to temporarily override
/// the optimization level on a per-function basis.
class OptLevelChanger {
SelectionDAGISel &IS;
CodeGenOpt::Level SavedOptLevel;
bool SavedFastISel;
OptLevelChanger(SelectionDAGISel &ISel,
CodeGenOpt::Level NewOptLevel) : IS(ISel) {
SavedOptLevel = IS.OptLevel;
if (NewOptLevel == SavedOptLevel)
IS.OptLevel = NewOptLevel;
DEBUG(dbgs() << "\nChanging optimization level for Function "
<< IS.MF->getFunction()->getName() << "\n");
DEBUG(dbgs() << "\tBefore: -O" << SavedOptLevel
<< " ; After: -O" << NewOptLevel << "\n");
SavedFastISel = IS.TM.Options.EnableFastISel;
if (NewOptLevel == CodeGenOpt::None) {
DEBUG(dbgs() << "\tFastISel is "
<< (IS.TM.Options.EnableFastISel ? "enabled" : "disabled")
<< "\n");
~OptLevelChanger() {
if (IS.OptLevel == SavedOptLevel)
DEBUG(dbgs() << "\nRestoring optimization level for Function "
<< IS.MF->getFunction()->getName() << "\n");
DEBUG(dbgs() << "\tBefore: -O" << IS.OptLevel
<< " ; After: -O" << SavedOptLevel << "\n");
IS.OptLevel = SavedOptLevel;
/// createDefaultScheduler - This creates an instruction scheduler appropriate
/// for the target.
ScheduleDAGSDNodes* createDefaultScheduler(SelectionDAGISel *IS,
CodeGenOpt::Level OptLevel) {
const TargetLowering *TLI = IS->TLI;
const TargetSubtargetInfo &ST = IS->MF->getSubtarget();
// Try first to see if the Target has its own way of selecting a scheduler
if (auto *SchedulerCtor = ST.getDAGScheduler(OptLevel)) {
return SchedulerCtor(IS, OptLevel);
if (OptLevel == CodeGenOpt::None ||
(ST.enableMachineScheduler() && ST.enableMachineSchedDefaultSched()) ||
TLI->getSchedulingPreference() == Sched::Source)
return createSourceListDAGScheduler(IS, OptLevel);
if (TLI->getSchedulingPreference() == Sched::RegPressure)
return createBURRListDAGScheduler(IS, OptLevel);
if (TLI->getSchedulingPreference() == Sched::Hybrid)
return createHybridListDAGScheduler(IS, OptLevel);
if (TLI->getSchedulingPreference() == Sched::VLIW)
return createVLIWDAGScheduler(IS, OptLevel);
assert(TLI->getSchedulingPreference() == Sched::ILP &&
"Unknown sched type!");
return createILPListDAGScheduler(IS, OptLevel);
} // end namespace llvm
// EmitInstrWithCustomInserter - This method should be implemented by targets
// that mark instructions with the 'usesCustomInserter' flag. These
// instructions are special in various ways, which require special support to
// insert. The specified MachineInstr is created but not inserted into any
// basic blocks, and this method is called to expand it into a sequence of
// instructions, potentially also creating new basic blocks and control flow.
// When new basic blocks are inserted and the edges from MBB to its successors
// are modified, the method should insert pairs of <OldSucc, NewSucc> into the
// DenseMap.
MachineBasicBlock *
TargetLowering::EmitInstrWithCustomInserter(MachineInstr &MI,
MachineBasicBlock *MBB) const {
#ifndef NDEBUG
dbgs() << "If a target marks an instruction with "
"'usesCustomInserter', it must implement "
void TargetLowering::AdjustInstrPostInstrSelection(MachineInstr &MI,
SDNode *Node) const {
assert(!MI.hasPostISelHook() &&
"If a target marks an instruction with 'hasPostISelHook', "
"it must implement TargetLowering::AdjustInstrPostInstrSelection!");
// SelectionDAGISel code
SelectionDAGISel::SelectionDAGISel(TargetMachine &tm,
CodeGenOpt::Level OL) :
MachineFunctionPass(ID), TM(tm),
FuncInfo(new FunctionLoweringInfo()),
CurDAG(new SelectionDAG(tm, OL)),
SDB(new SelectionDAGBuilder(*CurDAG, *FuncInfo, OL)),
AA(), GFI(),
DAGSize(0) {
SelectionDAGISel::~SelectionDAGISel() {
delete SDB;
delete CurDAG;
delete FuncInfo;
void SelectionDAGISel::getAnalysisUsage(AnalysisUsage &AU) const {
if (OptLevel != CodeGenOpt::None)
if (UseMBPI && OptLevel != CodeGenOpt::None)
/// SplitCriticalSideEffectEdges - Look for critical edges with a PHI value that
/// may trap on it. In this case we have to split the edge so that the path
/// through the predecessor block that doesn't go to the phi block doesn't
/// execute the possibly trapping instruction. If available, we pass domtree
/// and loop info to be updated when we split critical edges. This is because
/// SelectionDAGISel preserves these analyses.
/// This is required for correctness, so it must be done at -O0.
static void SplitCriticalSideEffectEdges(Function &Fn, DominatorTree *DT,
LoopInfo *LI) {
// Loop for blocks with phi nodes.
for (BasicBlock &BB : Fn) {
PHINode *PN = dyn_cast<PHINode>(BB.begin());
if (!PN) continue;
// For each block with a PHI node, check to see if any of the input values
// are potentially trapping constant expressions. Constant expressions are
// the only potentially trapping value that can occur as the argument to a
// PHI.
for (BasicBlock::iterator I = BB.begin(); (PN = dyn_cast<PHINode>(I)); ++I)
for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i) {
ConstantExpr *CE = dyn_cast<ConstantExpr>(PN->getIncomingValue(i));
if (!CE || !CE->canTrap()) continue;
// The only case we have to worry about is when the edge is critical.
// Since this block has a PHI Node, we assume it has multiple input
// edges: check to see if the pred has multiple successors.
BasicBlock *Pred = PN->getIncomingBlock(i);
if (Pred->getTerminator()->getNumSuccessors() == 1)
// Okay, we have to split this edge.
Pred->getTerminator(), GetSuccessorNumber(Pred, &BB),
CriticalEdgeSplittingOptions(DT, LI).setMergeIdenticalEdges());
goto ReprocessBlock;
bool SelectionDAGISel::runOnMachineFunction(MachineFunction &mf) {
// If we already selected that function, we do not need to run SDISel.
if (mf.getProperties().hasProperty(
return false;
// Do some sanity-checking on the command-line options.
assert((!EnableFastISelAbort || TM.Options.EnableFastISel) &&
"-fast-isel-abort > 0 requires -fast-isel");
const Function &Fn = *mf.getFunction();
MF = &mf;
// Reset the target options before resetting the optimization
// level below.
// FIXME: This is a horrible hack and should be processed via
// codegen looking at the optimization level explicitly when
// it wants to look at it.
// Reset OptLevel to None for optnone functions.
CodeGenOpt::Level NewOptLevel = OptLevel;
if (OptLevel != CodeGenOpt::None && skipFunction(Fn))
NewOptLevel = CodeGenOpt::None;
OptLevelChanger OLC(*this, NewOptLevel);
TII = MF->getSubtarget().getInstrInfo();
TLI = MF->getSubtarget().getTargetLowering();
RegInfo = &MF->getRegInfo();
LibInfo = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
GFI = Fn.hasGC() ? &getAnalysis<GCModuleInfo>().getFunctionInfo(Fn) : nullptr;
ORE = make_unique<OptimizationRemarkEmitter>(&Fn);
auto *DTWP = getAnalysisIfAvailable<DominatorTreeWrapperPass>();
DominatorTree *DT = DTWP ? &DTWP->getDomTree() : nullptr;
auto *LIWP = getAnalysisIfAvailable<LoopInfoWrapperPass>();
LoopInfo *LI = LIWP ? &LIWP->getLoopInfo() : nullptr;
DEBUG(dbgs() << "\n\n\n=== " << Fn.getName() << "\n");
SplitCriticalSideEffectEdges(const_cast<Function &>(Fn), DT, LI);
CurDAG->init(*MF, *ORE, this);
FuncInfo->set(Fn, *MF, CurDAG);
// Now get the optional analyzes if we want to.
// This is based on the possibly changed OptLevel (after optnone is taken
// into account). That's unfortunate but OK because it just means we won't
// ask for passes that have been required anyway.
if (UseMBPI && OptLevel != CodeGenOpt::None)
FuncInfo->BPI = &getAnalysis<BranchProbabilityInfoWrapperPass>().getBPI();
FuncInfo->BPI = nullptr;
if (OptLevel != CodeGenOpt::None)
AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
AA = nullptr;
SDB->init(GFI, AA, LibInfo);
FuncInfo->SplitCSR = false;
// We split CSR if the target supports it for the given function
// and the function has only return exits.
if (OptLevel != CodeGenOpt::None && TLI->supportSplitCSR(MF)) {
FuncInfo->SplitCSR = true;
// Collect all the return blocks.
for (const BasicBlock &BB : Fn) {
if (!succ_empty(&BB))
const TerminatorInst *Term = BB.getTerminator();
if (isa<UnreachableInst>(Term) || isa<ReturnInst>(Term))
// Bail out if the exit block is not Return nor Unreachable.
