Page MenuHomePhabricator

No OneTemporary

File Metadata

Created
Jan 24 2020, 4:13 PM
This file is larger than 256 KB, so syntax highlighting was skipped.
<
diff --git a/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp b/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp
index de4377e..2c3ac81 100644
--- a/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp
+++ b/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp
@@ -1,1994 +1,2012 @@
//===-- ARMExpandPseudoInsts.cpp - Expand pseudo instructions -------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file contains a pass that expands pseudo instructions into target
// instructions to allow proper scheduling, if-conversion, and other late
// optimizations. This pass should be run after register allocation but before
// the post-regalloc scheduling pass.
//
//===----------------------------------------------------------------------===//
#include "ARM.h"
#include "ARMBaseInstrInfo.h"
#include "ARMBaseRegisterInfo.h"
#include "ARMConstantPoolValue.h"
#include "ARMMachineFunctionInfo.h"
#include "ARMSubtarget.h"
#include "MCTargetDesc/ARMAddressingModes.h"
#include "llvm/CodeGen/LivePhysRegs.h"
#include "llvm/CodeGen/MachineFrameInfo.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/Support/Debug.h"
using namespace llvm;
#define DEBUG_TYPE "arm-pseudo"
static cl::opt<bool>
VerifyARMPseudo("verify-arm-pseudo-expand", cl::Hidden,
cl::desc("Verify machine code after expanding ARM pseudos"));
#define ARM_EXPAND_PSEUDO_NAME "ARM pseudo instruction expansion pass"
namespace {
class ARMExpandPseudo : public MachineFunctionPass {
public:
static char ID;
ARMExpandPseudo() : MachineFunctionPass(ID) {}
const ARMBaseInstrInfo *TII;
const TargetRegisterInfo *TRI;
const ARMSubtarget *STI;
ARMFunctionInfo *AFI;
bool runOnMachineFunction(MachineFunction &Fn) override;
MachineFunctionProperties getRequiredProperties() const override {
return MachineFunctionProperties().set(
MachineFunctionProperties::Property::NoVRegs);
}
StringRef getPassName() const override {
return ARM_EXPAND_PSEUDO_NAME;
}
private:
void TransferImpOps(MachineInstr &OldMI,
MachineInstrBuilder &UseMI, MachineInstrBuilder &DefMI);
bool ExpandMI(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI,
MachineBasicBlock::iterator &NextMBBI);
bool ExpandMBB(MachineBasicBlock &MBB);
void ExpandVLD(MachineBasicBlock::iterator &MBBI);
void ExpandVST(MachineBasicBlock::iterator &MBBI);
void ExpandLaneOp(MachineBasicBlock::iterator &MBBI);
void ExpandVTBL(MachineBasicBlock::iterator &MBBI,
unsigned Opc, bool IsExt);
void ExpandMOV32BitImm(MachineBasicBlock &MBB,
MachineBasicBlock::iterator &MBBI);
bool ExpandCMP_SWAP(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI, unsigned LdrexOp,
unsigned StrexOp, unsigned UxtOp,
MachineBasicBlock::iterator &NextMBBI);
bool ExpandCMP_SWAP_64(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI,
MachineBasicBlock::iterator &NextMBBI);
};
char ARMExpandPseudo::ID = 0;
}
INITIALIZE_PASS(ARMExpandPseudo, DEBUG_TYPE, ARM_EXPAND_PSEUDO_NAME, false,
false)
/// TransferImpOps - Transfer implicit operands on the pseudo instruction to
/// the instructions created from the expansion.
void ARMExpandPseudo::TransferImpOps(MachineInstr &OldMI,
MachineInstrBuilder &UseMI,
MachineInstrBuilder &DefMI) {
const MCInstrDesc &Desc = OldMI.getDesc();
for (unsigned i = Desc.getNumOperands(), e = OldMI.getNumOperands();
i != e; ++i) {
const MachineOperand &MO = OldMI.getOperand(i);
assert(MO.isReg() && MO.getReg());
if (MO.isUse())
UseMI.add(MO);
else
DefMI.add(MO);
}
}
namespace {
// Constants for register spacing in NEON load/store instructions.
// For quad-register load-lane and store-lane pseudo instructors, the
// spacing is initially assumed to be EvenDblSpc, and that is changed to
// OddDblSpc depending on the lane number operand.
enum NEONRegSpacing {
SingleSpc,
SingleLowSpc , // Single spacing, low registers, three and four vectors.
SingleHighQSpc, // Single spacing, high registers, four vectors.
SingleHighTSpc, // Single spacing, high registers, three vectors.
EvenDblSpc,
OddDblSpc
};
// Entries for NEON load/store information table. The table is sorted by
// PseudoOpc for fast binary-search lookups.
struct NEONLdStTableEntry {
uint16_t PseudoOpc;
uint16_t RealOpc;
bool IsLoad;
bool isUpdating;
bool hasWritebackOperand;
uint8_t RegSpacing; // One of type NEONRegSpacing
uint8_t NumRegs; // D registers loaded or stored
uint8_t RegElts; // elements per D register; used for lane ops
// FIXME: Temporary flag to denote whether the real instruction takes
// a single register (like the encoding) or all of the registers in
// the list (like the asm syntax and the isel DAG). When all definitions
// are converted to take only the single encoded register, this will
// go away.
bool copyAllListRegs;
// Comparison methods for binary search of the table.
bool operator<(const NEONLdStTableEntry &TE) const {
return PseudoOpc < TE.PseudoOpc;
}
friend bool operator<(const NEONLdStTableEntry &TE, unsigned PseudoOpc) {
return TE.PseudoOpc < PseudoOpc;
}
friend bool LLVM_ATTRIBUTE_UNUSED operator<(unsigned PseudoOpc,
const NEONLdStTableEntry &TE) {
return PseudoOpc < TE.PseudoOpc;
}
};
}
static const NEONLdStTableEntry NEONLdStTable[] = {
{ ARM::VLD1LNq16Pseudo, ARM::VLD1LNd16, true, false, false, EvenDblSpc, 1, 4 ,true},
{ ARM::VLD1LNq16Pseudo_UPD, ARM::VLD1LNd16_UPD, true, true, true, EvenDblSpc, 1, 4 ,true},
{ ARM::VLD1LNq32Pseudo, ARM::VLD1LNd32, true, false, false, EvenDblSpc, 1, 2 ,true},
{ ARM::VLD1LNq32Pseudo_UPD, ARM::VLD1LNd32_UPD, true, true, true, EvenDblSpc, 1, 2 ,true},
{ ARM::VLD1LNq8Pseudo, ARM::VLD1LNd8, true, false, false, EvenDblSpc, 1, 8 ,true},
{ ARM::VLD1LNq8Pseudo_UPD, ARM::VLD1LNd8_UPD, true, true, true, EvenDblSpc, 1, 8 ,true},
{ ARM::VLD1d16QPseudo, ARM::VLD1d16Q, true, false, false, SingleSpc, 4, 4 ,false},
{ ARM::VLD1d16TPseudo, ARM::VLD1d16T, true, false, false, SingleSpc, 3, 4 ,false},
{ ARM::VLD1d32QPseudo, ARM::VLD1d32Q, true, false, false, SingleSpc, 4, 2 ,false},
{ ARM::VLD1d32TPseudo, ARM::VLD1d32T, true, false, false, SingleSpc, 3, 2 ,false},
{ ARM::VLD1d64QPseudo, ARM::VLD1d64Q, true, false, false, SingleSpc, 4, 1 ,false},
{ ARM::VLD1d64QPseudoWB_fixed, ARM::VLD1d64Qwb_fixed, true, true, false, SingleSpc, 4, 1 ,false},
{ ARM::VLD1d64QPseudoWB_register, ARM::VLD1d64Qwb_register, true, true, true, SingleSpc, 4, 1 ,false},
{ ARM::VLD1d64TPseudo, ARM::VLD1d64T, true, false, false, SingleSpc, 3, 1 ,false},
{ ARM::VLD1d64TPseudoWB_fixed, ARM::VLD1d64Twb_fixed, true, true, false, SingleSpc, 3, 1 ,false},
{ ARM::VLD1d64TPseudoWB_register, ARM::VLD1d64Twb_register, true, true, true, SingleSpc, 3, 1 ,false},
{ ARM::VLD1d8QPseudo, ARM::VLD1d8Q, true, false, false, SingleSpc, 4, 8 ,false},
{ ARM::VLD1d8TPseudo, ARM::VLD1d8T, true, false, false, SingleSpc, 3, 8 ,false},
{ ARM::VLD1q16HighQPseudo, ARM::VLD1d16Q, true, false, false, SingleHighQSpc, 4, 4 ,false},
{ ARM::VLD1q16HighTPseudo, ARM::VLD1d16T, true, false, false, SingleHighTSpc, 3, 4 ,false},
{ ARM::VLD1q16LowQPseudo_UPD, ARM::VLD1d16Qwb_fixed, true, true, true, SingleLowSpc, 4, 4 ,false},
