llvm-6502/lib/Target/PowerPC/PPCTLSDynamicCall.cpp
Bill Schmidt 49b3971b70 [PowerPC] Fix reverted patch r227976 to avoid register assignment issues
See full discussion in http://reviews.llvm.org/D7491.

We now hide the add-immediate and call instructions together in a
separate pseudo-op, which is tagged to define GPR3 and clobber the
call-killed registers.  The PPCTLSDynamicCall pass prior to RA now
expands this op into the two separate addi and call ops, with explicit
definitions of GPR3 on both instructions, and explicit clobbers on the
call instruction.  The pass is now marked as requiring and preserving
the LiveIntervals and SlotIndexes analyses, and fixes these up after
the replacement sequences are introduced.

Self-hosting has been verified on LE P8 and BE P7 with various
optimization levels, etc.  It has also been verified with the
--no-tls-optimize flag workaround removed.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228725 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-10 19:09:05 +00:00

171 lines
5.5 KiB
C++

//===---------- PPCTLSDynamicCall.cpp - TLS Dynamic Call Fixup ------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This pass expands ADDItls{ld,gd}LADDR[32] machine instructions into
// separate ADDItls[gd]L[32] and GETtlsADDR[32] instructions, both of
// which define GPR3. A copy is added from GPR3 to the target virtual
// register of the original instruction. The GETtlsADDR[32] is really
// a call instruction, so its target register is constrained to be GPR3.
// This is not true of ADDItls[gd]L[32], but there is a legacy linker
// optimization bug that requires the target register of the addi of
// a local- or general-dynamic TLS access sequence to be GPR3.
//
// This is done in a late pass so that TLS variable accesses can be
// fully commoned by MachineCSE.
//
//===----------------------------------------------------------------------===//
#include "PPCInstrInfo.h"
#include "PPC.h"
#include "PPCInstrBuilder.h"
#include "PPCTargetMachine.h"
#include "llvm/CodeGen/LiveIntervalAnalysis.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/raw_ostream.h"
using namespace llvm;
#define DEBUG_TYPE "ppc-tls-dynamic-call"
namespace llvm {
void initializePPCTLSDynamicCallPass(PassRegistry&);
}
namespace {
struct PPCTLSDynamicCall : public MachineFunctionPass {
static char ID;
PPCTLSDynamicCall() : MachineFunctionPass(ID) {
initializePPCTLSDynamicCallPass(*PassRegistry::getPassRegistry());
}
const PPCTargetMachine *TM;
const PPCInstrInfo *TII;
LiveIntervals *LIS;
protected:
bool processBlock(MachineBasicBlock &MBB) {
bool Changed = false;
bool Is64Bit = TM->getSubtargetImpl()->isPPC64();
for (MachineBasicBlock::iterator I = MBB.begin(), IE = MBB.end();
I != IE; ++I) {
MachineInstr *MI = I;
if (MI->getOpcode() != PPC::ADDItlsgdLADDR &&
MI->getOpcode() != PPC::ADDItlsldLADDR &&
MI->getOpcode() != PPC::ADDItlsgdLADDR32 &&
MI->getOpcode() != PPC::ADDItlsldLADDR32)
continue;
DEBUG(dbgs() << "TLS Dynamic Call Fixup:\n " << *MI;);
unsigned OutReg = MI->getOperand(0).getReg();
unsigned InReg = MI->getOperand(1).getReg();
DebugLoc DL = MI->getDebugLoc();
unsigned GPR3 = Is64Bit ? PPC::X3 : PPC::R3;
unsigned Opc1, Opc2;
SmallVector<unsigned, 4> OrigRegs;
OrigRegs.push_back(OutReg);
OrigRegs.push_back(InReg);
OrigRegs.push_back(GPR3);
switch (MI->getOpcode()) {
default:
llvm_unreachable("Opcode inconsistency error");
case PPC::ADDItlsgdLADDR:
Opc1 = PPC::ADDItlsgdL;
Opc2 = PPC::GETtlsADDR;
break;
case PPC::ADDItlsldLADDR:
Opc1 = PPC::ADDItlsldL;
Opc2 = PPC::GETtlsldADDR;
break;
case PPC::ADDItlsgdLADDR32:
Opc1 = PPC::ADDItlsgdL32;
Opc2 = PPC::GETtlsADDR32;
break;
case PPC::ADDItlsldLADDR32:
Opc1 = PPC::ADDItlsldL32;
Opc2 = PPC::GETtlsldADDR32;
break;
}
// Expand into two ops built prior to the existing instruction.
MachineInstr *Addi = BuildMI(MBB, I, DL, TII->get(Opc1), GPR3)
.addReg(InReg);
Addi->addOperand(MI->getOperand(2));
// The ADDItls* instruction is the first instruction in the
// repair range.
MachineBasicBlock::iterator First = I;
--First;
MachineInstr *Call = (BuildMI(MBB, I, DL, TII->get(Opc2), GPR3)
.addReg(GPR3));
Call->addOperand(MI->getOperand(3));
BuildMI(MBB, I, DL, TII->get(TargetOpcode::COPY), OutReg)
.addReg(GPR3);
// The COPY is the last instruction in the repair range.
MachineBasicBlock::iterator Last = I;
--Last;
// Move past the original instruction and remove it.
++I;
MI->removeFromParent();
// Repair the live intervals.
LIS->repairIntervalsInRange(&MBB, First, Last, OrigRegs);
Changed = true;
}
return Changed;
}
public:
bool runOnMachineFunction(MachineFunction &MF) override {
TM = static_cast<const PPCTargetMachine *>(&MF.getTarget());
TII = TM->getSubtargetImpl()->getInstrInfo();
LIS = &getAnalysis<LiveIntervals>();
bool Changed = false;
for (MachineFunction::iterator I = MF.begin(); I != MF.end();) {
MachineBasicBlock &B = *I++;
if (processBlock(B))
Changed = true;
}
return Changed;
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.addRequired<LiveIntervals>();
AU.addPreserved<LiveIntervals>();
AU.addRequired<SlotIndexes>();
AU.addPreserved<SlotIndexes>();
MachineFunctionPass::getAnalysisUsage(AU);
}
};
}
INITIALIZE_PASS_BEGIN(PPCTLSDynamicCall, DEBUG_TYPE,
"PowerPC TLS Dynamic Call Fixup", false, false)
INITIALIZE_PASS_DEPENDENCY(LiveIntervals)
INITIALIZE_PASS_DEPENDENCY(SlotIndexes)
INITIALIZE_PASS_END(PPCTLSDynamicCall, DEBUG_TYPE,
"PowerPC TLS Dynamic Call Fixup", false, false)
char PPCTLSDynamicCall::ID = 0;
FunctionPass*
llvm::createPPCTLSDynamicCallPass() { return new PPCTLSDynamicCall(); }