FuncInfo->SplitCSR = false;
MachineBasicBlock *EntryMBB = &MF->front();
if (FuncInfo->SplitCSR)
// This performs initialization so lowering for SplitCSR will be correct.
if (FastISelFailed && EnableFastISelFallbackReport) {
DiagnosticInfoISelFallback DiagFallback(Fn);
// If the first basic block in the function has live ins that need to be
// copied into vregs, emit the copies into the top of the block before
// emitting the code for the block.
const TargetRegisterInfo &TRI = *MF->getSubtarget().getRegisterInfo();
RegInfo->EmitLiveInCopies(EntryMBB, TRI, *TII);
// Insert copies in the entry block and the return blocks.
if (FuncInfo->SplitCSR) {
SmallVector<MachineBasicBlock*, 4> Returns;
// Collect all the return blocks.
for (MachineBasicBlock &MBB : mf) {
if (!MBB.succ_empty())
MachineBasicBlock::iterator Term = MBB.getFirstTerminator();
if (Term != MBB.end() && Term->isReturn()) {
TLI->insertCopiesSplitCSR(EntryMBB, Returns);
DenseMap<unsigned, unsigned> LiveInMap;
if (!FuncInfo->ArgDbgValues.empty())
for (std::pair<unsigned, unsigned> LI : RegInfo->liveins())
if (LI.second)
// Insert DBG_VALUE instructions for function arguments to the entry block.
for (unsigned i = 0, e = FuncInfo->ArgDbgValues.size(); i != e; ++i) {
MachineInstr *MI = FuncInfo->ArgDbgValues[e-i-1];
bool hasFI = MI->getOperand(0).isFI();
unsigned Reg =
hasFI ? TRI.getFrameRegister(*MF) : MI->getOperand(0).getReg();
if (TargetRegisterInfo::isPhysicalRegister(Reg))
EntryMBB->insert(EntryMBB->begin(), MI);
else {
MachineInstr *Def = RegInfo->getVRegDef(Reg);
if (Def) {
MachineBasicBlock::iterator InsertPos = Def;
// FIXME: VR def may not be in entry block.
Def->getParent()->insert(std::next(InsertPos), MI);
} else
DEBUG(dbgs() << "Dropping debug info for dead vreg"
<< TargetRegisterInfo::virtReg2Index(Reg) << "\n");
// If Reg is live-in then update debug info to track its copy in a vreg.
DenseMap<unsigned, unsigned>::iterator LDI = LiveInMap.find(Reg);
if (LDI != LiveInMap.end()) {
assert(!hasFI && "There's no handling of frame pointer updating here yet "
"- add if needed");
MachineInstr *Def = RegInfo->getVRegDef(LDI->second);
MachineBasicBlock::iterator InsertPos = Def;
const MDNode *Variable = MI->getDebugVariable();
const MDNode *Expr = MI->getDebugExpression();
DebugLoc DL = MI->getDebugLoc();
bool IsIndirect = MI->isIndirectDebugValue();
if (IsIndirect)
assert(MI->getOperand(1).getImm() == 0 &&
"DBG_VALUE with nonzero offset");
assert(cast<DILocalVariable>(Variable)->isValidLocationForIntrinsic(DL) &&
"Expected inlined-at fields to agree");
// Def is never a terminator here, so it is ok to increment InsertPos.
BuildMI(*EntryMBB, ++InsertPos, DL, TII->get(TargetOpcode::DBG_VALUE),
IsIndirect, LDI->second, Variable, Expr);
// If this vreg is directly copied into an exported register then
// that COPY instructions also need DBG_VALUE, if it is the only
// user of LDI->second.
MachineInstr *CopyUseMI = nullptr;
for (MachineRegisterInfo::use_instr_iterator
UI = RegInfo->use_instr_begin(LDI->second),
E = RegInfo->use_instr_end(); UI != E; ) {
MachineInstr *UseMI = &*(UI++);
if (UseMI->isDebugValue()) continue;
if (UseMI->isCopy() && !CopyUseMI && UseMI->getParent() == EntryMBB) {
CopyUseMI = UseMI; continue;
// Otherwise this is another use or second copy use.
CopyUseMI = nullptr; break;
if (CopyUseMI) {
// Use MI's debug location, which describes where Variable was
// declared, rather than whatever is attached to CopyUseMI.
MachineInstr *NewMI =
BuildMI(*MF, DL, TII->get(TargetOpcode::DBG_VALUE), IsIndirect,
CopyUseMI->getOperand(0).getReg(), Variable, Expr);
MachineBasicBlock::iterator Pos = CopyUseMI;
EntryMBB->insertAfter(Pos, NewMI);
// Determine if there are any calls in this machine function.
MachineFrameInfo &MFI = MF->getFrameInfo();
for (const auto &MBB : *MF) {
if (MFI.hasCalls() && MF->hasInlineAsm())
for (const auto &MI : MBB) {
const MCInstrDesc &MCID = TII->get(MI.getOpcode());
if ((MCID.isCall() && !MCID.isReturn()) ||
MI.isStackAligningInlineAsm()) {
if (MI.isInlineAsm()) {
// Determine if there is a call to setjmp in the machine function.
// Replace forward-declared registers with the registers containing
// the desired value.
MachineRegisterInfo &MRI = MF->getRegInfo();
for (DenseMap<unsigned, unsigned>::iterator
I = FuncInfo->RegFixups.begin(), E = FuncInfo->RegFixups.end();
I != E; ++I) {
unsigned From = I->first;
unsigned To = I->second;
// If To is also scheduled to be replaced, find what its ultimate
// replacement is.
while (true) {
DenseMap<unsigned, unsigned>::iterator J = FuncInfo->RegFixups.find(To);
if (J == E) break;
To = J->second;
// Make sure the new register has a sufficiently constrained register class.
if (TargetRegisterInfo::isVirtualRegister(From) &&
MRI.constrainRegClass(To, MRI.getRegClass(From));
// Replace it.
// Replacing one register with another won't touch the kill flags.
// We need to conservatively clear the kill flags as a kill on the old
// register might dominate existing uses of the new register.
if (!MRI.use_empty(To))
MRI.replaceRegWith(From, To);
// Release function-specific state. SDB and CurDAG are already cleared
// at this point.
DEBUG(dbgs() << "*** MachineFunction at end of ISel ***\n");
return true;
static void reportFastISelFailure(MachineFunction &MF,
OptimizationRemarkEmitter &ORE,
OptimizationRemarkMissed &R,
bool ShouldAbort) {
// Print the function name explicitly if we don't have a debug location (which
// makes the diagnostic less useful) or if we're going to emit a raw error.
if (!R.getLocation().isValid() || ShouldAbort)
R << (" (in function: " + MF.getName() + ")").str();
if (ShouldAbort)
void SelectionDAGISel::SelectBasicBlock(BasicBlock::const_iterator Begin,
BasicBlock::const_iterator End,
bool &HadTailCall) {
// Allow creating illegal types during DAG building for the basic block.
CurDAG->NewNodesMustHaveLegalTypes = false;
// Lower the instructions. If a call is emitted as a tail call, cease emitting
// nodes for this block.
for (BasicBlock::const_iterator I = Begin; I != End && !SDB->HasTailCall; ++I) {
if (!ElidedArgCopyInstrs.count(&*I))
// Make sure the root of the DAG is up-to-date.
HadTailCall = SDB->HasTailCall;
// Final step, emit the lowered DAG as machine code.
void SelectionDAGISel::ComputeLiveOutVRegInfo() {
SmallPtrSet<SDNode*, 16> VisitedNodes;
SmallVector<SDNode*, 128> Worklist;
KnownBits Known;
do {
SDNode *N = Worklist.pop_back_val();
// If we've already seen this node, ignore it.
if (!VisitedNodes.insert(N).second)
// Otherwise, add all chain operands to the worklist.
for (const SDValue &Op : N->op_values())
if (Op.getValueType() == MVT::Other)
// If this is a CopyToReg with a vreg dest, process it.
if (N->getOpcode() != ISD::CopyToReg)
unsigned DestReg = cast<RegisterSDNode>(N->getOperand(1))->getReg();
if (!TargetRegisterInfo::isVirtualRegister(DestReg))
// Ignore non-scalar or non-integer values.
SDValue Src = N->getOperand(2);
EVT SrcVT = Src.getValueType();
if (!SrcVT.isInteger() || SrcVT.isVector())
unsigned NumSignBits = CurDAG->ComputeNumSignBits(Src);
CurDAG->computeKnownBits(Src, Known);
FuncInfo->AddLiveOutRegInfo(DestReg, NumSignBits, Known);
} while (!Worklist.empty());
void SelectionDAGISel::CodeGenAndEmitDAG() {
StringRef GroupName = "sdag";
StringRef GroupDescription = "Instruction Selection and Scheduling";
std::string BlockName;
int BlockNumber = -1;
bool MatchFilterBB = false; (void)MatchFilterBB;
// Pre-type legalization allow creation of any node types.
CurDAG->NewNodesMustHaveLegalTypes = false;
#ifndef NDEBUG
MatchFilterBB = (FilterDAGBasicBlockName.empty() ||
FilterDAGBasicBlockName ==
#ifdef NDEBUG
if (ViewDAGCombine1 || ViewLegalizeTypesDAGs || ViewLegalizeDAGs ||
ViewDAGCombine2 || ViewDAGCombineLT || ViewISelDAGs || ViewSchedDAGs ||
BlockNumber = FuncInfo->MBB->getNumber();
BlockName =
(MF->getName() + ":" + FuncInfo->MBB->getBasicBlock()->getName()).str();
DEBUG(dbgs() << "Initial selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
if (ViewDAGCombine1 && MatchFilterBB)
CurDAG->viewGraph("dag-combine1 input for " + BlockName);
// Run the DAG combiner in pre-legalize mode.
NamedRegionTimer T("combine1", "DAG Combining 1", GroupName,
GroupDescription, TimePassesIsEnabled);
CurDAG->Combine(BeforeLegalizeTypes, AA, OptLevel);
DEBUG(dbgs() << "Optimized lowered selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
// Second step, hack on the DAG until it only uses operations and types that
// the target supports.
if (ViewLegalizeTypesDAGs && MatchFilterBB)
CurDAG->viewGraph("legalize-types input for " + BlockName);
bool Changed;
NamedRegionTimer T("legalize_types", "Type Legalization", GroupName,
GroupDescription, TimePassesIsEnabled);
Changed = CurDAG->LegalizeTypes();
DEBUG(dbgs() << "Type-legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
// Only allow creation of legal node types.
CurDAG->NewNodesMustHaveLegalTypes = true;
if (Changed) {
if (ViewDAGCombineLT && MatchFilterBB)
CurDAG->viewGraph("dag-combine-lt input for " + BlockName);
// Run the DAG combiner in post-type-legalize mode.