{ ARM::VLD1q16LowTPseudo_UPD, ARM::VLD1d16Twb_fixed, true, true, true, SingleLowSpc, 3, 4 ,false},
{ ARM::VLD1q32HighQPseudo, ARM::VLD1d32Q, true, false, false, SingleHighQSpc, 4, 2 ,false},
{ ARM::VLD1q32HighTPseudo, ARM::VLD1d32T, true, false, false, SingleHighTSpc, 3, 2 ,false},
{ ARM::VLD1q32LowQPseudo_UPD, ARM::VLD1d32Qwb_fixed, true, true, true, SingleLowSpc, 4, 2 ,false},
{ ARM::VLD1q32LowTPseudo_UPD, ARM::VLD1d32Twb_fixed, true, true, true, SingleLowSpc, 3, 2 ,false},
{ ARM::VLD1q64HighQPseudo, ARM::VLD1d64Q, true, false, false, SingleHighQSpc, 4, 1 ,false},
{ ARM::VLD1q64HighTPseudo, ARM::VLD1d64T, true, false, false, SingleHighTSpc, 3, 1 ,false},
{ ARM::VLD1q64LowQPseudo_UPD, ARM::VLD1d64Qwb_fixed, true, true, true, SingleLowSpc, 4, 1 ,false},
{ ARM::VLD1q64LowTPseudo_UPD, ARM::VLD1d64Twb_fixed, true, true, true, SingleLowSpc, 3, 1 ,false},
{ ARM::VLD1q8HighQPseudo, ARM::VLD1d8Q, true, false, false, SingleHighQSpc, 4, 8 ,false},
{ ARM::VLD1q8HighTPseudo, ARM::VLD1d8T, true, false, false, SingleHighTSpc, 3, 8 ,false},
{ ARM::VLD1q8LowQPseudo_UPD, ARM::VLD1d8Qwb_fixed, true, true, true, SingleLowSpc, 4, 8 ,false},
{ ARM::VLD1q8LowTPseudo_UPD, ARM::VLD1d8Twb_fixed, true, true, true, SingleLowSpc, 3, 8 ,false},
{ ARM::VLD2DUPq16EvenPseudo, ARM::VLD2DUPd16x2, true, false, false, EvenDblSpc, 2, 4 ,false},
{ ARM::VLD2DUPq16OddPseudo, ARM::VLD2DUPd16x2, true, false, false, OddDblSpc, 2, 4 ,false},
{ ARM::VLD2DUPq32EvenPseudo, ARM::VLD2DUPd32x2, true, false, false, EvenDblSpc, 2, 2 ,false},
{ ARM::VLD2DUPq32OddPseudo, ARM::VLD2DUPd32x2, true, false, false, OddDblSpc, 2, 2 ,false},
{ ARM::VLD2DUPq8EvenPseudo, ARM::VLD2DUPd8x2, true, false, false, EvenDblSpc, 2, 8 ,false},
{ ARM::VLD2DUPq8OddPseudo, ARM::VLD2DUPd8x2, true, false, false, OddDblSpc, 2, 8 ,false},
{ ARM::VLD2LNd16Pseudo, ARM::VLD2LNd16, true, false, false, SingleSpc, 2, 4 ,true},
{ ARM::VLD2LNd16Pseudo_UPD, ARM::VLD2LNd16_UPD, true, true, true, SingleSpc, 2, 4 ,true},
{ ARM::VLD2LNd32Pseudo, ARM::VLD2LNd32, true, false, false, SingleSpc, 2, 2 ,true},
{ ARM::VLD2LNd32Pseudo_UPD, ARM::VLD2LNd32_UPD, true, true, true, SingleSpc, 2, 2 ,true},
{ ARM::VLD2LNd8Pseudo, ARM::VLD2LNd8, true, false, false, SingleSpc, 2, 8 ,true},
{ ARM::VLD2LNd8Pseudo_UPD, ARM::VLD2LNd8_UPD, true, true, true, SingleSpc, 2, 8 ,true},
{ ARM::VLD2LNq16Pseudo, ARM::VLD2LNq16, true, false, false, EvenDblSpc, 2, 4 ,true},
{ ARM::VLD2LNq16Pseudo_UPD, ARM::VLD2LNq16_UPD, true, true, true, EvenDblSpc, 2, 4 ,true},
{ ARM::VLD2LNq32Pseudo, ARM::VLD2LNq32, true, false, false, EvenDblSpc, 2, 2 ,true},
{ ARM::VLD2LNq32Pseudo_UPD, ARM::VLD2LNq32_UPD, true, true, true, EvenDblSpc, 2, 2 ,true},
{ ARM::VLD2q16Pseudo, ARM::VLD2q16, true, false, false, SingleSpc, 4, 4 ,false},
{ ARM::VLD2q16PseudoWB_fixed, ARM::VLD2q16wb_fixed, true, true, false, SingleSpc, 4, 4 ,false},
{ ARM::VLD2q16PseudoWB_register, ARM::VLD2q16wb_register, true, true, true, SingleSpc, 4, 4 ,false},
{ ARM::VLD2q32Pseudo, ARM::VLD2q32, true, false, false, SingleSpc, 4, 2 ,false},
{ ARM::VLD2q32PseudoWB_fixed, ARM::VLD2q32wb_fixed, true, true, false, SingleSpc, 4, 2 ,false},
{ ARM::VLD2q32PseudoWB_register, ARM::VLD2q32wb_register, true, true, true, SingleSpc, 4, 2 ,false},
{ ARM::VLD2q8Pseudo, ARM::VLD2q8, true, false, false, SingleSpc, 4, 8 ,false},
{ ARM::VLD2q8PseudoWB_fixed, ARM::VLD2q8wb_fixed, true, true, false, SingleSpc, 4, 8 ,false},
{ ARM::VLD2q8PseudoWB_register, ARM::VLD2q8wb_register, true, true, true, SingleSpc, 4, 8 ,false},
{ ARM::VLD3DUPd16Pseudo, ARM::VLD3DUPd16, true, false, false, SingleSpc, 3, 4,true},
{ ARM::VLD3DUPd16Pseudo_UPD, ARM::VLD3DUPd16_UPD, true, true, true, SingleSpc, 3, 4,true},
{ ARM::VLD3DUPd32Pseudo, ARM::VLD3DUPd32, true, false, false, SingleSpc, 3, 2,true},
{ ARM::VLD3DUPd32Pseudo_UPD, ARM::VLD3DUPd32_UPD, true, true, true, SingleSpc, 3, 2,true},
{ ARM::VLD3DUPd8Pseudo, ARM::VLD3DUPd8, true, false, false, SingleSpc, 3, 8,true},
{ ARM::VLD3DUPd8Pseudo_UPD, ARM::VLD3DUPd8_UPD, true, true, true, SingleSpc, 3, 8,true},
{ ARM::VLD3DUPq16EvenPseudo, ARM::VLD3DUPq16, true, false, false, EvenDblSpc, 3, 4 ,true},
{ ARM::VLD3DUPq16OddPseudo, ARM::VLD3DUPq16, true, false, false, OddDblSpc, 3, 4 ,true},
{ ARM::VLD3DUPq32EvenPseudo, ARM::VLD3DUPq32, true, false, false, EvenDblSpc, 3, 2 ,true},
{ ARM::VLD3DUPq32OddPseudo, ARM::VLD3DUPq32, true, false, false, OddDblSpc, 3, 2 ,true},
{ ARM::VLD3DUPq8EvenPseudo, ARM::VLD3DUPq8, true, false, false, EvenDblSpc, 3, 8 ,true},
{ ARM::VLD3DUPq8OddPseudo, ARM::VLD3DUPq8, true, false, false, OddDblSpc, 3, 8 ,true},
{ ARM::VLD3LNd16Pseudo, ARM::VLD3LNd16, true, false, false, SingleSpc, 3, 4 ,true},
{ ARM::VLD3LNd16Pseudo_UPD, ARM::VLD3LNd16_UPD, true, true, true, SingleSpc, 3, 4 ,true},
{ ARM::VLD3LNd32Pseudo, ARM::VLD3LNd32, true, false, false, SingleSpc, 3, 2 ,true},
{ ARM::VLD3LNd32Pseudo_UPD, ARM::VLD3LNd32_UPD, true, true, true, SingleSpc, 3, 2 ,true},
{ ARM::VLD3LNd8Pseudo, ARM::VLD3LNd8, true, false, false, SingleSpc, 3, 8 ,true},
{ ARM::VLD3LNd8Pseudo_UPD, ARM::VLD3LNd8_UPD, true, true, true, SingleSpc, 3, 8 ,true},
{ ARM::VLD3LNq16Pseudo, ARM::VLD3LNq16, true, false, false, EvenDblSpc, 3, 4 ,true},
{ ARM::VLD3LNq16Pseudo_UPD, ARM::VLD3LNq16_UPD, true, true, true, EvenDblSpc, 3, 4 ,true},
{ ARM::VLD3LNq32Pseudo, ARM::VLD3LNq32, true, false, false, EvenDblSpc, 3, 2 ,true},
{ ARM::VLD3LNq32Pseudo_UPD, ARM::VLD3LNq32_UPD, true, true, true, EvenDblSpc, 3, 2 ,true},
{ ARM::VLD3d16Pseudo, ARM::VLD3d16, true, false, false, SingleSpc, 3, 4 ,true},
{ ARM::VLD3d16Pseudo_UPD, ARM::VLD3d16_UPD, true, true, true, SingleSpc, 3, 4 ,true},
{ ARM::VLD3d32Pseudo, ARM::VLD3d32, true, false, false, SingleSpc, 3, 2 ,true},
{ ARM::VLD3d32Pseudo_UPD, ARM::VLD3d32_UPD, true, true, true, SingleSpc, 3, 2 ,true},
{ ARM::VLD3d8Pseudo, ARM::VLD3d8, true, false, false, SingleSpc, 3, 8 ,true},
{ ARM::VLD3d8Pseudo_UPD, ARM::VLD3d8_UPD, true, true, true, SingleSpc, 3, 8 ,true},
{ ARM::VLD3q16Pseudo_UPD, ARM::VLD3q16_UPD, true, true, true, EvenDblSpc, 3, 4 ,true},
{ ARM::VLD3q16oddPseudo, ARM::VLD3q16, true, false, false, OddDblSpc, 3, 4 ,true},
{ ARM::VLD3q16oddPseudo_UPD, ARM::VLD3q16_UPD, true, true, true, OddDblSpc, 3, 4 ,true},
{ ARM::VLD3q32Pseudo_UPD, ARM::VLD3q32_UPD, true, true, true, EvenDblSpc, 3, 2 ,true},
{ ARM::VLD3q32oddPseudo, ARM::VLD3q32, true, false, false, OddDblSpc, 3, 2 ,true},
{ ARM::VLD3q32oddPseudo_UPD, ARM::VLD3q32_UPD, true, true, true, OddDblSpc, 3, 2 ,true},
{ ARM::VLD3q8Pseudo_UPD, ARM::VLD3q8_UPD, true, true, true, EvenDblSpc, 3, 8 ,true},
{ ARM::VLD3q8oddPseudo, ARM::VLD3q8, true, false, false, OddDblSpc, 3, 8 ,true},
{ ARM::VLD3q8oddPseudo_UPD, ARM::VLD3q8_UPD, true, true, true, OddDblSpc, 3, 8 ,true},
{ ARM::VLD4DUPd16Pseudo, ARM::VLD4DUPd16, true, false, false, SingleSpc, 4, 4,true},
{ ARM::VLD4DUPd16Pseudo_UPD, ARM::VLD4DUPd16_UPD, true, true, true, SingleSpc, 4, 4,true},
{ ARM::VLD4DUPd32Pseudo, ARM::VLD4DUPd32, true, false, false, SingleSpc, 4, 2,true},
{ ARM::VLD4DUPd32Pseudo_UPD, ARM::VLD4DUPd32_UPD, true, true, true, SingleSpc, 4, 2,true},
{ ARM::VLD4DUPd8Pseudo, ARM::VLD4DUPd8, true, false, false, SingleSpc, 4, 8,true},
{ ARM::VLD4DUPd8Pseudo_UPD, ARM::VLD4DUPd8_UPD, true, true, true, SingleSpc, 4, 8,true},
{ ARM::VLD4DUPq16EvenPseudo, ARM::VLD4DUPq16, true, false, false, EvenDblSpc, 4, 4 ,true},
{ ARM::VLD4DUPq16OddPseudo, ARM::VLD4DUPq16, true, false, false, OddDblSpc, 4, 4 ,true},
{ ARM::VLD4DUPq32EvenPseudo, ARM::VLD4DUPq32, true, false, false, EvenDblSpc, 4, 2 ,true},
{ ARM::VLD4DUPq32OddPseudo, ARM::VLD4DUPq32, true, false, false, OddDblSpc, 4, 2 ,true},
{ ARM::VLD4DUPq8EvenPseudo, ARM::VLD4DUPq8, true, false, false, EvenDblSpc, 4, 8 ,true},
{ ARM::VLD4DUPq8OddPseudo, ARM::VLD4DUPq8, true, false, false, OddDblSpc, 4, 8 ,true},
{ ARM::VLD4LNd16Pseudo, ARM::VLD4LNd16, true, false, false, SingleSpc, 4, 4 ,true},
{ ARM::VLD4LNd16Pseudo_UPD, ARM::VLD4LNd16_UPD, true, true, true, SingleSpc, 4, 4 ,true},
{ ARM::VLD4LNd32Pseudo, ARM::VLD4LNd32, true, false, false, SingleSpc, 4, 2 ,true},
{ ARM::VLD4LNd32Pseudo_UPD, ARM::VLD4LNd32_UPD, true, true, true, SingleSpc, 4, 2 ,true},
{ ARM::VLD4LNd8Pseudo, ARM::VLD4LNd8, true, false, false, SingleSpc, 4, 8 ,true},
{ ARM::VLD4LNd8Pseudo_UPD, ARM::VLD4LNd8_UPD, true, true, true, SingleSpc, 4, 8 ,true},
{ ARM::VLD4LNq16Pseudo, ARM::VLD4LNq16, true, false, false, EvenDblSpc, 4, 4 ,true},
{ ARM::VLD4LNq16Pseudo_UPD, ARM::VLD4LNq16_UPD, true, true, true, EvenDblSpc, 4, 4 ,true},
{ ARM::VLD4LNq32Pseudo, ARM::VLD4LNq32, true, false, false, EvenDblSpc, 4, 2 ,true},
{ ARM::VLD4LNq32Pseudo_UPD, ARM::VLD4LNq32_UPD, true, true, true, EvenDblSpc, 4, 2 ,true},
{ ARM::VLD4d16Pseudo, ARM::VLD4d16, true, false, false, SingleSpc, 4, 4 ,true},
{ ARM::VLD4d16Pseudo_UPD, ARM::VLD4d16_UPD, true, true, true, SingleSpc, 4, 4 ,true},
{ ARM::VLD4d32Pseudo, ARM::VLD4d32, true, false, false, SingleSpc, 4, 2 ,true},
{ ARM::VLD4d32Pseudo_UPD, ARM::VLD4d32_UPD, true, true, true, SingleSpc, 4, 2 ,true},
{ ARM::VLD4d8Pseudo, ARM::VLD4d8, true, false, false, SingleSpc, 4, 8 ,true},
{ ARM::VLD4d8Pseudo_UPD, ARM::VLD4d8_UPD, true, true, true, SingleSpc, 4, 8 ,true},
{ ARM::VLD4q16Pseudo_UPD, ARM::VLD4q16_UPD, true, true, true, EvenDblSpc, 4, 4 ,true},
{ ARM::VLD4q16oddPseudo, ARM::VLD4q16, true, false, false, OddDblSpc, 4, 4 ,true},
{ ARM::VLD4q16oddPseudo_UPD, ARM::VLD4q16_UPD, true, true, true, OddDblSpc, 4, 4 ,true},
{ ARM::VLD4q32Pseudo_UPD, ARM::VLD4q32_UPD, true, true, true, EvenDblSpc, 4, 2 ,true},
{ ARM::VLD4q32oddPseudo, ARM::VLD4q32, true, false, false, OddDblSpc, 4, 2 ,true},
{ ARM::VLD4q32oddPseudo_UPD, ARM::VLD4q32_UPD, true, true, true, OddDblSpc, 4, 2 ,true},
{ ARM::VLD4q8Pseudo_UPD, ARM::VLD4q8_UPD, true, true, true, EvenDblSpc, 4, 8 ,true},
{ ARM::VLD4q8oddPseudo, ARM::VLD4q8, true, false, false, OddDblSpc, 4, 8 ,true},
{ ARM::VLD4q8oddPseudo_UPD, ARM::VLD4q8_UPD, true, true, true, OddDblSpc, 4, 8 ,true},
{ ARM::VST1LNq16Pseudo, ARM::VST1LNd16, false, false, false, EvenDblSpc, 1, 4 ,true},
{ ARM::VST1LNq16Pseudo_UPD, ARM::VST1LNd16_UPD, false, true, true, EvenDblSpc, 1, 4 ,true},
{ ARM::VST1LNq32Pseudo, ARM::VST1LNd32, false, false, false, EvenDblSpc, 1, 2 ,true},
{ ARM::VST1LNq32Pseudo_UPD, ARM::VST1LNd32_UPD, false, true, true, EvenDblSpc, 1, 2 ,true},
{ ARM::VST1LNq8Pseudo, ARM::VST1LNd8, false, false, false, EvenDblSpc, 1, 8 ,true},
{ ARM::VST1LNq8Pseudo_UPD, ARM::VST1LNd8_UPD, false, true, true, EvenDblSpc, 1, 8 ,true},
{ ARM::VST1d16QPseudo, ARM::VST1d16Q, false, false, false, SingleSpc, 4, 4 ,false},
{ ARM::VST1d16TPseudo, ARM::VST1d16T, false, false, false, SingleSpc, 3, 4 ,false},
{ ARM::VST1d32QPseudo, ARM::VST1d32Q, false, false, false, SingleSpc, 4, 2 ,false},
{ ARM::VST1d32TPseudo, ARM::VST1d32T, false, false, false, SingleSpc, 3, 2 ,false},
{ ARM::VST1d64QPseudo, ARM::VST1d64Q, false, false, false, SingleSpc, 4, 1 ,false},
{ ARM::VST1d64QPseudoWB_fixed, ARM::VST1d64Qwb_fixed, false, true, false, SingleSpc, 4, 1 ,false},
{ ARM::VST1d64QPseudoWB_register, ARM::VST1d64Qwb_register, false, true, true, SingleSpc, 4, 1 ,false},
{ ARM::VST1d64TPseudo, ARM::VST1d64T, false, false, false, SingleSpc, 3, 1 ,false},
{ ARM::VST1d64TPseudoWB_fixed, ARM::VST1d64Twb_fixed, false, true, false, SingleSpc, 3, 1 ,false},
{ ARM::VST1d64TPseudoWB_register, ARM::VST1d64Twb_register, false, true, true, SingleSpc, 3, 1 ,false},
{ ARM::VST1d8QPseudo, ARM::VST1d8Q, false, false, false, SingleSpc, 4, 8 ,false},
{ ARM::VST1d8TPseudo, ARM::VST1d8T, false, false, false, SingleSpc, 3, 8 ,false},
{ ARM::VST1q16HighQPseudo, ARM::VST1d16Q, false, false, false, SingleHighQSpc, 4, 4 ,false},
{ ARM::VST1q16HighTPseudo, ARM::VST1d16T, false, false, false, SingleHighTSpc, 3, 4 ,false},
{ ARM::VST1q16LowQPseudo_UPD, ARM::VST1d16Qwb_fixed, false, true, true, SingleLowSpc, 4, 4 ,false},
{ ARM::VST1q16LowTPseudo_UPD, ARM::VST1d16Twb_fixed, false, true, true, SingleLowSpc, 3, 4 ,false},
{ ARM::VST1q32HighQPseudo, ARM::VST1d32Q, false, false, false, SingleHighQSpc, 4, 2 ,false},
{ ARM::VST1q32HighTPseudo, ARM::VST1d32T, false, false, false, SingleHighTSpc, 3, 2 ,false},
{ ARM::VST1q32LowQPseudo_UPD, ARM::VST1d32Qwb_fixed, false, true, true, SingleLowSpc, 4, 2 ,false},
{ ARM::VST1q32LowTPseudo_UPD, ARM::VST1d32Twb_fixed, false, true, true, SingleLowSpc, 3, 2 ,false},
{ ARM::VST1q64HighQPseudo, ARM::VST1d64Q, false, false, false, SingleHighQSpc, 4, 1 ,false},
{ ARM::VST1q64HighTPseudo, ARM::VST1d64T, false, false, false, SingleHighTSpc, 3, 1 ,false},
{ ARM::VST1q64LowQPseudo_UPD, ARM::VST1d64Qwb_fixed, false, true, true, SingleLowSpc, 4, 1 ,false},
{ ARM::VST1q64LowTPseudo_UPD, ARM::VST1d64Twb_fixed, false, true, true, SingleLowSpc, 3, 1 ,false},
{ ARM::VST1q8HighQPseudo, ARM::VST1d8Q, false, false, false, SingleHighQSpc, 4, 8 ,false},
{ ARM::VST1q8HighTPseudo, ARM::VST1d8T, false, false, false, SingleHighTSpc, 3, 8 ,false},
{ ARM::VST1q8LowQPseudo_UPD, ARM::VST1d8Qwb_fixed, false, true, true, SingleLowSpc, 4, 8 ,false},
{ ARM::VST1q8LowTPseudo_UPD, ARM::VST1d8Twb_fixed, false, true, true, SingleLowSpc, 3, 8 ,false},
{ ARM::VST2LNd16Pseudo, ARM::VST2LNd16, false, false, false, SingleSpc, 2, 4 ,true},
{ ARM::VST2LNd16Pseudo_UPD, ARM::VST2LNd16_UPD, false, true, true, SingleSpc, 2, 4 ,true},
{ ARM::VST2LNd32Pseudo, ARM::VST2LNd32, false, false, false, SingleSpc, 2, 2 ,true},
{ ARM::VST2LNd32Pseudo_UPD, ARM::VST2LNd32_UPD, false, true, true, SingleSpc, 2, 2 ,true},
{ ARM::VST2LNd8Pseudo, ARM::VST2LNd8, false, false, false, SingleSpc, 2, 8 ,true},
{ ARM::VST2LNd8Pseudo_UPD, ARM::VST2LNd8_UPD, false, true, true, SingleSpc, 2, 8 ,true},
{ ARM::VST2LNq16Pseudo, ARM::VST2LNq16, false, false, false, EvenDblSpc, 2, 4,true},
{ ARM::VST2LNq16Pseudo_UPD, ARM::VST2LNq16_UPD, false, true, true, EvenDblSpc, 2, 4,true},
{ ARM::VST2LNq32Pseudo, ARM::VST2LNq32, false, false, false, EvenDblSpc, 2, 2,true},
{ ARM::VST2LNq32Pseudo_UPD, ARM::VST2LNq32_UPD, false, true, true, EvenDblSpc, 2, 2,true},
{ ARM::VST2q16Pseudo, ARM::VST2q16, false, false, false, SingleSpc, 4, 4 ,false},
{ ARM::VST2q16PseudoWB_fixed, ARM::VST2q16wb_fixed, false, true, false, SingleSpc, 4, 4 ,false},
{ ARM::VST2q16PseudoWB_register, ARM::VST2q16wb_register, false, true, true, SingleSpc, 4, 4 ,false},