NamedRegionTimer T("combine_lt", "DAG Combining after legalize types",
GroupName, GroupDescription, TimePassesIsEnabled);
CurDAG->Combine(AfterLegalizeTypes, AA, OptLevel);
DEBUG(dbgs() << "Optimized type-legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
NamedRegionTimer T("legalize_vec", "Vector Legalization", GroupName,
GroupDescription, TimePassesIsEnabled);
Changed = CurDAG->LegalizeVectors();
if (Changed) {
DEBUG(dbgs() << "Vector-legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
NamedRegionTimer T("legalize_types2", "Type Legalization 2", GroupName,
GroupDescription, TimePassesIsEnabled);
DEBUG(dbgs() << "Vector/type-legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
if (ViewDAGCombineLT && MatchFilterBB)
CurDAG->viewGraph("dag-combine-lv input for " + BlockName);
// Run the DAG combiner in post-type-legalize mode.
NamedRegionTimer T("combine_lv", "DAG Combining after legalize vectors",
GroupName, GroupDescription, TimePassesIsEnabled);
CurDAG->Combine(AfterLegalizeVectorOps, AA, OptLevel);
DEBUG(dbgs() << "Optimized vector-legalized selection DAG: BB#"
<< BlockNumber << " '" << BlockName << "'\n"; CurDAG->dump());
if (ViewLegalizeDAGs && MatchFilterBB)
CurDAG->viewGraph("legalize input for " + BlockName);
NamedRegionTimer T("legalize", "DAG Legalization", GroupName,
GroupDescription, TimePassesIsEnabled);
DEBUG(dbgs() << "Legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
if (ViewDAGCombine2 && MatchFilterBB)
CurDAG->viewGraph("dag-combine2 input for " + BlockName);
// Run the DAG combiner in post-legalize mode.
NamedRegionTimer T("combine2", "DAG Combining 2", GroupName,
GroupDescription, TimePassesIsEnabled);
CurDAG->Combine(AfterLegalizeDAG, AA, OptLevel);
DEBUG(dbgs() << "Optimized legalized selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
if (OptLevel != CodeGenOpt::None)
if (ViewISelDAGs && MatchFilterBB)
CurDAG->viewGraph("isel input for " + BlockName);
// Third, instruction select all of the operations to machine code, adding the
// code to the MachineBasicBlock.
NamedRegionTimer T("isel", "Instruction Selection", GroupName,
GroupDescription, TimePassesIsEnabled);
DEBUG(dbgs() << "Selected selection DAG: BB#" << BlockNumber
<< " '" << BlockName << "'\n"; CurDAG->dump());
if (ViewSchedDAGs && MatchFilterBB)
CurDAG->viewGraph("scheduler input for " + BlockName);
// Schedule machine code.
ScheduleDAGSDNodes *Scheduler = CreateScheduler();
NamedRegionTimer T("sched", "Instruction Scheduling", GroupName,
GroupDescription, TimePassesIsEnabled);
Scheduler->Run(CurDAG, FuncInfo->MBB);
if (ViewSUnitDAGs && MatchFilterBB)
// Emit machine code to BB. This can change 'BB' to the last block being
// inserted into.
MachineBasicBlock *FirstMBB = FuncInfo->MBB, *LastMBB;
NamedRegionTimer T("emit", "Instruction Creation", GroupName,
GroupDescription, TimePassesIsEnabled);
// FuncInfo->InsertPt is passed by reference and set to the end of the
// scheduled instructions.
LastMBB = FuncInfo->MBB = Scheduler->EmitSchedule(FuncInfo->InsertPt);
// If the block was split, make sure we update any references that are used to
// update PHI nodes later on.
if (FirstMBB != LastMBB)
SDB->UpdateSplitBlock(FirstMBB, LastMBB);
// Free the scheduler state.
NamedRegionTimer T("cleanup", "Instruction Scheduling Cleanup", GroupName,
GroupDescription, TimePassesIsEnabled);
delete Scheduler;
// Free the SelectionDAG state, now that we're finished with it.
namespace {
/// ISelUpdater - helper class to handle updates of the instruction selection
/// graph.
class ISelUpdater : public SelectionDAG::DAGUpdateListener {
SelectionDAG::allnodes_iterator &ISelPosition;
ISelUpdater(SelectionDAG &DAG, SelectionDAG::allnodes_iterator &isp)
: SelectionDAG::DAGUpdateListener(DAG), ISelPosition(isp) {}
/// NodeDeleted - Handle nodes deleted from the graph. If the node being
/// deleted is the current ISelPosition node, update ISelPosition.
void NodeDeleted(SDNode *N, SDNode *E) override {
if (ISelPosition == SelectionDAG::allnodes_iterator(N))
} // end anonymous namespace
void SelectionDAGISel::DoInstructionSelection() {
DEBUG(dbgs() << "===== Instruction selection begins: BB#"
<< FuncInfo->MBB->getNumber()
<< " '" << FuncInfo->MBB->getName() << "'\n");
// Select target instructions for the DAG.
// Number all nodes with a topological order and set DAGSize.
DAGSize = CurDAG->AssignTopologicalOrder();
// Create a dummy node (which is not added to allnodes), that adds
// a reference to the root node, preventing it from being deleted,
// and tracking any changes of the root.
HandleSDNode Dummy(CurDAG->getRoot());
SelectionDAG::allnodes_iterator ISelPosition (CurDAG->getRoot().getNode());
// Make sure that ISelPosition gets properly updated when nodes are deleted
// in calls made from this function.
ISelUpdater ISU(*CurDAG, ISelPosition);
// The AllNodes list is now topological-sorted. Visit the
// nodes by starting at the end of the list (the root of the
// graph) and preceding back toward the beginning (the entry
// node).
while (ISelPosition != CurDAG->allnodes_begin()) {
SDNode *Node = &*--ISelPosition;
// Skip dead nodes. DAGCombiner is expected to eliminate all dead nodes,
// but there are currently some corner cases that it misses. Also, this
// makes it theoretically possible to disable the DAGCombiner.
if (Node->use_empty())
// When we are using non-default rounding modes or FP exception behavior
// FP operations are represented by StrictFP pseudo-operations. They
// need to be simplified here so that the target-specific instruction
// selectors know how to handle them.
// If the current node is a strict FP pseudo-op, the isStrictFPOp()
// function will provide the corresponding normal FP opcode to which the
// node should be mutated.
// FIXME: The backends need a way to handle FP constraints.
if (Node->isStrictFPOpcode())
Node = CurDAG->mutateStrictFPToFP(Node);
DEBUG(dbgs() << "===== Instruction selection ends:\n");
static bool hasExceptionPointerOrCodeUser(const CatchPadInst *CPI) {
for (const User *U : CPI->users()) {
if (const IntrinsicInst *EHPtrCall = dyn_cast<IntrinsicInst>(U)) {
Intrinsic::ID IID = EHPtrCall->getIntrinsicID();
if (IID == Intrinsic::eh_exceptionpointer ||
IID == Intrinsic::eh_exceptioncode)
return true;
return false;
/// PrepareEHLandingPad - Emit an EH_LABEL, set up live-in registers, and
/// do other setup for EH landing-pad blocks.
bool SelectionDAGISel::PrepareEHLandingPad() {
MachineBasicBlock *MBB = FuncInfo->MBB;
const Constant *PersonalityFn = FuncInfo->Fn->getPersonalityFn();
const BasicBlock *LLVMBB = MBB->getBasicBlock();
const TargetRegisterClass *PtrRC =
// Catchpads have one live-in register, which typically holds the exception
// pointer or code.
if (const auto *CPI = dyn_cast<CatchPadInst>(LLVMBB->getFirstNonPHI())) {
if (hasExceptionPointerOrCodeUser(CPI)) {
// Get or create the virtual register to hold the pointer or code. Mark
// the live in physreg and copy into the vreg.
MCPhysReg EHPhysReg = TLI->getExceptionPointerRegister(PersonalityFn);
assert(EHPhysReg && "target lacks exception pointer register");
unsigned VReg = FuncInfo->getCatchPadExceptionPointerVReg(CPI, PtrRC);
BuildMI(*MBB, FuncInfo->InsertPt, SDB->getCurDebugLoc(),
TII->get(TargetOpcode::COPY), VReg)
.addReg(EHPhysReg, RegState::Kill);
return true;
if (!LLVMBB->isLandingPad())
return true;
// Add a label to mark the beginning of the landing pad. Deletion of the
// landing pad can thus be detected via the MachineModuleInfo.
MCSymbol *Label = MF->addLandingPad(MBB);
// Assign the call site to the landing pad's begin label.
MF->setCallSiteLandingPad(Label, SDB->LPadToCallSiteMap[MBB]);
const MCInstrDesc &II = TII->get(TargetOpcode::EH_LABEL);
BuildMI(*MBB, FuncInfo->InsertPt, SDB->getCurDebugLoc(), II)
// Mark exception register as live in.
if (unsigned Reg = TLI->getExceptionPointerRegister(PersonalityFn))
FuncInfo->ExceptionPointerVirtReg = MBB->addLiveIn(Reg, PtrRC);
// Mark exception selector register as live in.
if (unsigned Reg = TLI->getExceptionSelectorRegister(PersonalityFn))
FuncInfo->ExceptionSelectorVirtReg = MBB->addLiveIn(Reg, PtrRC);
return true;
/// isFoldedOrDeadInstruction - Return true if the specified instruction is
/// side-effect free and is either dead or folded into a generated instruction.
/// Return false if it needs to be emitted.
static bool isFoldedOrDeadInstruction(const Instruction *I,
FunctionLoweringInfo *FuncInfo) {
return !I->mayWriteToMemory() && // Side-effecting instructions aren't folded.
!isa<TerminatorInst>(I) && // Terminators aren't folded.
!isa<DbgInfoIntrinsic>(I) && // Debug instructions aren't folded.
!I->isEHPad() && // EH pad instructions aren't folded.
!FuncInfo->isExportedInst(I); // Exported instrs must be computed.
/// Set up SwiftErrorVals by going through the function. If the function has
/// swifterror argument, it will be the first entry.
static void setupSwiftErrorVals(const Function &Fn, const TargetLowering *TLI,
FunctionLoweringInfo *FuncInfo) {
if (!TLI->supportSwiftError())
FuncInfo->SwiftErrorArg = nullptr;
// Check if function has a swifterror argument.
bool HaveSeenSwiftErrorArg = false;
for (Function::const_arg_iterator AI = Fn.arg_begin(), AE = Fn.arg_end();
AI != AE; ++AI)
if (AI->hasSwiftErrorAttr()) {
assert(!HaveSeenSwiftErrorArg &&
"Must have only one swifterror parameter");
(void)HaveSeenSwiftErrorArg; // silence warning.
HaveSeenSwiftErrorArg = true;
FuncInfo->SwiftErrorArg = &*AI;
for (const auto &LLVMBB : Fn)
for (const auto &Inst : LLVMBB) {
if (const AllocaInst *Alloca = dyn_cast<AllocaInst>(&Inst))
if (Alloca->isSwiftError())
static void createSwiftErrorEntriesInEntryBlock(FunctionLoweringInfo *FuncInfo,
FastISel *FastIS,
const TargetLowering *TLI,
const TargetInstrInfo *TII,
SelectionDAGBuilder *SDB) {
if (!TLI->supportSwiftError())
// We only need to do this when we have swifterror parameter or swifterror
// alloc.
if (FuncInfo->SwiftErrorVals.empty())
assert(FuncInfo->MBB == &*FuncInfo->MF->begin() &&
"expected to insert into entry block");
auto &DL = FuncInfo->MF->getDataLayout();
auto const *RC = TLI->getRegClassFor(TLI->getPointerTy(DL));
for (const auto *SwiftErrorVal : FuncInfo->SwiftErrorVals) {
// We will always generate a copy from the argument. It is always used at
// least by the 'return' of the swifterror.
if (FuncInfo->SwiftErrorArg && FuncInfo->SwiftErrorArg == SwiftErrorVal)
unsigned VReg = FuncInfo->MF->getRegInfo().createVirtualRegister(RC);
// Assign Undef to Vreg. We construct MI directly to make sure it works
// with FastISel.
BuildMI(*FuncInfo->MBB, FuncInfo->MBB->getFirstNonPHI(),
SDB->getCurDebugLoc(), TII->get(TargetOpcode::IMPLICIT_DEF),
// Keep FastIS informed about the value we just inserted.
if (FastIS)
FuncInfo->setCurrentSwiftErrorVReg(FuncInfo->MBB, SwiftErrorVal, VReg);
/// Collect llvm.dbg.declare information. This is done after argument lowering
/// in case the declarations refer to arguments.
static void processDbgDeclares(FunctionLoweringInfo *FuncInfo) {
MachineFunction *MF = FuncInfo->MF;
const DataLayout &DL = MF->getDataLayout();
for (const BasicBlock &BB : *FuncInfo->Fn) {
for (const Instruction &I : BB) {
const DbgDeclareInst *DI = dyn_cast<DbgDeclareInst>(&I);
if (!DI)
assert(DI->getVariable() && "Missing variable");
assert(DI->getDebugLoc() && "Missing location");
const Value *Address = DI->getAddress();
if (!Address)
// Look through casts and constant offset GEPs. These mostly come from
// inalloca.
APInt Offset(DL.getPointerSizeInBits(0), 0);
Address = Address->stripAndAccumulateInBoundsConstantOffsets(DL, Offset);
// Check if the variable is a static alloca or a byval or inalloca
// argument passed in memory. If it is not, then we will ignore this
// intrinsic and handle this during isel like dbg.value.
int FI = std::numeric_limits<int>::max();
if (const auto *AI = dyn_cast<AllocaInst>(Address)) {
auto SI = FuncInfo->StaticAllocaMap.find(AI);
if (SI != FuncInfo->StaticAllocaMap.end())
FI = SI->second;
} else if (const auto *Arg = dyn_cast<Argument>(Address))
FI = FuncInfo->getArgumentFrameIndex(Arg);
if (FI == std::numeric_limits<int>::max())
DIExpression *Expr = DI->getExpression();
if (Offset.getBoolValue())
Expr = DIExpression::prepend(Expr, DIExpression::NoDeref,
MF->setVariableDbgInfo(DI->getVariable(), Expr, FI, DI->getDebugLoc());
/// Propagate swifterror values through the machine function CFG.
static void propagateSwiftErrorVRegs(FunctionLoweringInfo *FuncInfo) {
auto *TLI = FuncInfo->TLI;
if (!TLI->supportSwiftError())
// We only need to do this when we have swifterror parameter or swifterror
// alloc.
if (FuncInfo->SwiftErrorVals.empty())
// For each machine basic block in reverse post order.
ReversePostOrderTraversal<MachineFunction *> RPOT(FuncInfo->MF);
for (MachineBasicBlock *MBB : RPOT) {
// For each swifterror value in the function.
for(const auto *SwiftErrorVal : FuncInfo->SwiftErrorVals) {
auto Key = std::make_pair(MBB, SwiftErrorVal);
auto UUseIt = FuncInfo->SwiftErrorVRegUpwardsUse.find(Key);
auto VRegDefIt = FuncInfo->SwiftErrorVRegDefMap.find(Key);
bool UpwardsUse = UUseIt != FuncInfo->SwiftErrorVRegUpwardsUse.end();
unsigned UUseVReg = UpwardsUse ? UUseIt->second : 0;
bool DownwardDef = VRegDefIt != FuncInfo->SwiftErrorVRegDefMap.end();
assert(!(UpwardsUse && !DownwardDef) &&
"We can't have an upwards use but no downwards def");
// If there is no upwards exposed use and an entry for the swifterror in
// the def map for this value we don't need to do anything: We already
// have a downward def for this basic block.
if (!UpwardsUse && DownwardDef)
// Otherwise we either have an upwards exposed use vreg that we need to
// materialize or need to forward the downward def from predecessors.
// Check whether we have a single vreg def from all predecessors.
// Otherwise we need a phi.
SmallVector<std::pair<MachineBasicBlock *, unsigned>, 4> VRegs;
SmallSet<const MachineBasicBlock*, 8> Visited;
for (auto *Pred : MBB->predecessors()) {
if (!Visited.insert(Pred).second)
Pred, FuncInfo->getOrCreateSwiftErrorVReg(Pred, SwiftErrorVal)));
if (Pred != MBB)
// We have a self-edge.
// If there was no upwards use in this basic block there is now one: the
// phi needs to use it self.
if (!UpwardsUse) {
UpwardsUse = true;
UUseIt = FuncInfo->SwiftErrorVRegUpwardsUse.find(Key);
assert(UUseIt != FuncInfo->SwiftErrorVRegUpwardsUse.end());
UUseVReg = UUseIt->second;
// We need a phi node if we have more than one predecessor with different
// downward defs.
bool needPHI =
VRegs.size() >= 1 &&
VRegs.begin(), VRegs.end(),
[&](const std::pair<const MachineBasicBlock *, unsigned> &V)
-> bool { return V.second != VRegs[0].second; }) !=
// If there is no upwards exposed used and we don't need a phi just
// forward the swifterror vreg from the predecessor(s).
if (!UpwardsUse && !needPHI) {
assert(!VRegs.empty() &&
"No predecessors? The entry block should bail out earlier");
// Just forward the swifterror vreg from the predecessor(s).
FuncInfo->setCurrentSwiftErrorVReg(MBB, SwiftErrorVal, VRegs[0].second);
auto DLoc = isa<Instruction>(SwiftErrorVal)
? dyn_cast<Instruction>(SwiftErrorVal)->getDebugLoc()
: DebugLoc();
const auto *TII = FuncInfo->MF->getSubtarget().getInstrInfo();
// If we don't need a phi create a copy to the upward exposed vreg.
if (!needPHI) {
assert(!VRegs.empty() &&
"No predecessors? Is the Calling Convention correct?");
unsigned DestReg = UUseVReg;
BuildMI(*MBB, MBB->getFirstNonPHI(), DLoc, TII->get(TargetOpcode::COPY),
// We need a phi: if there is an upwards exposed use we already have a
// destination virtual register number otherwise we generate a new one.
auto &DL = FuncInfo->MF->getDataLayout();
auto const *RC = TLI->getRegClassFor(TLI->getPointerTy(DL));
unsigned PHIVReg =
UpwardsUse ? UUseVReg
: FuncInfo->MF->getRegInfo().createVirtualRegister(RC);
MachineInstrBuilder SwiftErrorPHI =
BuildMI(*MBB, MBB->getFirstNonPHI(), DLoc,
TII->get(TargetOpcode::PHI), PHIVReg);
for (auto BBRegPair : VRegs) {
// We did not have a definition in this block before: store the phi's vreg
// as this block downward exposed def.
if (!UpwardsUse)
FuncInfo->setCurrentSwiftErrorVReg(MBB, SwiftErrorVal, PHIVReg);
static void preassignSwiftErrorRegs(const TargetLowering *TLI,
FunctionLoweringInfo *FuncInfo,
BasicBlock::const_iterator Begin,
BasicBlock::const_iterator End) {
if (!TLI->supportSwiftError() || FuncInfo->SwiftErrorVals.empty())
// Iterator over instructions and assign vregs to swifterror defs and uses.
for (auto It = Begin; It != End; ++It) {
ImmutableCallSite CS(&*It);
if (CS) {
// A call-site with a swifterror argument is both use and def.
const Value *SwiftErrorAddr = nullptr;
for (auto &Arg : CS.args()) {
if (!Arg->isSwiftError())
// Use of swifterror.
assert(!SwiftErrorAddr && "Cannot have multiple swifterror arguments");
SwiftErrorAddr = &*Arg;
assert(SwiftErrorAddr->isSwiftError() &&
"Must have a swifterror value argument");
unsigned VReg; bool CreatedReg;
std::tie(VReg, CreatedReg) = FuncInfo->getOrCreateSwiftErrorVRegUseAt(
&*It, FuncInfo->MBB, SwiftErrorAddr);
if (!SwiftErrorAddr)
// Def of swifterror.
unsigned VReg; bool CreatedReg;
std::tie(VReg, CreatedReg) =
FuncInfo->setCurrentSwiftErrorVReg(FuncInfo->MBB, SwiftErrorAddr, VReg);
// A load is a use.
} else if (const LoadInst *LI = dyn_cast<const LoadInst>(&*It)) {
const Value *V = LI->getOperand(0);
if (!V->isSwiftError())
unsigned VReg; bool CreatedReg;
std::tie(VReg, CreatedReg) =
FuncInfo->getOrCreateSwiftErrorVRegUseAt(LI, FuncInfo->MBB, V);
// A store is a def.