{ ARM::VST2q32Pseudo, ARM::VST2q32, false, false, false, SingleSpc, 4, 2 ,false},
{ ARM::VST2q32PseudoWB_fixed, ARM::VST2q32wb_fixed, false, true, false, SingleSpc, 4, 2 ,false},
{ ARM::VST2q32PseudoWB_register, ARM::VST2q32wb_register, false, true, true, SingleSpc, 4, 2 ,false},
{ ARM::VST2q8Pseudo, ARM::VST2q8, false, false, false, SingleSpc, 4, 8 ,false},
{ ARM::VST2q8PseudoWB_fixed, ARM::VST2q8wb_fixed, false, true, false, SingleSpc, 4, 8 ,false},
{ ARM::VST2q8PseudoWB_register, ARM::VST2q8wb_register, false, true, true, SingleSpc, 4, 8 ,false},
{ ARM::VST3LNd16Pseudo, ARM::VST3LNd16, false, false, false, SingleSpc, 3, 4 ,true},
{ ARM::VST3LNd16Pseudo_UPD, ARM::VST3LNd16_UPD, false, true, true, SingleSpc, 3, 4 ,true},
{ ARM::VST3LNd32Pseudo, ARM::VST3LNd32, false, false, false, SingleSpc, 3, 2 ,true},
{ ARM::VST3LNd32Pseudo_UPD, ARM::VST3LNd32_UPD, false, true, true, SingleSpc, 3, 2 ,true},
{ ARM::VST3LNd8Pseudo, ARM::VST3LNd8, false, false, false, SingleSpc, 3, 8 ,true},
{ ARM::VST3LNd8Pseudo_UPD, ARM::VST3LNd8_UPD, false, true, true, SingleSpc, 3, 8 ,true},
{ ARM::VST3LNq16Pseudo, ARM::VST3LNq16, false, false, false, EvenDblSpc, 3, 4,true},
{ ARM::VST3LNq16Pseudo_UPD, ARM::VST3LNq16_UPD, false, true, true, EvenDblSpc, 3, 4,true},
{ ARM::VST3LNq32Pseudo, ARM::VST3LNq32, false, false, false, EvenDblSpc, 3, 2,true},
{ ARM::VST3LNq32Pseudo_UPD, ARM::VST3LNq32_UPD, false, true, true, EvenDblSpc, 3, 2,true},
{ ARM::VST3d16Pseudo, ARM::VST3d16, false, false, false, SingleSpc, 3, 4 ,true},
{ ARM::VST3d16Pseudo_UPD, ARM::VST3d16_UPD, false, true, true, SingleSpc, 3, 4 ,true},
{ ARM::VST3d32Pseudo, ARM::VST3d32, false, false, false, SingleSpc, 3, 2 ,true},
{ ARM::VST3d32Pseudo_UPD, ARM::VST3d32_UPD, false, true, true, SingleSpc, 3, 2 ,true},
{ ARM::VST3d8Pseudo, ARM::VST3d8, false, false, false, SingleSpc, 3, 8 ,true},
{ ARM::VST3d8Pseudo_UPD, ARM::VST3d8_UPD, false, true, true, SingleSpc, 3, 8 ,true},
{ ARM::VST3q16Pseudo_UPD, ARM::VST3q16_UPD, false, true, true, EvenDblSpc, 3, 4 ,true},
{ ARM::VST3q16oddPseudo, ARM::VST3q16, false, false, false, OddDblSpc, 3, 4 ,true},
{ ARM::VST3q16oddPseudo_UPD, ARM::VST3q16_UPD, false, true, true, OddDblSpc, 3, 4 ,true},
{ ARM::VST3q32Pseudo_UPD, ARM::VST3q32_UPD, false, true, true, EvenDblSpc, 3, 2 ,true},
{ ARM::VST3q32oddPseudo, ARM::VST3q32, false, false, false, OddDblSpc, 3, 2 ,true},
{ ARM::VST3q32oddPseudo_UPD, ARM::VST3q32_UPD, false, true, true, OddDblSpc, 3, 2 ,true},
{ ARM::VST3q8Pseudo_UPD, ARM::VST3q8_UPD, false, true, true, EvenDblSpc, 3, 8 ,true},
{ ARM::VST3q8oddPseudo, ARM::VST3q8, false, false, false, OddDblSpc, 3, 8 ,true},
{ ARM::VST3q8oddPseudo_UPD, ARM::VST3q8_UPD, false, true, true, OddDblSpc, 3, 8 ,true},
{ ARM::VST4LNd16Pseudo, ARM::VST4LNd16, false, false, false, SingleSpc, 4, 4 ,true},
{ ARM::VST4LNd16Pseudo_UPD, ARM::VST4LNd16_UPD, false, true, true, SingleSpc, 4, 4 ,true},
{ ARM::VST4LNd32Pseudo, ARM::VST4LNd32, false, false, false, SingleSpc, 4, 2 ,true},
{ ARM::VST4LNd32Pseudo_UPD, ARM::VST4LNd32_UPD, false, true, true, SingleSpc, 4, 2 ,true},
{ ARM::VST4LNd8Pseudo, ARM::VST4LNd8, false, false, false, SingleSpc, 4, 8 ,true},
{ ARM::VST4LNd8Pseudo_UPD, ARM::VST4LNd8_UPD, false, true, true, SingleSpc, 4, 8 ,true},
{ ARM::VST4LNq16Pseudo, ARM::VST4LNq16, false, false, false, EvenDblSpc, 4, 4,true},
{ ARM::VST4LNq16Pseudo_UPD, ARM::VST4LNq16_UPD, false, true, true, EvenDblSpc, 4, 4,true},
{ ARM::VST4LNq32Pseudo, ARM::VST4LNq32, false, false, false, EvenDblSpc, 4, 2,true},
{ ARM::VST4LNq32Pseudo_UPD, ARM::VST4LNq32_UPD, false, true, true, EvenDblSpc, 4, 2,true},
{ ARM::VST4d16Pseudo, ARM::VST4d16, false, false, false, SingleSpc, 4, 4 ,true},
{ ARM::VST4d16Pseudo_UPD, ARM::VST4d16_UPD, false, true, true, SingleSpc, 4, 4 ,true},
{ ARM::VST4d32Pseudo, ARM::VST4d32, false, false, false, SingleSpc, 4, 2 ,true},
{ ARM::VST4d32Pseudo_UPD, ARM::VST4d32_UPD, false, true, true, SingleSpc, 4, 2 ,true},
{ ARM::VST4d8Pseudo, ARM::VST4d8, false, false, false, SingleSpc, 4, 8 ,true},
{ ARM::VST4d8Pseudo_UPD, ARM::VST4d8_UPD, false, true, true, SingleSpc, 4, 8 ,true},
{ ARM::VST4q16Pseudo_UPD, ARM::VST4q16_UPD, false, true, true, EvenDblSpc, 4, 4 ,true},
{ ARM::VST4q16oddPseudo, ARM::VST4q16, false, false, false, OddDblSpc, 4, 4 ,true},
{ ARM::VST4q16oddPseudo_UPD, ARM::VST4q16_UPD, false, true, true, OddDblSpc, 4, 4 ,true},
{ ARM::VST4q32Pseudo_UPD, ARM::VST4q32_UPD, false, true, true, EvenDblSpc, 4, 2 ,true},
{ ARM::VST4q32oddPseudo, ARM::VST4q32, false, false, false, OddDblSpc, 4, 2 ,true},
{ ARM::VST4q32oddPseudo_UPD, ARM::VST4q32_UPD, false, true, true, OddDblSpc, 4, 2 ,true},
{ ARM::VST4q8Pseudo_UPD, ARM::VST4q8_UPD, false, true, true, EvenDblSpc, 4, 8 ,true},
{ ARM::VST4q8oddPseudo, ARM::VST4q8, false, false, false, OddDblSpc, 4, 8 ,true},
{ ARM::VST4q8oddPseudo_UPD, ARM::VST4q8_UPD, false, true, true, OddDblSpc, 4, 8 ,true}
};
/// LookupNEONLdSt - Search the NEONLdStTable for information about a NEON
/// load or store pseudo instruction.
static const NEONLdStTableEntry *LookupNEONLdSt(unsigned Opcode) {
#ifndef NDEBUG
// Make sure the table is sorted.
static std::atomic<bool> TableChecked(false);
if (!TableChecked.load(std::memory_order_relaxed)) {
assert(std::is_sorted(std::begin(NEONLdStTable), std::end(NEONLdStTable)) &&
"NEONLdStTable is not sorted!");
TableChecked.store(true, std::memory_order_relaxed);
}
#endif
auto I = llvm::lower_bound(NEONLdStTable, Opcode);
if (I != std::end(NEONLdStTable) && I->PseudoOpc == Opcode)
return I;
return nullptr;
}
/// GetDSubRegs - Get 4 D subregisters of a Q, QQ, or QQQQ register,
/// corresponding to the specified register spacing. Not all of the results
/// are necessarily valid, e.g., a Q register only has 2 D subregisters.
static void GetDSubRegs(unsigned Reg, NEONRegSpacing RegSpc,
const TargetRegisterInfo *TRI, unsigned &D0,
unsigned &D1, unsigned &D2, unsigned &D3) {
if (RegSpc == SingleSpc || RegSpc == SingleLowSpc) {
D0 = TRI->getSubReg(Reg, ARM::dsub_0);
D1 = TRI->getSubReg(Reg, ARM::dsub_1);
D2 = TRI->getSubReg(Reg, ARM::dsub_2);
D3 = TRI->getSubReg(Reg, ARM::dsub_3);
} else if (RegSpc == SingleHighQSpc) {
D0 = TRI->getSubReg(Reg, ARM::dsub_4);
D1 = TRI->getSubReg(Reg, ARM::dsub_5);
D2 = TRI->getSubReg(Reg, ARM::dsub_6);
D3 = TRI->getSubReg(Reg, ARM::dsub_7);
} else if (RegSpc == SingleHighTSpc) {
D0 = TRI->getSubReg(Reg, ARM::dsub_3);
D1 = TRI->getSubReg(Reg, ARM::dsub_4);
D2 = TRI->getSubReg(Reg, ARM::dsub_5);
D3 = TRI->getSubReg(Reg, ARM::dsub_6);
} else if (RegSpc == EvenDblSpc) {
D0 = TRI->getSubReg(Reg, ARM::dsub_0);
D1 = TRI->getSubReg(Reg, ARM::dsub_2);
D2 = TRI->getSubReg(Reg, ARM::dsub_4);
D3 = TRI->getSubReg(Reg, ARM::dsub_6);
} else {
assert(RegSpc == OddDblSpc && "unknown register spacing");
D0 = TRI->getSubReg(Reg, ARM::dsub_1);
D1 = TRI->getSubReg(Reg, ARM::dsub_3);
D2 = TRI->getSubReg(Reg, ARM::dsub_5);
D3 = TRI->getSubReg(Reg, ARM::dsub_7);
}
}
/// ExpandVLD - Translate VLD pseudo instructions with Q, QQ or QQQQ register
/// operands to real VLD instructions with D register operands.
void ARMExpandPseudo::ExpandVLD(MachineBasicBlock::iterator &MBBI) {
MachineInstr &MI = *MBBI;
MachineBasicBlock &MBB = *MI.getParent();
LLVM_DEBUG(dbgs() << "Expanding: "; MI.dump());
const NEONLdStTableEntry *TableEntry = LookupNEONLdSt(MI.getOpcode());
assert(TableEntry && TableEntry->IsLoad && "NEONLdStTable lookup failed");
NEONRegSpacing RegSpc = (NEONRegSpacing)TableEntry->RegSpacing;
unsigned NumRegs = TableEntry->NumRegs;
MachineInstrBuilder MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(TableEntry->RealOpc));
unsigned OpIdx = 0;
bool DstIsDead = MI.getOperand(OpIdx).isDead();
Register DstReg = MI.getOperand(OpIdx++).getReg();
if(TableEntry->RealOpc == ARM::VLD2DUPd8x2 ||
TableEntry->RealOpc == ARM::VLD2DUPd16x2 ||
TableEntry->RealOpc == ARM::VLD2DUPd32x2) {
unsigned SubRegIndex;
if (RegSpc == EvenDblSpc) {
SubRegIndex = ARM::dsub_0;
} else {
assert(RegSpc == OddDblSpc && "Unexpected spacing!");
SubRegIndex = ARM::dsub_1;
}
Register SubReg = TRI->getSubReg(DstReg, SubRegIndex);
unsigned DstRegPair = TRI->getMatchingSuperReg(SubReg, ARM::dsub_0,
&ARM::DPairSpcRegClass);
MIB.addReg(DstRegPair, RegState::Define | getDeadRegState(DstIsDead));
} else {
unsigned D0, D1, D2, D3;
GetDSubRegs(DstReg, RegSpc, TRI, D0, D1, D2, D3);
MIB.addReg(D0, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 1 && TableEntry->copyAllListRegs)
MIB.addReg(D1, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 2 && TableEntry->copyAllListRegs)
MIB.addReg(D2, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 3 && TableEntry->copyAllListRegs)
MIB.addReg(D3, RegState::Define | getDeadRegState(DstIsDead));
}
if (TableEntry->isUpdating)
MIB.add(MI.getOperand(OpIdx++));
// Copy the addrmode6 operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Copy the am6offset operand.
if (TableEntry->hasWritebackOperand) {
// TODO: The writing-back pseudo instructions we translate here are all
// defined to take am6offset nodes that are capable to represent both fixed
// and register forms. Some real instructions, however, do not rely on
// am6offset and have separate definitions for such forms. When this is the
// case, fixed forms do not take any offset nodes, so here we skip them for
// such instructions. Once all real and pseudo writing-back instructions are
// rewritten without use of am6offset nodes, this code will go away.
const MachineOperand &AM6Offset = MI.getOperand(OpIdx++);
if (TableEntry->RealOpc == ARM::VLD1d8Qwb_fixed ||
TableEntry->RealOpc == ARM::VLD1d16Qwb_fixed ||
TableEntry->RealOpc == ARM::VLD1d32Qwb_fixed ||
TableEntry->RealOpc == ARM::VLD1d64Qwb_fixed ||
TableEntry->RealOpc == ARM::VLD1d8Twb_fixed ||
TableEntry->RealOpc == ARM::VLD1d16Twb_fixed ||
TableEntry->RealOpc == ARM::VLD1d32Twb_fixed ||
TableEntry->RealOpc == ARM::VLD1d64Twb_fixed) {
assert(AM6Offset.getReg() == 0 &&
"A fixed writing-back pseudo instruction provides an offset "
"register!");
} else {
MIB.add(AM6Offset);
}
}
// For an instruction writing double-spaced subregs, the pseudo instruction
// has an extra operand that is a use of the super-register. Record the
// operand index and skip over it.
unsigned SrcOpIdx = 0;
if(TableEntry->RealOpc != ARM::VLD2DUPd8x2 &&
TableEntry->RealOpc != ARM::VLD2DUPd16x2 &&
TableEntry->RealOpc != ARM::VLD2DUPd32x2) {
if (RegSpc == EvenDblSpc || RegSpc == OddDblSpc ||
RegSpc == SingleLowSpc || RegSpc == SingleHighQSpc ||
RegSpc == SingleHighTSpc)
SrcOpIdx = OpIdx++;
}
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Copy the super-register source operand used for double-spaced subregs over
// to the new instruction as an implicit operand.
if (SrcOpIdx != 0) {
MachineOperand MO = MI.getOperand(SrcOpIdx);
MO.setImplicit(true);
MIB.add(MO);
}
// Add an implicit def for the super-register.
MIB.addReg(DstReg, RegState::ImplicitDefine | getDeadRegState(DstIsDead));
TransferImpOps(MI, MIB, MIB);
// Transfer memoperands.
MIB.cloneMemRefs(MI);
MI.eraseFromParent();
LLVM_DEBUG(dbgs() << "To: "; MIB.getInstr()->dump(););
}
/// ExpandVST - Translate VST pseudo instructions with Q, QQ or QQQQ register
/// operands to real VST instructions with D register operands.
void ARMExpandPseudo::ExpandVST(MachineBasicBlock::iterator &MBBI) {
MachineInstr &MI = *MBBI;
MachineBasicBlock &MBB = *MI.getParent();
LLVM_DEBUG(dbgs() << "Expanding: "; MI.dump());
const NEONLdStTableEntry *TableEntry = LookupNEONLdSt(MI.getOpcode());
assert(TableEntry && !TableEntry->IsLoad && "NEONLdStTable lookup failed");
NEONRegSpacing RegSpc = (NEONRegSpacing)TableEntry->RegSpacing;
unsigned NumRegs = TableEntry->NumRegs;
MachineInstrBuilder MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(TableEntry->RealOpc));
unsigned OpIdx = 0;
if (TableEntry->isUpdating)
MIB.add(MI.getOperand(OpIdx++));
// Copy the addrmode6 operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
if (TableEntry->hasWritebackOperand) {
// TODO: The writing-back pseudo instructions we translate here are all
// defined to take am6offset nodes that are capable to represent both fixed
// and register forms. Some real instructions, however, do not rely on
// am6offset and have separate definitions for such forms. When this is the
// case, fixed forms do not take any offset nodes, so here we skip them for
// such instructions. Once all real and pseudo writing-back instructions are
// rewritten without use of am6offset nodes, this code will go away.
const MachineOperand &AM6Offset = MI.getOperand(OpIdx++);
if (TableEntry->RealOpc == ARM::VST1d8Qwb_fixed ||
TableEntry->RealOpc == ARM::VST1d16Qwb_fixed ||
TableEntry->RealOpc == ARM::VST1d32Qwb_fixed ||
TableEntry->RealOpc == ARM::VST1d64Qwb_fixed ||
TableEntry->RealOpc == ARM::VST1d8Twb_fixed ||
TableEntry->RealOpc == ARM::VST1d16Twb_fixed ||
TableEntry->RealOpc == ARM::VST1d32Twb_fixed ||
TableEntry->RealOpc == ARM::VST1d64Twb_fixed) {
assert(AM6Offset.getReg() == 0 &&
"A fixed writing-back pseudo instruction provides an offset "
"register!");
} else {
MIB.add(AM6Offset);
}
}
bool SrcIsKill = MI.getOperand(OpIdx).isKill();
bool SrcIsUndef = MI.getOperand(OpIdx).isUndef();
Register SrcReg = MI.getOperand(OpIdx++).getReg();
unsigned D0, D1, D2, D3;
GetDSubRegs(SrcReg, RegSpc, TRI, D0, D1, D2, D3);
MIB.addReg(D0, getUndefRegState(SrcIsUndef));
if (NumRegs > 1 && TableEntry->copyAllListRegs)
MIB.addReg(D1, getUndefRegState(SrcIsUndef));
if (NumRegs > 2 && TableEntry->copyAllListRegs)
MIB.addReg(D2, getUndefRegState(SrcIsUndef));
if (NumRegs > 3 && TableEntry->copyAllListRegs)
MIB.addReg(D3, getUndefRegState(SrcIsUndef));
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
if (SrcIsKill && !SrcIsUndef) // Add an implicit kill for the super-reg.
MIB->addRegisterKilled(SrcReg, TRI, true);
else if (!SrcIsUndef)
MIB.addReg(SrcReg, RegState::Implicit); // Add implicit uses for src reg.
TransferImpOps(MI, MIB, MIB);
// Transfer memoperands.
MIB.cloneMemRefs(MI);
MI.eraseFromParent();
LLVM_DEBUG(dbgs() << "To: "; MIB.getInstr()->dump(););
}
/// ExpandLaneOp - Translate VLD*LN and VST*LN instructions with Q, QQ or QQQQ
/// register operands to real instructions with D register operands.
void ARMExpandPseudo::ExpandLaneOp(MachineBasicBlock::iterator &MBBI) {
MachineInstr &MI = *MBBI;
MachineBasicBlock &MBB = *MI.getParent();
LLVM_DEBUG(dbgs() << "Expanding: "; MI.dump());
const NEONLdStTableEntry *TableEntry = LookupNEONLdSt(MI.getOpcode());
assert(TableEntry && "NEONLdStTable lookup failed");
NEONRegSpacing RegSpc = (NEONRegSpacing)TableEntry->RegSpacing;
unsigned NumRegs = TableEntry->NumRegs;
unsigned RegElts = TableEntry->RegElts;
MachineInstrBuilder MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(TableEntry->RealOpc));
unsigned OpIdx = 0;
// The lane operand is always the 3rd from last operand, before the 2
// predicate operands.
unsigned Lane = MI.getOperand(MI.getDesc().getNumOperands() - 3).getImm();
// Adjust the lane and spacing as needed for Q registers.
assert(RegSpc != OddDblSpc && "unexpected register spacing for VLD/VST-lane");
if (RegSpc == EvenDblSpc && Lane >= RegElts) {
RegSpc = OddDblSpc;
Lane -= RegElts;
}
assert(Lane < RegElts && "out of range lane for VLD/VST-lane");
unsigned D0 = 0, D1 = 0, D2 = 0, D3 = 0;
unsigned DstReg = 0;
bool DstIsDead = false;
if (TableEntry->IsLoad) {
DstIsDead = MI.getOperand(OpIdx).isDead();
DstReg = MI.getOperand(OpIdx++).getReg();
GetDSubRegs(DstReg, RegSpc, TRI, D0, D1, D2, D3);
MIB.addReg(D0, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 1)
MIB.addReg(D1, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 2)
MIB.addReg(D2, RegState::Define | getDeadRegState(DstIsDead));
if (NumRegs > 3)
MIB.addReg(D3, RegState::Define | getDeadRegState(DstIsDead));
}
if (TableEntry->isUpdating)
MIB.add(MI.getOperand(OpIdx++));
// Copy the addrmode6 operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Copy the am6offset operand.
if (TableEntry->hasWritebackOperand)
MIB.add(MI.getOperand(OpIdx++));
// Grab the super-register source.
MachineOperand MO = MI.getOperand(OpIdx++);
if (!TableEntry->IsLoad)
GetDSubRegs(MO.getReg(), RegSpc, TRI, D0, D1, D2, D3);
// Add the subregs as sources of the new instruction.
unsigned SrcFlags = (getUndefRegState(MO.isUndef()) |
getKillRegState(MO.isKill()));
MIB.addReg(D0, SrcFlags);
if (NumRegs > 1)
MIB.addReg(D1, SrcFlags);
if (NumRegs > 2)
MIB.addReg(D2, SrcFlags);
if (NumRegs > 3)
MIB.addReg(D3, SrcFlags);
// Add the lane number operand.