} else if (const StoreInst *SI = dyn_cast<const StoreInst>(&*It)) {
const Value *SwiftErrorAddr = SI->getOperand(1);
if (!SwiftErrorAddr->isSwiftError())
// Def of swifterror.
unsigned VReg; bool CreatedReg;
std::tie(VReg, CreatedReg) =
FuncInfo->setCurrentSwiftErrorVReg(FuncInfo->MBB, SwiftErrorAddr, VReg);
// A return in a swiferror returning function is a use.
} else if (const ReturnInst *R = dyn_cast<const ReturnInst>(&*It)) {
const Function *F = R->getParent()->getParent();
unsigned VReg; bool CreatedReg;
std::tie(VReg, CreatedReg) = FuncInfo->getOrCreateSwiftErrorVRegUseAt(
R, FuncInfo->MBB, FuncInfo->SwiftErrorArg);
void SelectionDAGISel::SelectAllBasicBlocks(const Function &Fn) {
FastISelFailed = false;
// Initialize the Fast-ISel state, if needed.
FastISel *FastIS = nullptr;
if (TM.Options.EnableFastISel)
FastIS = TLI->createFastISel(*FuncInfo, LibInfo);
setupSwiftErrorVals(Fn, TLI, FuncInfo);
ReversePostOrderTraversal<const Function*> RPOT(&Fn);
// Lower arguments up front. An RPO iteration always visits the entry block
// first.
assert(*RPOT.begin() == &Fn.getEntryBlock());
// Set up FuncInfo for ISel. Entry blocks never have PHIs.
FuncInfo->MBB = FuncInfo->MBBMap[&Fn.getEntryBlock()];
FuncInfo->InsertPt = FuncInfo->MBB->begin();
if (!FastIS) {
} else {
// See if fast isel can lower the arguments.
if (!FastIS->lowerArguments()) {
FastISelFailed = true;
// Fast isel failed to lower these arguments
OptimizationRemarkMissed R("sdagisel", "FastISelFailure",
R << "FastISel didn't lower all arguments: "
<< ore::NV("Prototype", Fn.getType());
reportFastISelFailure(*MF, *ORE, R, EnableFastISelAbort > 1);
// Use SelectionDAG argument lowering
// If we inserted any instructions at the beginning, make a note of
// where they are, so we can be sure to emit subsequent instructions
// after them.
if (FuncInfo->InsertPt != FuncInfo->MBB->begin())
createSwiftErrorEntriesInEntryBlock(FuncInfo, FastIS, TLI, TII, SDB);
// Iterate over all basic blocks in the function.
for (const BasicBlock *LLVMBB : RPOT) {
if (OptLevel != CodeGenOpt::None) {
bool AllPredsVisited = true;
for (const_pred_iterator PI = pred_begin(LLVMBB), PE = pred_end(LLVMBB);
PI != PE; ++PI) {
if (!FuncInfo->VisitedBBs.count(*PI)) {
AllPredsVisited = false;
if (AllPredsVisited) {
for (BasicBlock::const_iterator I = LLVMBB->begin();
const PHINode *PN = dyn_cast<PHINode>(I); ++I)
} else {
for (BasicBlock::const_iterator I = LLVMBB->begin();
const PHINode *PN = dyn_cast<PHINode>(I); ++I)
BasicBlock::const_iterator const Begin =
BasicBlock::const_iterator const End = LLVMBB->end();
BasicBlock::const_iterator BI = End;
FuncInfo->MBB = FuncInfo->MBBMap[LLVMBB];
if (!FuncInfo->MBB)
continue; // Some blocks like catchpads have no code or MBB.
// Insert new instructions after any phi or argument setup code.
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Setup an EH landing-pad block.
FuncInfo->ExceptionPointerVirtReg = 0;
FuncInfo->ExceptionSelectorVirtReg = 0;
if (LLVMBB->isEHPad())
if (!PrepareEHLandingPad())
// Before doing SelectionDAG ISel, see if FastISel has been requested.
if (FastIS) {
if (LLVMBB != &Fn.getEntryBlock())
unsigned NumFastIselRemaining = std::distance(Begin, End);
// Pre-assign swifterror vregs.
preassignSwiftErrorRegs(TLI, FuncInfo, Begin, End);
// Do FastISel on as many instructions as possible.
for (; BI != Begin; --BI) {
const Instruction *Inst = &*std::prev(BI);
// If we no longer require this instruction, skip it.
if (isFoldedOrDeadInstruction(Inst, FuncInfo) ||
ElidedArgCopyInstrs.count(Inst)) {
// Bottom-up: reset the insert pos at the top, after any local-value
// instructions.
// Try to select the instruction with FastISel.
if (FastIS->selectInstruction(Inst)) {
// If fast isel succeeded, skip over all the folded instructions, and
// then see if there is a load right before the selected instructions.
// Try to fold the load if so.
const Instruction *BeforeInst = Inst;
while (BeforeInst != &*Begin) {
BeforeInst = &*std::prev(BasicBlock::const_iterator(BeforeInst));
if (!isFoldedOrDeadInstruction(BeforeInst, FuncInfo))
if (BeforeInst != Inst && isa<LoadInst>(BeforeInst) &&
BeforeInst->hasOneUse() &&
FastIS->tryToFoldLoad(cast<LoadInst>(BeforeInst), Inst)) {
// If we succeeded, don't re-select the load.
BI = std::next(BasicBlock::const_iterator(BeforeInst));
FastISelFailed = true;
// Then handle certain instructions as single-LLVM-Instruction blocks.
// We cannot separate out GCrelocates to their own blocks since we need
// to keep track of gc-relocates for a particular gc-statepoint. This is
// done by SelectionDAGBuilder::LowerAsSTATEPOINT, called before
// visitGCRelocate.
if (isa<CallInst>(Inst) && !isStatepoint(Inst) && !isGCRelocate(Inst)) {
OptimizationRemarkMissed R("sdagisel", "FastISelFailure",
Inst->getDebugLoc(), LLVMBB);
R << "FastISel missed call";
if (R.isEnabled() || EnableFastISelAbort) {
std::string InstStrStorage;
raw_string_ostream InstStr(InstStrStorage);
InstStr << *Inst;
R << ": " << InstStr.str();
reportFastISelFailure(*MF, *ORE, R, EnableFastISelAbort > 2);
if (!Inst->getType()->isVoidTy() && !Inst->getType()->isTokenTy() &&
!Inst->use_empty()) {
unsigned &R = FuncInfo->ValueMap[Inst];
if (!R)
R = FuncInfo->CreateRegs(Inst->getType());
bool HadTailCall = false;
MachineBasicBlock::iterator SavedInsertPt = FuncInfo->InsertPt;
SelectBasicBlock(Inst->getIterator(), BI, HadTailCall);
// If the call was emitted as a tail call, we're done with the block.
// We also need to delete any previously emitted instructions.
if (HadTailCall) {
FastIS->removeDeadCode(SavedInsertPt, FuncInfo->MBB->end());
// Recompute NumFastIselRemaining as Selection DAG instruction
// selection may have handled the call, input args, etc.
unsigned RemainingNow = std::distance(Begin, BI);
NumFastIselFailures += NumFastIselRemaining - RemainingNow;
NumFastIselRemaining = RemainingNow;
OptimizationRemarkMissed R("sdagisel", "FastISelFailure",
Inst->getDebugLoc(), LLVMBB);
bool ShouldAbort = EnableFastISelAbort;
if (isa<TerminatorInst>(Inst)) {
// Use a different message for terminator misses.
R << "FastISel missed terminator";
// Don't abort for terminator unless the level is really high
ShouldAbort = (EnableFastISelAbort > 2);
} else {
R << "FastISel missed";
if (R.isEnabled() || EnableFastISelAbort) {
std::string InstStrStorage;
raw_string_ostream InstStr(InstStrStorage);
InstStr << *Inst;
R << ": " << InstStr.str();
reportFastISelFailure(*MF, *ORE, R, ShouldAbort);
NumFastIselFailures += NumFastIselRemaining;
if (getAnalysis<StackProtector>().shouldEmitSDCheck(*LLVMBB)) {
bool FunctionBasedInstrumentation =
SDB->SPDescriptor.initialize(LLVMBB, FuncInfo->MBBMap[LLVMBB],
if (Begin != BI)
if (Begin != BI) {
// Run SelectionDAG instruction selection on the remainder of the block
// not handled by FastISel. If FastISel is not run, this is the entire
// block.
bool HadTailCall;
SelectBasicBlock(Begin, BI, HadTailCall);
// But if FastISel was run, we already selected some of the block.
// If we emitted a tail-call, we need to delete any previously emitted
// instruction that follows it.
if (HadTailCall && FuncInfo->InsertPt != FuncInfo->MBB->end())
FastIS->removeDeadCode(FuncInfo->InsertPt, FuncInfo->MBB->end());
delete FastIS;
/// Given that the input MI is before a partial terminator sequence TSeq, return
/// true if M + TSeq also a partial terminator sequence.
/// A Terminator sequence is a sequence of MachineInstrs which at this point in
/// lowering copy vregs into physical registers, which are then passed into
/// terminator instructors so we can satisfy ABI constraints. A partial
/// terminator sequence is an improper subset of a terminator sequence (i.e. it
/// may be the whole terminator sequence).
static bool MIIsInTerminatorSequence(const MachineInstr &MI) {
// If we do not have a copy or an implicit def, we return true if and only if
// MI is a debug value.
if (!MI.isCopy() && !MI.isImplicitDef())
// Sometimes DBG_VALUE MI sneak in between the copies from the vregs to the
// physical registers if there is debug info associated with the terminator
// of our mbb. We want to include said debug info in our terminator
// sequence, so we return true in that case.
return MI.isDebugValue();
// We have left the terminator sequence if we are not doing one of the
// following:
// 1. Copying a vreg into a physical register.
// 2. Copying a vreg into a vreg.
// 3. Defining a register via an implicit def.
// OPI should always be a register definition...
MachineInstr::const_mop_iterator OPI = MI.operands_begin();
if (!OPI->isReg() || !OPI->isDef())
return false;
// Defining any register via an implicit def is always ok.
if (MI.isImplicitDef())
return true;
// Grab the copy source...
MachineInstr::const_mop_iterator OPI2 = OPI;
assert(OPI2 != MI.operands_end()
&& "Should have a copy implying we should have 2 arguments.");
// Make sure that the copy dest is not a vreg when the copy source is a
// physical register.
if (!OPI2->isReg() ||
(!TargetRegisterInfo::isPhysicalRegister(OPI->getReg()) &&
return false;
return true;
/// Find the split point at which to splice the end of BB into its success stack
/// protector check machine basic block.
/// On many platforms, due to ABI constraints, terminators, even before register
/// allocation, use physical registers. This creates an issue for us since
/// physical registers at this point can not travel across basic
/// blocks. Luckily, selectiondag always moves physical registers into vregs
/// when they enter functions and moves them through a sequence of copies back
/// into the physical registers right before the terminator creating a
/// ``Terminator Sequence''. This function is searching for the beginning of the
/// terminator sequence so that we can ensure that we splice off not just the
/// terminator, but additionally the copies that move the vregs into the
/// physical registers.
static MachineBasicBlock::iterator
FindSplitPointForStackProtector(MachineBasicBlock *BB) {
MachineBasicBlock::iterator SplitPoint = BB->getFirstTerminator();
if (SplitPoint == BB->begin())
return SplitPoint;
MachineBasicBlock::iterator Start = BB->begin();
MachineBasicBlock::iterator Previous = SplitPoint;
while (MIIsInTerminatorSequence(*Previous)) {
SplitPoint = Previous;
if (Previous == Start)
return SplitPoint;
SelectionDAGISel::FinishBasicBlock() {
DEBUG(dbgs() << "Total amount of phi nodes to update: "
<< FuncInfo->PHINodesToUpdate.size() << "\n";
for (unsigned i = 0, e = FuncInfo->PHINodesToUpdate.size(); i != e; ++i)
dbgs() << "Node " << i << " : ("
<< FuncInfo->PHINodesToUpdate[i].first
<< ", " << FuncInfo->PHINodesToUpdate[i].second << ")\n");
// Next, now that we know what the last MBB the LLVM BB expanded is, update
// PHI nodes in successors.
for (unsigned i = 0, e = FuncInfo->PHINodesToUpdate.size(); i != e; ++i) {
MachineInstrBuilder PHI(*MF, FuncInfo->PHINodesToUpdate[i].first);
assert(PHI->isPHI() &&
"This is not a machine PHI node that we are updating!");
if (!FuncInfo->MBB->isSuccessor(PHI->getParent()))
// Handle stack protector.
if (SDB->SPDescriptor.shouldEmitFunctionBasedCheckStackProtector()) {
// The target provides a guard check function. There is no need to
// generate error handling code or to split current basic block.
MachineBasicBlock *ParentMBB = SDB->SPDescriptor.getParentMBB();
// Add load and check to the basicblock.
FuncInfo->MBB = ParentMBB;
FuncInfo->InsertPt =
SDB->visitSPDescriptorParent(SDB->SPDescriptor, ParentMBB);
// Clear the Per-BB State.
} else if (SDB->SPDescriptor.shouldEmitStackProtector()) {
MachineBasicBlock *ParentMBB = SDB->SPDescriptor.getParentMBB();
MachineBasicBlock *SuccessMBB = SDB->SPDescriptor.getSuccessMBB();
// Find the split point to split the parent mbb. At the same time copy all
// physical registers used in the tail of parent mbb into virtual registers
// before the split point and back into physical registers after the split
// point. This prevents us needing to deal with Live-ins and many other
// register allocation issues caused by us splitting the parent mbb. The
// register allocator will clean up said virtual copies later on.
MachineBasicBlock::iterator SplitPoint =
// Splice the terminator of ParentMBB into SuccessMBB.
SuccessMBB->splice(SuccessMBB->end(), ParentMBB,
// Add compare/jump on neq/jump to the parent BB.
FuncInfo->MBB = ParentMBB;
FuncInfo->InsertPt = ParentMBB->end();
SDB->visitSPDescriptorParent(SDB->SPDescriptor, ParentMBB);
// CodeGen Failure MBB if we have not codegened it yet.
MachineBasicBlock *FailureMBB = SDB->SPDescriptor.getFailureMBB();
if (FailureMBB->empty()) {
FuncInfo->MBB = FailureMBB;
FuncInfo->InsertPt = FailureMBB->end();
// Clear the Per-BB State.
// Lower each BitTestBlock.
for (auto &BTB : SDB->BitTestCases) {
// Lower header first, if it wasn't already lowered
if (!BTB.Emitted) {
// Set the current basic block to the mbb we wish to insert the code into
FuncInfo->MBB = BTB.Parent;
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Emit the code
SDB->visitBitTestHeader(BTB, FuncInfo->MBB);
BranchProbability UnhandledProb = BTB.Prob;
for (unsigned j = 0, ej = BTB.Cases.size(); j != ej; ++j) {
UnhandledProb -= BTB.Cases[j].ExtraProb;
// Set the current basic block to the mbb we wish to insert the code into
FuncInfo->MBB = BTB.Cases[j].ThisBB;
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Emit the code
// If all cases cover a contiguous range, it is not necessary to jump to
// the default block after the last bit test fails. This is because the
// range check during bit test header creation has guaranteed that every
// case here doesn't go outside the range. In this case, there is no need
// to perform the last bit test, as it will always be true. Instead, make
// the second-to-last bit-test fall through to the target of the last bit
// test, and delete the last bit test.
MachineBasicBlock *NextMBB;
if (BTB.ContiguousRange && j + 2 == ej) {
// Second-to-last bit-test with contiguous range: fall through to the
// target of the final bit test.
NextMBB = BTB.Cases[j + 1].TargetBB;
} else if (j + 1 == ej) {
// For the last bit test, fall through to Default.
NextMBB = BTB.Default;
} else {
// Otherwise, fall through to the next bit test.
NextMBB = BTB.Cases[j + 1].ThisBB;
SDB->visitBitTestCase(BTB, NextMBB, UnhandledProb, BTB.Reg, BTB.Cases[j],
if (BTB.ContiguousRange && j + 2 == ej) {
// Since we're not going to use the final bit test, remove it.
// Update PHI Nodes
for (unsigned pi = 0, pe = FuncInfo->PHINodesToUpdate.size();
pi != pe; ++pi) {
MachineInstrBuilder PHI(*MF, FuncInfo->PHINodesToUpdate[pi].first);
MachineBasicBlock *PHIBB = PHI->getParent();
assert(PHI->isPHI() &&
"This is not a machine PHI node that we are updating!");
// This is "default" BB. We have two jumps to it. From "header" BB and
// from last "case" BB, unless the latter was skipped.
if (PHIBB == BTB.Default) {
if (!BTB.ContiguousRange) {
// One of "cases" BB.
for (unsigned j = 0, ej = BTB.Cases.size();
j != ej; ++j) {
MachineBasicBlock* cBB = BTB.Cases[j].ThisBB;
if (cBB->isSuccessor(PHIBB))
// If the JumpTable record is filled in, then we need to emit a jump table.
// Updating the PHI nodes is tricky in this case, since we need to determine
// whether the PHI is a successor of the range check MBB or the jump table MBB
for (unsigned i = 0, e = SDB->JTCases.size(); i != e; ++i) {
// Lower header first, if it wasn't already lowered
if (!SDB->JTCases[i].first.Emitted) {
// Set the current basic block to the mbb we wish to insert the code into
FuncInfo->MBB = SDB->JTCases[i].first.HeaderBB;
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Emit the code
SDB->visitJumpTableHeader(SDB->JTCases[i].second, SDB->JTCases[i].first,
// Set the current basic block to the mbb we wish to insert the code into
FuncInfo->MBB = SDB->JTCases[i].second.MBB;
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Emit the code
// Update PHI Nodes
for (unsigned pi = 0, pe = FuncInfo->PHINodesToUpdate.size();
pi != pe; ++pi) {
MachineInstrBuilder PHI(*MF, FuncInfo->PHINodesToUpdate[pi].first);
MachineBasicBlock *PHIBB = PHI->getParent();
assert(PHI->isPHI() &&
"This is not a machine PHI node that we are updating!");
// "default" BB. We can go there only from header BB.
if (PHIBB == SDB->JTCases[i].second.Default)
// JT BB. Just iterate over successors here
if (FuncInfo->MBB->isSuccessor(PHIBB))
// If we generated any switch lowering information, build and codegen any
// additional DAGs necessary.
for (unsigned i = 0, e = SDB->SwitchCases.size(); i != e; ++i) {
// Set the current basic block to the mbb we wish to insert the code into
FuncInfo->MBB = SDB->SwitchCases[i].ThisBB;
FuncInfo->InsertPt = FuncInfo->MBB->end();
// Determine the unique successors.
SmallVector<MachineBasicBlock *, 2> Succs;
if (SDB->SwitchCases[i].TrueBB != SDB->SwitchCases[i].FalseBB)
// Emit the code. Note that this could result in FuncInfo->MBB being split.
SDB->visitSwitchCase(SDB->SwitchCases[i], FuncInfo->MBB);
// Remember the last block, now that any splitting is done, for use in
// populating PHI nodes in successors.
MachineBasicBlock *ThisBB = FuncInfo->MBB;
// Handle any PHI nodes in successors of this chunk, as if we were coming
// from the original BB before switch expansion. Note that PHI nodes can
// occur multiple times in PHINodesToUpdate. We have to be very careful to
// handle them the right number of times.
for (unsigned i = 0, e = Succs.size(); i != e; ++i) {
FuncInfo->MBB = Succs[i];
FuncInfo->InsertPt = FuncInfo->MBB->end();
// FuncInfo->MBB may have been removed from the CFG if a branch was
// constant folded.
if (ThisBB->isSuccessor(FuncInfo->MBB)) {
for (MachineBasicBlock::iterator
MBBI = FuncInfo->MBB->begin(), MBBE = FuncInfo->MBB->end();
MBBI != MBBE && MBBI->isPHI(); ++MBBI) {
MachineInstrBuilder PHI(*MF, MBBI);
// This value for this PHI node is recorded in PHINodesToUpdate.
for (unsigned pn = 0; ; ++pn) {
assert(pn != FuncInfo->PHINodesToUpdate.size() &&
"Didn't find PHI entry!");
if (FuncInfo->PHINodesToUpdate[pn].first == PHI) {
/// Create the scheduler. If a specific scheduler was specified
/// via the SchedulerRegistry, use it, otherwise select the
/// one preferred by the target.