MIB.addImm(Lane);
OpIdx += 1;
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Copy the super-register source to be an implicit source.
MO.setImplicit(true);
MIB.add(MO);
if (TableEntry->IsLoad)
// Add an implicit def for the super-register.
MIB.addReg(DstReg, RegState::ImplicitDefine | getDeadRegState(DstIsDead));
TransferImpOps(MI, MIB, MIB);
// Transfer memoperands.
MIB.cloneMemRefs(MI);
MI.eraseFromParent();
}
/// ExpandVTBL - Translate VTBL and VTBX pseudo instructions with Q or QQ
/// register operands to real instructions with D register operands.
void ARMExpandPseudo::ExpandVTBL(MachineBasicBlock::iterator &MBBI,
unsigned Opc, bool IsExt) {
MachineInstr &MI = *MBBI;
MachineBasicBlock &MBB = *MI.getParent();
LLVM_DEBUG(dbgs() << "Expanding: "; MI.dump());
MachineInstrBuilder MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc));
unsigned OpIdx = 0;
// Transfer the destination register operand.
MIB.add(MI.getOperand(OpIdx++));
if (IsExt) {
MachineOperand VdSrc(MI.getOperand(OpIdx++));
MIB.add(VdSrc);
}
bool SrcIsKill = MI.getOperand(OpIdx).isKill();
Register SrcReg = MI.getOperand(OpIdx++).getReg();
unsigned D0, D1, D2, D3;
GetDSubRegs(SrcReg, SingleSpc, TRI, D0, D1, D2, D3);
MIB.addReg(D0);
// Copy the other source register operand.
MachineOperand VmSrc(MI.getOperand(OpIdx++));
MIB.add(VmSrc);
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Add an implicit kill and use for the super-reg.
MIB.addReg(SrcReg, RegState::Implicit | getKillRegState(SrcIsKill));
TransferImpOps(MI, MIB, MIB);
MI.eraseFromParent();
LLVM_DEBUG(dbgs() << "To: "; MIB.getInstr()->dump(););
}
static bool IsAnAddressOperand(const MachineOperand &MO) {
// This check is overly conservative. Unless we are certain that the machine
// operand is not a symbol reference, we return that it is a symbol reference.
// This is important as the load pair may not be split up Windows.
switch (MO.getType()) {
case MachineOperand::MO_Register:
case MachineOperand::MO_Immediate:
case MachineOperand::MO_CImmediate:
case MachineOperand::MO_FPImmediate:
case MachineOperand::MO_ShuffleMask:
return false;
case MachineOperand::MO_MachineBasicBlock:
return true;
case MachineOperand::MO_FrameIndex:
return false;
case MachineOperand::MO_ConstantPoolIndex:
case MachineOperand::MO_TargetIndex:
case MachineOperand::MO_JumpTableIndex:
case MachineOperand::MO_ExternalSymbol:
case MachineOperand::MO_GlobalAddress:
case MachineOperand::MO_BlockAddress:
return true;
case MachineOperand::MO_RegisterMask:
case MachineOperand::MO_RegisterLiveOut:
return false;
case MachineOperand::MO_Metadata:
case MachineOperand::MO_MCSymbol:
return true;
case MachineOperand::MO_CFIIndex:
return false;
case MachineOperand::MO_IntrinsicID:
case MachineOperand::MO_Predicate:
llvm_unreachable("should not exist post-isel");
}
llvm_unreachable("unhandled machine operand type");
}
static MachineOperand makeImplicit(const MachineOperand &MO) {
MachineOperand NewMO = MO;
NewMO.setImplicit();
return NewMO;
}
void ARMExpandPseudo::ExpandMOV32BitImm(MachineBasicBlock &MBB,
MachineBasicBlock::iterator &MBBI) {
MachineInstr &MI = *MBBI;
unsigned Opcode = MI.getOpcode();
unsigned PredReg = 0;
ARMCC::CondCodes Pred = getInstrPredicate(MI, PredReg);
Register DstReg = MI.getOperand(0).getReg();
bool DstIsDead = MI.getOperand(0).isDead();
bool isCC = Opcode == ARM::MOVCCi32imm || Opcode == ARM::t2MOVCCi32imm;
const MachineOperand &MO = MI.getOperand(isCC ? 2 : 1);
bool RequiresBundling = STI->isTargetWindows() && IsAnAddressOperand(MO);
MachineInstrBuilder LO16, HI16;
LLVM_DEBUG(dbgs() << "Expanding: "; MI.dump());
if (!STI->hasV6T2Ops() &&
(Opcode == ARM::MOVi32imm || Opcode == ARM::MOVCCi32imm)) {
// FIXME Windows CE supports older ARM CPUs
assert(!STI->isTargetWindows() && "Windows on ARM requires ARMv7+");
// Expand into a movi + orr.
LO16 = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::MOVi), DstReg);
HI16 = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::ORRri))
.addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead))
.addReg(DstReg);
assert (MO.isImm() && "MOVi32imm w/ non-immediate source operand!");
unsigned ImmVal = (unsigned)MO.getImm();
unsigned SOImmValV1 = ARM_AM::getSOImmTwoPartFirst(ImmVal);
unsigned SOImmValV2 = ARM_AM::getSOImmTwoPartSecond(ImmVal);
LO16 = LO16.addImm(SOImmValV1);
HI16 = HI16.addImm(SOImmValV2);
LO16.cloneMemRefs(MI);
HI16.cloneMemRefs(MI);
LO16.addImm(Pred).addReg(PredReg).add(condCodeOp());
HI16.addImm(Pred).addReg(PredReg).add(condCodeOp());
if (isCC)
LO16.add(makeImplicit(MI.getOperand(1)));
TransferImpOps(MI, LO16, HI16);
MI.eraseFromParent();
return;
}
unsigned LO16Opc = 0;
unsigned HI16Opc = 0;
if (Opcode == ARM::t2MOVi32imm || Opcode == ARM::t2MOVCCi32imm) {
LO16Opc = ARM::t2MOVi16;
HI16Opc = ARM::t2MOVTi16;
} else {
LO16Opc = ARM::MOVi16;
HI16Opc = ARM::MOVTi16;
}
LO16 = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(LO16Opc), DstReg);
HI16 = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(HI16Opc))
.addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead))
.addReg(DstReg);
switch (MO.getType()) {
case MachineOperand::MO_Immediate: {
unsigned Imm = MO.getImm();
unsigned Lo16 = Imm & 0xffff;
unsigned Hi16 = (Imm >> 16) & 0xffff;
LO16 = LO16.addImm(Lo16);
HI16 = HI16.addImm(Hi16);
break;
}
case MachineOperand::MO_ExternalSymbol: {
const char *ES = MO.getSymbolName();
unsigned TF = MO.getTargetFlags();
LO16 = LO16.addExternalSymbol(ES, TF | ARMII::MO_LO16);
HI16 = HI16.addExternalSymbol(ES, TF | ARMII::MO_HI16);
break;
}
default: {
const GlobalValue *GV = MO.getGlobal();
unsigned TF = MO.getTargetFlags();
LO16 = LO16.addGlobalAddress(GV, MO.getOffset(), TF | ARMII::MO_LO16);
HI16 = HI16.addGlobalAddress(GV, MO.getOffset(), TF | ARMII::MO_HI16);
break;
}
}
LO16.cloneMemRefs(MI);
HI16.cloneMemRefs(MI);
LO16.addImm(Pred).addReg(PredReg);
HI16.addImm(Pred).addReg(PredReg);
if (RequiresBundling)
finalizeBundle(MBB, LO16->getIterator(), MBBI->getIterator());
if (isCC)
LO16.add(makeImplicit(MI.getOperand(1)));
TransferImpOps(MI, LO16, HI16);
MI.eraseFromParent();
LLVM_DEBUG(dbgs() << "To: "; LO16.getInstr()->dump(););
LLVM_DEBUG(dbgs() << "And: "; HI16.getInstr()->dump(););
}
/// Expand a CMP_SWAP pseudo-inst to an ldrex/strex loop as simply as
/// possible. This only gets used at -O0 so we don't care about efficiency of
/// the generated code.
bool ARMExpandPseudo::ExpandCMP_SWAP(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI,
unsigned LdrexOp, unsigned StrexOp,
unsigned UxtOp,
MachineBasicBlock::iterator &NextMBBI) {
bool IsThumb = STI->isThumb();
MachineInstr &MI = *MBBI;
DebugLoc DL = MI.getDebugLoc();
const MachineOperand &Dest = MI.getOperand(0);
Register TempReg = MI.getOperand(1).getReg();
// Duplicating undef operands into 2 instructions does not guarantee the same
// value on both; However undef should be replaced by xzr anyway.
assert(!MI.getOperand(2).isUndef() && "cannot handle undef");
Register AddrReg = MI.getOperand(2).getReg();
Register DesiredReg = MI.getOperand(3).getReg();
Register NewReg = MI.getOperand(4).getReg();
MachineFunction *MF = MBB.getParent();
auto LoadCmpBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
auto StoreBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
auto DoneBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
MF->insert(++MBB.getIterator(), LoadCmpBB);
MF->insert(++LoadCmpBB->getIterator(), StoreBB);
MF->insert(++StoreBB->getIterator(), DoneBB);
if (UxtOp) {
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, DL, TII->get(UxtOp), DesiredReg)
.addReg(DesiredReg, RegState::Kill);
if (!IsThumb)
MIB.addImm(0);
MIB.add(predOps(ARMCC::AL));
}
// .Lloadcmp:
// ldrex rDest, [rAddr]
// cmp rDest, rDesired
// bne .Ldone
MachineInstrBuilder MIB;
MIB = BuildMI(LoadCmpBB, DL, TII->get(LdrexOp), Dest.getReg());
MIB.addReg(AddrReg);
if (LdrexOp == ARM::t2LDREX)
MIB.addImm(0); // a 32-bit Thumb ldrex (only) allows an offset.
MIB.add(predOps(ARMCC::AL));
unsigned CMPrr = IsThumb ? ARM::tCMPhir : ARM::CMPrr;
BuildMI(LoadCmpBB, DL, TII->get(CMPrr))
.addReg(Dest.getReg(), getKillRegState(Dest.isDead()))
.addReg(DesiredReg)
.add(predOps(ARMCC::AL));
unsigned Bcc = IsThumb ? ARM::tBcc : ARM::Bcc;
BuildMI(LoadCmpBB, DL, TII->get(Bcc))
.addMBB(DoneBB)
.addImm(ARMCC::NE)
.addReg(ARM::CPSR, RegState::Kill);
LoadCmpBB->addSuccessor(DoneBB);
LoadCmpBB->addSuccessor(StoreBB);
// .Lstore:
// strex rTempReg, rNew, [rAddr]
// cmp rTempReg, #0
// bne .Lloadcmp
MIB = BuildMI(StoreBB, DL, TII->get(StrexOp), TempReg)
.addReg(NewReg)
.addReg(AddrReg);
if (StrexOp == ARM::t2STREX)
MIB.addImm(0); // a 32-bit Thumb strex (only) allows an offset.
MIB.add(predOps(ARMCC::AL));
unsigned CMPri = IsThumb ? ARM::t2CMPri : ARM::CMPri;
BuildMI(StoreBB, DL, TII->get(CMPri))
.addReg(TempReg, RegState::Kill)
.addImm(0)
.add(predOps(ARMCC::AL));
BuildMI(StoreBB, DL, TII->get(Bcc))
.addMBB(LoadCmpBB)
.addImm(ARMCC::NE)
.addReg(ARM::CPSR, RegState::Kill);
StoreBB->addSuccessor(LoadCmpBB);
StoreBB->addSuccessor(DoneBB);
DoneBB->splice(DoneBB->end(), &MBB, MI, MBB.end());
DoneBB->transferSuccessors(&MBB);
MBB.addSuccessor(LoadCmpBB);
NextMBBI = MBB.end();
MI.eraseFromParent();
// Recompute livein lists.
LivePhysRegs LiveRegs;
computeAndAddLiveIns(LiveRegs, *DoneBB);
computeAndAddLiveIns(LiveRegs, *StoreBB);
computeAndAddLiveIns(LiveRegs, *LoadCmpBB);
// Do an extra pass around the loop to get loop carried registers right.
StoreBB->clearLiveIns();
computeAndAddLiveIns(LiveRegs, *StoreBB);
LoadCmpBB->clearLiveIns();
computeAndAddLiveIns(LiveRegs, *LoadCmpBB);
return true;
}
/// ARM's ldrexd/strexd take a consecutive register pair (represented as a
/// single GPRPair register), Thumb's take two separate registers so we need to
/// extract the subregs from the pair.
static void addExclusiveRegPair(MachineInstrBuilder &MIB, MachineOperand &Reg,
unsigned Flags, bool IsThumb,
const TargetRegisterInfo *TRI) {
if (IsThumb) {
Register RegLo = TRI->getSubReg(Reg.getReg(), ARM::gsub_0);
Register RegHi = TRI->getSubReg(Reg.getReg(), ARM::gsub_1);
MIB.addReg(RegLo, Flags);
MIB.addReg(RegHi, Flags);
} else
MIB.addReg(Reg.getReg(), Flags);
}
/// Expand a 64-bit CMP_SWAP to an ldrexd/strexd loop.
bool ARMExpandPseudo::ExpandCMP_SWAP_64(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI,
MachineBasicBlock::iterator &NextMBBI) {
bool IsThumb = STI->isThumb();
MachineInstr &MI = *MBBI;
DebugLoc DL = MI.getDebugLoc();
MachineOperand &Dest = MI.getOperand(0);
Register TempReg = MI.getOperand(1).getReg();
// Duplicating undef operands into 2 instructions does not guarantee the same
// value on both; However undef should be replaced by xzr anyway.
assert(!MI.getOperand(2).isUndef() && "cannot handle undef");
Register AddrReg = MI.getOperand(2).getReg();
Register DesiredReg = MI.getOperand(3).getReg();
MachineOperand New = MI.getOperand(4);
New.setIsKill(false);
Register DestLo = TRI->getSubReg(Dest.getReg(), ARM::gsub_0);
Register DestHi = TRI->getSubReg(Dest.getReg(), ARM::gsub_1);
Register DesiredLo = TRI->getSubReg(DesiredReg, ARM::gsub_0);
Register DesiredHi = TRI->getSubReg(DesiredReg, ARM::gsub_1);
MachineFunction *MF = MBB.getParent();
auto LoadCmpBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
auto StoreBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
auto DoneBB = MF->CreateMachineBasicBlock(MBB.getBasicBlock());
MF->insert(++MBB.getIterator(), LoadCmpBB);
MF->insert(++LoadCmpBB->getIterator(), StoreBB);
MF->insert(++StoreBB->getIterator(), DoneBB);
// .Lloadcmp:
// ldrexd rDestLo, rDestHi, [rAddr]
// cmp rDestLo, rDesiredLo
// sbcs dead rTempReg, rDestHi, rDesiredHi
// bne .Ldone
unsigned LDREXD = IsThumb ? ARM::t2LDREXD : ARM::LDREXD;
MachineInstrBuilder MIB;
MIB = BuildMI(LoadCmpBB, DL, TII->get(LDREXD));
addExclusiveRegPair(MIB, Dest, RegState::Define, IsThumb, TRI);
MIB.addReg(AddrReg).add(predOps(ARMCC::AL));
unsigned CMPrr = IsThumb ? ARM::tCMPhir : ARM::CMPrr;
BuildMI(LoadCmpBB, DL, TII->get(CMPrr))
.addReg(DestLo, getKillRegState(Dest.isDead()))
.addReg(DesiredLo)
.add(predOps(ARMCC::AL));
BuildMI(LoadCmpBB, DL, TII->get(CMPrr))
.addReg(DestHi, getKillRegState(Dest.isDead()))
.addReg(DesiredHi)
.addImm(ARMCC::EQ).addReg(ARM::CPSR, RegState::Kill);
unsigned Bcc = IsThumb ? ARM::tBcc : ARM::Bcc;
BuildMI(LoadCmpBB, DL, TII->get(Bcc))
.addMBB(DoneBB)
.addImm(ARMCC::NE)
.addReg(ARM::CPSR, RegState::Kill);
LoadCmpBB->addSuccessor(DoneBB);
LoadCmpBB->addSuccessor(StoreBB);
// .Lstore:
// strexd rTempReg, rNewLo, rNewHi, [rAddr]
// cmp rTempReg, #0
// bne .Lloadcmp
unsigned STREXD = IsThumb ? ARM::t2STREXD : ARM::STREXD;
MIB = BuildMI(StoreBB, DL, TII->get(STREXD), TempReg);
unsigned Flags = getKillRegState(New.isDead());
addExclusiveRegPair(MIB, New, Flags, IsThumb, TRI);
MIB.addReg(AddrReg).add(predOps(ARMCC::AL));
unsigned CMPri = IsThumb ? ARM::t2CMPri : ARM::CMPri;
BuildMI(StoreBB, DL, TII->get(CMPri))
.addReg(TempReg, RegState::Kill)
.addImm(0)
.add(predOps(ARMCC::AL));
BuildMI(StoreBB, DL, TII->get(Bcc))
.addMBB(LoadCmpBB)
.addImm(ARMCC::NE)
.addReg(ARM::CPSR, RegState::Kill);
StoreBB->addSuccessor(LoadCmpBB);
StoreBB->addSuccessor(DoneBB);
DoneBB->splice(DoneBB->end(), &MBB, MI, MBB.end());
DoneBB->transferSuccessors(&MBB);
MBB.addSuccessor(LoadCmpBB);
NextMBBI = MBB.end();
MI.eraseFromParent();
// Recompute livein lists.
LivePhysRegs LiveRegs;
computeAndAddLiveIns(LiveRegs, *DoneBB);
computeAndAddLiveIns(LiveRegs, *StoreBB);
computeAndAddLiveIns(LiveRegs, *LoadCmpBB);
// Do an extra pass around the loop to get loop carried registers right.
StoreBB->clearLiveIns();
computeAndAddLiveIns(LiveRegs, *StoreBB);
LoadCmpBB->clearLiveIns();
computeAndAddLiveIns(LiveRegs, *LoadCmpBB);
return true;
}
bool ARMExpandPseudo::ExpandMI(MachineBasicBlock &MBB,
MachineBasicBlock::iterator MBBI,
MachineBasicBlock::iterator &NextMBBI) {
MachineInstr &MI = *MBBI;
unsigned Opcode = MI.getOpcode();
switch (Opcode) {
default:
return false;
case ARM::TCRETURNdi:
case ARM::TCRETURNri: {
MachineBasicBlock::iterator MBBI = MBB.getLastNonDebugInstr();
assert(MBBI->isReturn() &&
"Can only insert epilog into returning blocks");
unsigned RetOpcode = MBBI->getOpcode();
DebugLoc dl = MBBI->getDebugLoc();
const ARMBaseInstrInfo &TII = *static_cast<const ARMBaseInstrInfo *>(
MBB.getParent()->getSubtarget().getInstrInfo());
// Tail call return: adjust the stack pointer and jump to callee.
MBBI = MBB.getLastNonDebugInstr();
MachineOperand &JumpTarget = MBBI->getOperand(0);
// Jump to label or value in register.
if (RetOpcode == ARM::TCRETURNdi) {
unsigned TCOpcode =
STI->isThumb()
? (STI->isTargetMachO() ? ARM::tTAILJMPd : ARM::tTAILJMPdND)
: ARM::TAILJMPd;
MachineInstrBuilder MIB = BuildMI(MBB, MBBI, dl, TII.get(TCOpcode));
if (JumpTarget.isGlobal())
MIB.addGlobalAddress(JumpTarget.getGlobal(), JumpTarget.getOffset(),
JumpTarget.getTargetFlags());
else {
assert(JumpTarget.isSymbol());
MIB.addExternalSymbol(JumpTarget.getSymbolName(),
JumpTarget.getTargetFlags());
}
// Add the default predicate in Thumb mode.
if (STI->isThumb())
MIB.add(predOps(ARMCC::AL));
} else if (RetOpcode == ARM::TCRETURNri) {
unsigned Opcode =
STI->isThumb() ? ARM::tTAILJMPr
: (STI->hasV4TOps() ? ARM::TAILJMPr : ARM::TAILJMPr4);
BuildMI(MBB, MBBI, dl,
TII.get(Opcode))
.addReg(JumpTarget.getReg(), RegState::Kill);
}
auto NewMI = std::prev(MBBI);
for (unsigned i = 1, e = MBBI->getNumOperands(); i != e; ++i)
NewMI->addOperand(MBBI->getOperand(i));
// Update call site info and delete the pseudo instruction TCRETURN.