ScheduleDAGSDNodes *SelectionDAGISel::CreateScheduler() {
return ISHeuristic(this, OptLevel);
// Helper functions used by the generated instruction selector.
// Calls to these methods are generated by tblgen.
/// CheckAndMask - The isel is trying to match something like (and X, 255). If
/// the dag combiner simplified the 255, we still want to match. RHS is the
/// actual value in the DAG on the RHS of an AND, and DesiredMaskS is the value
/// specified in the .td file (e.g. 255).
bool SelectionDAGISel::CheckAndMask(SDValue LHS, ConstantSDNode *RHS,
int64_t DesiredMaskS) const {
const APInt &ActualMask = RHS->getAPIntValue();
const APInt &DesiredMask = APInt(LHS.getValueSizeInBits(), DesiredMaskS);
// If the actual mask exactly matches, success!
if (ActualMask == DesiredMask)
return true;
// If the actual AND mask is allowing unallowed bits, this doesn't match.
if (ActualMask.intersects(~DesiredMask))
return false;
// Otherwise, the DAG Combiner may have proven that the value coming in is
// either already zero or is not demanded. Check for known zero input bits.
APInt NeededMask = DesiredMask & ~ActualMask;
if (CurDAG->MaskedValueIsZero(LHS, NeededMask))
return true;
// TODO: check to see if missing bits are just not demanded.
// Otherwise, this pattern doesn't match.
return false;
/// CheckOrMask - The isel is trying to match something like (or X, 255). If
/// the dag combiner simplified the 255, we still want to match. RHS is the
/// actual value in the DAG on the RHS of an OR, and DesiredMaskS is the value
/// specified in the .td file (e.g. 255).
bool SelectionDAGISel::CheckOrMask(SDValue LHS, ConstantSDNode *RHS,
int64_t DesiredMaskS) const {
const APInt &ActualMask = RHS->getAPIntValue();
const APInt &DesiredMask = APInt(LHS.getValueSizeInBits(), DesiredMaskS);
// If the actual mask exactly matches, success!
if (ActualMask == DesiredMask)
return true;
// If the actual AND mask is allowing unallowed bits, this doesn't match.
if (ActualMask.intersects(~DesiredMask))
return false;
// Otherwise, the DAG Combiner may have proven that the value coming in is
// either already zero or is not demanded. Check for known zero input bits.
APInt NeededMask = DesiredMask & ~ActualMask;
KnownBits Known;
CurDAG->computeKnownBits(LHS, Known);
// If all the missing bits in the or are already known to be set, match!
if (NeededMask.isSubsetOf(Known.One))
return true;
// TODO: check to see if missing bits are just not demanded.
// Otherwise, this pattern doesn't match.
return false;
/// SelectInlineAsmMemoryOperands - Calls to this are automatically generated
/// by tblgen. Others should not call it.
void SelectionDAGISel::SelectInlineAsmMemoryOperands(std::vector<SDValue> &Ops,
const SDLoc &DL) {
std::vector<SDValue> InOps;
std::swap(InOps, Ops);
Ops.push_back(InOps[InlineAsm::Op_InputChain]); // 0
Ops.push_back(InOps[InlineAsm::Op_AsmString]); // 1
Ops.push_back(InOps[InlineAsm::Op_MDNode]); // 2, !srcloc
Ops.push_back(InOps[InlineAsm::Op_ExtraInfo]); // 3 (SideEffect, AlignStack)
unsigned i = InlineAsm::Op_FirstOperand, e = InOps.size();
if (InOps[e-1].getValueType() == MVT::Glue)
--e; // Don't process a glue operand if it is here.
while (i != e) {
unsigned Flags = cast<ConstantSDNode>(InOps[i])->getZExtValue();
if (!InlineAsm::isMemKind(Flags)) {
// Just skip over this operand, copying the operands verbatim.
Ops.insert(Ops.end(), InOps.begin()+i,
InOps.begin()+i+InlineAsm::getNumOperandRegisters(Flags) + 1);
i += InlineAsm::getNumOperandRegisters(Flags) + 1;
} else {
assert(InlineAsm::getNumOperandRegisters(Flags) == 1 &&
"Memory operand with multiple values?");
unsigned TiedToOperand;
if (InlineAsm::isUseOperandTiedToDef(Flags, TiedToOperand)) {
// We need the constraint ID from the operand this is tied to.
unsigned CurOp = InlineAsm::Op_FirstOperand;
Flags = cast<ConstantSDNode>(InOps[CurOp])->getZExtValue();
for (; TiedToOperand; --TiedToOperand) {
CurOp += InlineAsm::getNumOperandRegisters(Flags)+1;
Flags = cast<ConstantSDNode>(InOps[CurOp])->getZExtValue();
// Otherwise, this is a memory operand. Ask the target to select it.
std::vector<SDValue> SelOps;
unsigned ConstraintID = InlineAsm::getMemoryConstraintID(Flags);
if (SelectInlineAsmMemoryOperand(InOps[i+1], ConstraintID, SelOps))
report_fatal_error("Could not match memory address. Inline asm"
" failure!");
// Add this to the output node.
unsigned NewFlags =
InlineAsm::getFlagWord(InlineAsm::Kind_Mem, SelOps.size());
NewFlags = InlineAsm::getFlagWordForMem(NewFlags, ConstraintID);
Ops.push_back(CurDAG->getTargetConstant(NewFlags, DL, MVT::i32));
Ops.insert(Ops.end(), SelOps.begin(), SelOps.end());
i += 2;
// Add the glue input back if present.
if (e != InOps.size())
/// findGlueUse - Return use of MVT::Glue value produced by the specified
/// SDNode.
static SDNode *findGlueUse(SDNode *N) {
unsigned FlagResNo = N->getNumValues()-1;
for (SDNode::use_iterator I = N->use_begin(), E = N->use_end(); I != E; ++I) {
SDUse &Use = I.getUse();
if (Use.getResNo() == FlagResNo)
return Use.getUser();
return nullptr;
/// findNonImmUse - Return true if "Use" is a non-immediate use of "Def".
/// This function iteratively traverses up the operand chain, ignoring
/// certain nodes.
static bool findNonImmUse(SDNode *Use, SDNode* Def, SDNode *ImmedUse,
SDNode *Root, SmallPtrSetImpl<SDNode*> &Visited,
bool IgnoreChains) {
// The NodeID's are given uniques ID's where a node ID is guaranteed to be
// greater than all of its (recursive) operands. If we scan to a point where
// 'use' is smaller than the node we're scanning for, then we know we will
// never find it.
// The Use may be -1 (unassigned) if it is a newly allocated node. This can
// happen because we scan down to newly selected nodes in the case of glue
// uses.
std::vector<SDNode *> WorkList;
while (!WorkList.empty()) {
Use = WorkList.back();
if (Use->getNodeId() < Def->getNodeId() && Use->getNodeId() != -1)
// Don't revisit nodes if we already scanned it and didn't fail, we know we
// won't fail if we scan it again.
if (!Visited.insert(Use).second)
for (const SDValue &Op : Use->op_values()) {
// Ignore chain uses, they are validated by HandleMergeInputChains.
if (Op.getValueType() == MVT::Other && IgnoreChains)
SDNode *N = Op.getNode();
if (N == Def) {
if (Use == ImmedUse || Use == Root)
continue; // We are not looking for immediate use.
assert(N != Root);
return true;
// Traverse up the operand chain.
return false;
/// IsProfitableToFold - Returns true if it's profitable to fold the specific
/// operand node N of U during instruction selection that starts at Root.
bool SelectionDAGISel::IsProfitableToFold(SDValue N, SDNode *U,
SDNode *Root) const {
if (OptLevel == CodeGenOpt::None) return false;
return N.hasOneUse();
/// IsLegalToFold - Returns true if the specific operand node N of
/// U can be folded during instruction selection that starts at Root.
bool SelectionDAGISel::IsLegalToFold(SDValue N, SDNode *U, SDNode *Root,
CodeGenOpt::Level OptLevel,
bool IgnoreChains) {
if (OptLevel == CodeGenOpt::None) return false;
// If Root use can somehow reach N through a path that that doesn't contain
// U then folding N would create a cycle. e.g. In the following
// diagram, Root can reach N through X. If N is folded into into Root, then
// X is both a predecessor and a successor of U.
// [N*] //
// ^ ^ //
// / \ //
// [U*] [X]? //
// ^ ^ //
// \ / //
// \ / //
// [Root*] //
// * indicates nodes to be folded together.
// If Root produces glue, then it gets (even more) interesting. Since it
// will be "glued" together with its glue use in the scheduler, we need to
// check if it might reach N.
// [N*] //
// ^ ^ //
// / \ //
// [U*] [X]? //
// ^ ^ //
// \ \ //
// \ | //
// [Root*] | //
// ^ | //
// f | //
// | / //
// [Y] / //
// ^ / //
// f / //
// | / //
// [GU] //
// If GU (glue use) indirectly reaches N (the load), and Root folds N
// (call it Fold), then X is a predecessor of GU and a successor of
// Fold. But since Fold and GU are glued together, this will create
// a cycle in the scheduling graph.
// If the node has glue, walk down the graph to the "lowest" node in the
// glueged set.
EVT VT = Root->getValueType(Root->getNumValues()-1);
while (VT == MVT::Glue) {
SDNode *GU = findGlueUse(Root);
if (!GU)
Root = GU;
VT = Root->getValueType(Root->getNumValues()-1);
// If our query node has a glue result with a use, we've walked up it. If
// the user (which has already been selected) has a chain or indirectly uses
// the chain, our WalkChainUsers predicate will not consider it. Because of
// this, we cannot ignore chains in this predicate.
IgnoreChains = false;
SmallPtrSet<SDNode*, 16> Visited;
return !findNonImmUse(Root, N.getNode(), U, Root, Visited, IgnoreChains);
void SelectionDAGISel::Select_INLINEASM(SDNode *N) {
SDLoc DL(N);
std::vector<SDValue> Ops(N->op_begin(), N->op_end());
SelectInlineAsmMemoryOperands(Ops, DL);
const EVT VTs[] = {MVT::Other, MVT::Glue};
SDValue New = CurDAG->getNode(ISD::INLINEASM, DL, VTs, Ops);
ReplaceUses(N, New.getNode());
void SelectionDAGISel::Select_READ_REGISTER(SDNode *Op) {
SDLoc dl(Op);
MDNodeSDNode *MD = dyn_cast<MDNodeSDNode>(Op->getOperand(1));
const MDString *RegStr = dyn_cast<MDString>(MD->getMD()->getOperand(0));
unsigned Reg =
TLI->getRegisterByName(RegStr->getString().data(), Op->getValueType(0),
SDValue New = CurDAG->getCopyFromReg(
Op->getOperand(0), dl, Reg, Op->getValueType(0));
ReplaceUses(Op, New.getNode());
void SelectionDAGISel::Select_WRITE_REGISTER(SDNode *Op) {
SDLoc dl(Op);
MDNodeSDNode *MD = dyn_cast<MDNodeSDNode>(Op->getOperand(1));
const MDString *RegStr = dyn_cast<MDString>(MD->getMD()->getOperand(0));
unsigned Reg = TLI->getRegisterByName(RegStr->getString().data(),
SDValue New = CurDAG->getCopyToReg(
Op->getOperand(0), dl, Reg, Op->getOperand(2));
ReplaceUses(Op, New.getNode());
void SelectionDAGISel::Select_UNDEF(SDNode *N) {
CurDAG->SelectNodeTo(N, TargetOpcode::IMPLICIT_DEF, N->getValueType(0));
/// GetVBR - decode a vbr encoding whose top bit is set.
LLVM_ATTRIBUTE_ALWAYS_INLINE static inline uint64_t
GetVBR(uint64_t Val, const unsigned char *MatcherTable, unsigned &Idx) {
assert(Val >= 128 && "Not a VBR");
Val &= 127; // Remove first vbr bit.
unsigned Shift = 7;
uint64_t NextBits;
do {
NextBits = MatcherTable[Idx++];
Val |= (NextBits&127) << Shift;
Shift += 7;
} while (NextBits & 128);
return Val;
/// When a match is complete, this method updates uses of interior chain results
/// to use the new results.
void SelectionDAGISel::UpdateChains(
SDNode *NodeToMatch, SDValue InputChain,
SmallVectorImpl<SDNode *> &ChainNodesMatched, bool isMorphNodeTo) {
SmallVector<SDNode*, 4> NowDeadNodes;
// Now that all the normal results are replaced, we replace the chain and
// glue results if present.
if (!ChainNodesMatched.empty()) {
assert(InputChain.getNode() &&
"Matched input chains but didn't produce a chain");
// Loop over all of the nodes we matched that produced a chain result.
// Replace all the chain results with the final chain we ended up with.
for (unsigned i = 0, e = ChainNodesMatched.size(); i != e; ++i) {
SDNode *ChainNode = ChainNodesMatched[i];
// If ChainNode is null, it's because we replaced it on a previous
// iteration and we cleared it out of the map. Just skip it.
if (!ChainNode)
assert(ChainNode->getOpcode() != ISD::DELETED_NODE &&
"Deleted node left in chain");
// Don't replace the results of the root node if we're doing a
// MorphNodeTo.
if (ChainNode == NodeToMatch && isMorphNodeTo)
SDValue ChainVal = SDValue(ChainNode, ChainNode->getNumValues()-1);
if (ChainVal.getValueType() == MVT::Glue)
ChainVal = ChainVal.getValue(ChainVal->getNumValues()-2);
assert(ChainVal.getValueType() == MVT::Other && "Not a chain?");
SelectionDAG::DAGNodeDeletedListener NDL(
*CurDAG, [&](SDNode *N, SDNode *E) {
std::replace(ChainNodesMatched.begin(), ChainNodesMatched.end(), N,
static_cast<SDNode *>(nullptr));
CurDAG->ReplaceAllUsesOfValueWith(ChainVal, InputChain);
// If the node became dead and we haven't already seen it, delete it.
if (ChainNode != NodeToMatch && ChainNode->use_empty() &&
!std::count(NowDeadNodes.begin(), NowDeadNodes.end(), ChainNode))
if (!NowDeadNodes.empty())
DEBUG(dbgs() << "ISEL: Match complete!\n");
enum ChainResult {
/// WalkChainUsers - Walk down the users of the specified chained node that is
/// part of the pattern we're matching, looking at all of the users we find.
/// This determines whether something is an interior node, whether we have a
/// non-pattern node in between two pattern nodes (which prevent folding because
/// it would induce a cycle) and whether we have a TokenFactor node sandwiched
/// between pattern nodes (in which case the TF becomes part of the pattern).
/// The walk we do here is guaranteed to be small because we quickly get down to
/// already selected nodes "below" us.
static ChainResult
WalkChainUsers(const SDNode *ChainedNode,
SmallVectorImpl<SDNode *> &ChainedNodesInPattern,
DenseMap<const SDNode *, ChainResult> &TokenFactorResult,
SmallVectorImpl<SDNode *> &InteriorChainedNodes) {
ChainResult Result = CR_Simple;
for (SDNode::use_iterator UI = ChainedNode->use_begin(),
E = ChainedNode->use_end(); UI != E; ++UI) {
// Make sure the use is of the chain, not some other value we produce.
if (UI.getUse().getValueType() != MVT::Other) continue;
SDNode *User = *UI;
if (User->getOpcode() == ISD::HANDLENODE) // Root of the graph.
// If we see an already-selected machine node, then we've gone beyond the
// pattern that we're selecting down into the already selected chunk of the
// DAG.
unsigned UserOpcode = User->getOpcode();
if (User->isMachineOpcode() ||
UserOpcode == ISD::CopyToReg ||
UserOpcode == ISD::CopyFromReg ||
UserOpcode == ISD::INLINEASM ||
UserOpcode == ISD::EH_LABEL ||
UserOpcode == ISD::LIFETIME_START ||
UserOpcode == ISD::LIFETIME_END) {
// If their node ID got reset to -1 then they've already been selected.
// Treat them like a MachineOpcode.
if (User->getNodeId() == -1)
// If we have a TokenFactor, we handle it specially.
if (User->getOpcode() != ISD::TokenFactor) {
// If the node isn't a token factor and isn't part of our pattern, then it
// must be a random chained node in between two nodes we're selecting.
// This happens when we have something like:
// x = load ptr
// call
// y = x+4
// store y -> ptr
// Because we structurally match the load/store as a read/modify/write,
// but the call is chained between them. We cannot fold in this case
// because it would induce a cycle in the graph.
if (!std::count(ChainedNodesInPattern.begin(),
ChainedNodesInPattern.end(), User))
return CR_InducesCycle;
// Otherwise we found a node that is part of our pattern. For example in:
// x = load ptr
// y = x+4
// store y -> ptr
// This would happen when we're scanning down from the load and see the
// store as a user. Record that there is a use of ChainedNode that is
// part of the pattern and keep scanning uses.
Result = CR_LeadsToInteriorNode;
// If we found a TokenFactor, there are two cases to consider: first if the
// TokenFactor is just hanging "below" the pattern we're matching (i.e. no
// uses of the TF are in our pattern) we just want to ignore it. Second,
// the TokenFactor can be sandwiched in between two chained nodes, like so:
// [Load chain]
// ^
// |
// [Load]
// ^ ^
// | \ DAG's like cheese
// / \ do you?
// / |
// [TokenFactor] [Op]
// ^ ^
// | |
// \ /
// \ /
// [Store]
// In this case, the TokenFactor becomes part of our match and we rewrite it
// as a new TokenFactor.
// To distinguish these two cases, do a recursive walk down the uses.
auto MemoizeResult = TokenFactorResult.find(User);
bool Visited = MemoizeResult != TokenFactorResult.end();
// Recursively walk chain users only if the result is not memoized.
if (!Visited) {
auto Res = WalkChainUsers(User, ChainedNodesInPattern, TokenFactorResult,
MemoizeResult = TokenFactorResult.insert(std::make_pair(User, Res)).first;
switch (MemoizeResult->second) {
case CR_Simple:
// If the uses of the TokenFactor are just already-selected nodes, ignore
// it, it is "below" our pattern.
case CR_InducesCycle:
// If the uses of the TokenFactor lead to nodes that are not part of our
// pattern that are not selected, folding would turn this into a cycle,
// bail out now.
return CR_InducesCycle;
case CR_LeadsToInteriorNode:
break; // Otherwise, keep processing.
// Okay, we know we're in the interesting interior case. The TokenFactor
// is now going to be considered part of the pattern so that we rewrite its
// uses (it may have uses that are not part of the pattern) with the
// ultimate chain result of the generated code. We will also add its chain
// inputs as inputs to the ultimate TokenFactor we create.
Result = CR_LeadsToInteriorNode;
if (!Visited) {
return Result;
/// HandleMergeInputChains - This implements the OPC_EmitMergeInputChains
/// operation for when the pattern matched at least one node with a chains. The
/// input vector contains a list of all of the chained nodes that we match. We
/// must determine if this is a valid thing to cover (i.e. matching it won't
/// induce cycles in the DAG) and if so, creating a TokenFactor node. that will
/// be used as the input node chain for the generated nodes.
static SDValue
HandleMergeInputChains(SmallVectorImpl<SDNode*> &ChainNodesMatched,
SelectionDAG *CurDAG) {
// Used for memoization. Without it WalkChainUsers could take exponential
// time to run.
DenseMap<const SDNode *, ChainResult> TokenFactorResult;
// Walk all of the chained nodes we've matched, recursively scanning down the
// users of the chain result. This adds any TokenFactor nodes that are caught
// in between chained nodes to the chained and interior nodes list.
SmallVector<SDNode*, 3> InteriorChainedNodes;
for (unsigned i = 0, e = ChainNodesMatched.size(); i != e; ++i) {