MBB.getParent()->moveCallSiteInfo(&MI, &*NewMI);
MBB.erase(MBBI);
MBBI = NewMI;
return true;
}
case ARM::VMOVHcc:
case ARM::VMOVScc:
case ARM::VMOVDcc: {
unsigned newOpc = Opcode != ARM::VMOVDcc ? ARM::VMOVS : ARM::VMOVD;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(newOpc),
MI.getOperand(1).getReg())
.add(MI.getOperand(2))
.addImm(MI.getOperand(3).getImm()) // 'pred'
.add(MI.getOperand(4))
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::t2MOVCCr:
case ARM::MOVCCr: {
unsigned Opc = AFI->isThumbFunction() ? ARM::t2MOVr : ARM::MOVr;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc),
MI.getOperand(1).getReg())
.add(MI.getOperand(2))
.addImm(MI.getOperand(3).getImm()) // 'pred'
.add(MI.getOperand(4))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::MOVCCsi: {
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::MOVsi),
(MI.getOperand(1).getReg()))
.add(MI.getOperand(2))
.addImm(MI.getOperand(3).getImm())
.addImm(MI.getOperand(4).getImm()) // 'pred'
.add(MI.getOperand(5))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::MOVCCsr: {
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::MOVsr),
(MI.getOperand(1).getReg()))
.add(MI.getOperand(2))
.add(MI.getOperand(3))
.addImm(MI.getOperand(4).getImm())
.addImm(MI.getOperand(5).getImm()) // 'pred'
.add(MI.getOperand(6))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::t2MOVCCi16:
case ARM::MOVCCi16: {
unsigned NewOpc = AFI->isThumbFunction() ? ARM::t2MOVi16 : ARM::MOVi16;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(NewOpc),
MI.getOperand(1).getReg())
.addImm(MI.getOperand(2).getImm())
.addImm(MI.getOperand(3).getImm()) // 'pred'
.add(MI.getOperand(4))
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::t2MOVCCi:
case ARM::MOVCCi: {
unsigned Opc = AFI->isThumbFunction() ? ARM::t2MOVi : ARM::MOVi;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc),
MI.getOperand(1).getReg())
.addImm(MI.getOperand(2).getImm())
.addImm(MI.getOperand(3).getImm()) // 'pred'
.add(MI.getOperand(4))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::t2MVNCCi:
case ARM::MVNCCi: {
unsigned Opc = AFI->isThumbFunction() ? ARM::t2MVNi : ARM::MVNi;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc),
MI.getOperand(1).getReg())
.addImm(MI.getOperand(2).getImm())
.addImm(MI.getOperand(3).getImm()) // 'pred'
.add(MI.getOperand(4))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::t2MOVCClsl:
case ARM::t2MOVCClsr:
case ARM::t2MOVCCasr:
case ARM::t2MOVCCror: {
unsigned NewOpc;
switch (Opcode) {
case ARM::t2MOVCClsl: NewOpc = ARM::t2LSLri; break;
case ARM::t2MOVCClsr: NewOpc = ARM::t2LSRri; break;
case ARM::t2MOVCCasr: NewOpc = ARM::t2ASRri; break;
case ARM::t2MOVCCror: NewOpc = ARM::t2RORri; break;
default: llvm_unreachable("unexpeced conditional move");
}
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(NewOpc),
MI.getOperand(1).getReg())
.add(MI.getOperand(2))
.addImm(MI.getOperand(3).getImm())
.addImm(MI.getOperand(4).getImm()) // 'pred'
.add(MI.getOperand(5))
.add(condCodeOp()) // 's' bit
.add(makeImplicit(MI.getOperand(1)));
MI.eraseFromParent();
return true;
}
case ARM::Int_eh_sjlj_dispatchsetup: {
MachineFunction &MF = *MI.getParent()->getParent();
const ARMBaseInstrInfo *AII =
static_cast<const ARMBaseInstrInfo*>(TII);
const ARMBaseRegisterInfo &RI = AII->getRegisterInfo();
// For functions using a base pointer, we rematerialize it (via the frame
// pointer) here since eh.sjlj.setjmp and eh.sjlj.longjmp don't do it
// for us. Otherwise, expand to nothing.
if (RI.hasBasePointer(MF)) {
int32_t NumBytes = AFI->getFramePtrSpillOffset();
Register FramePtr = RI.getFrameRegister(MF);
assert(MF.getSubtarget().getFrameLowering()->hasFP(MF) &&
"base pointer without frame pointer?");
if (AFI->isThumb2Function()) {
emitT2RegPlusImmediate(MBB, MBBI, MI.getDebugLoc(), ARM::R6,
FramePtr, -NumBytes, ARMCC::AL, 0, *TII);
} else if (AFI->isThumbFunction()) {
emitThumbRegPlusImmediate(MBB, MBBI, MI.getDebugLoc(), ARM::R6,
FramePtr, -NumBytes, *TII, RI);
} else {
emitARMRegPlusImmediate(MBB, MBBI, MI.getDebugLoc(), ARM::R6,
FramePtr, -NumBytes, ARMCC::AL, 0,
*TII);
}
// If there's dynamic realignment, adjust for it.
if (RI.needsStackRealignment(MF)) {
MachineFrameInfo &MFI = MF.getFrameInfo();
unsigned MaxAlign = MFI.getMaxAlignment();
assert (!AFI->isThumb1OnlyFunction());
// Emit bic r6, r6, MaxAlign
assert(MaxAlign <= 256 && "The BIC instruction cannot encode "
"immediates larger than 256 with all lower "
"bits set.");
unsigned bicOpc = AFI->isThumbFunction() ?
ARM::t2BICri : ARM::BICri;
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(bicOpc), ARM::R6)
.addReg(ARM::R6, RegState::Kill)
.addImm(MaxAlign - 1)
.add(predOps(ARMCC::AL))
.add(condCodeOp());
}
}
MI.eraseFromParent();
return true;
}
case ARM::MOVsrl_flag:
case ARM::MOVsra_flag: {
// These are just fancy MOVs instructions.
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::MOVsi),
MI.getOperand(0).getReg())
.add(MI.getOperand(1))
.addImm(ARM_AM::getSORegOpc(
(Opcode == ARM::MOVsrl_flag ? ARM_AM::lsr : ARM_AM::asr), 1))
.add(predOps(ARMCC::AL))
.addReg(ARM::CPSR, RegState::Define);
MI.eraseFromParent();
return true;
}
case ARM::RRX: {
// This encodes as "MOVs Rd, Rm, rrx
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::MOVsi),
MI.getOperand(0).getReg())
.add(MI.getOperand(1))
.addImm(ARM_AM::getSORegOpc(ARM_AM::rrx, 0))
.add(predOps(ARMCC::AL))
.add(condCodeOp());
TransferImpOps(MI, MIB, MIB);
MI.eraseFromParent();
return true;
}
case ARM::tTPsoft:
case ARM::TPsoft: {
const bool Thumb = Opcode == ARM::tTPsoft;
MachineInstrBuilder MIB;
if (STI->genLongCalls()) {
MachineFunction *MF = MBB.getParent();
MachineConstantPool *MCP = MF->getConstantPool();
unsigned PCLabelID = AFI->createPICLabelUId();
MachineConstantPoolValue *CPV =
ARMConstantPoolSymbol::Create(MF->getFunction().getContext(),
"__aeabi_read_tp", PCLabelID, 0);
Register Reg = MI.getOperand(0).getReg();
MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(Thumb ? ARM::tLDRpci : ARM::LDRi12), Reg)
.addConstantPoolIndex(MCP->getConstantPoolIndex(CPV, 4));
if (!Thumb)
MIB.addImm(0);
MIB.add(predOps(ARMCC::AL));
MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(Thumb ? ARM::tBLXr : ARM::BLX));
if (Thumb)
MIB.add(predOps(ARMCC::AL));
MIB.addReg(Reg, RegState::Kill);
} else {
MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(Thumb ? ARM::tBL : ARM::BL));
if (Thumb)
MIB.add(predOps(ARMCC::AL));
MIB.addExternalSymbol("__aeabi_read_tp", 0);
}
MIB.cloneMemRefs(MI);
TransferImpOps(MI, MIB, MIB);
MI.getMF()->moveCallSiteInfo(&MI, &*MIB);
MI.eraseFromParent();
return true;
}
case ARM::tLDRpci_pic:
case ARM::t2LDRpci_pic: {
unsigned NewLdOpc = (Opcode == ARM::tLDRpci_pic)
? ARM::tLDRpci : ARM::t2LDRpci;
Register DstReg = MI.getOperand(0).getReg();
bool DstIsDead = MI.getOperand(0).isDead();
MachineInstrBuilder MIB1 =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(NewLdOpc), DstReg)
.add(MI.getOperand(1))
.add(predOps(ARMCC::AL));
MIB1.cloneMemRefs(MI);
MachineInstrBuilder MIB2 =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::tPICADD))
.addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead))
.addReg(DstReg)
.add(MI.getOperand(2));
TransferImpOps(MI, MIB1, MIB2);
MI.eraseFromParent();
return true;
}
case ARM::LDRLIT_ga_abs:
case ARM::LDRLIT_ga_pcrel:
case ARM::LDRLIT_ga_pcrel_ldr:
case ARM::tLDRLIT_ga_abs:
case ARM::tLDRLIT_ga_pcrel: {
Register DstReg = MI.getOperand(0).getReg();
bool DstIsDead = MI.getOperand(0).isDead();
const MachineOperand &MO1 = MI.getOperand(1);
auto Flags = MO1.getTargetFlags();
const GlobalValue *GV = MO1.getGlobal();
bool IsARM =
Opcode != ARM::tLDRLIT_ga_pcrel && Opcode != ARM::tLDRLIT_ga_abs;
bool IsPIC =
Opcode != ARM::LDRLIT_ga_abs && Opcode != ARM::tLDRLIT_ga_abs;
unsigned LDRLITOpc = IsARM ? ARM::LDRi12 : ARM::tLDRpci;
unsigned PICAddOpc =
IsARM
? (Opcode == ARM::LDRLIT_ga_pcrel_ldr ? ARM::PICLDR : ARM::PICADD)
: ARM::tPICADD;
// We need a new const-pool entry to load from.
MachineConstantPool *MCP = MBB.getParent()->getConstantPool();
unsigned ARMPCLabelIndex = 0;
MachineConstantPoolValue *CPV;
if (IsPIC) {
unsigned PCAdj = IsARM ? 8 : 4;
auto Modifier = (Flags & ARMII::MO_GOT)
? ARMCP::GOT_PREL
: ARMCP::no_modifier;
ARMPCLabelIndex = AFI->createPICLabelUId();
CPV = ARMConstantPoolConstant::Create(
GV, ARMPCLabelIndex, ARMCP::CPValue, PCAdj, Modifier,
/*AddCurrentAddr*/ Modifier == ARMCP::GOT_PREL);
} else
CPV = ARMConstantPoolConstant::Create(GV, ARMCP::no_modifier);
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(LDRLITOpc), DstReg)
.addConstantPoolIndex(MCP->getConstantPoolIndex(CPV, 4));
if (IsARM)
MIB.addImm(0);
MIB.add(predOps(ARMCC::AL));
if (IsPIC) {
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(PICAddOpc))
.addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead))
.addReg(DstReg)
.addImm(ARMPCLabelIndex);
if (IsARM)
MIB.add(predOps(ARMCC::AL));
}
MI.eraseFromParent();
return true;
}
case ARM::MOV_ga_pcrel:
case ARM::MOV_ga_pcrel_ldr:
case ARM::t2MOV_ga_pcrel: {
// Expand into movw + movw. Also "add pc" / ldr [pc] in PIC mode.
unsigned LabelId = AFI->createPICLabelUId();
Register DstReg = MI.getOperand(0).getReg();
bool DstIsDead = MI.getOperand(0).isDead();
const MachineOperand &MO1 = MI.getOperand(1);
const GlobalValue *GV = MO1.getGlobal();
unsigned TF = MO1.getTargetFlags();
bool isARM = Opcode != ARM::t2MOV_ga_pcrel;
unsigned LO16Opc = isARM ? ARM::MOVi16_ga_pcrel : ARM::t2MOVi16_ga_pcrel;
unsigned HI16Opc = isARM ? ARM::MOVTi16_ga_pcrel :ARM::t2MOVTi16_ga_pcrel;
unsigned LO16TF = TF | ARMII::MO_LO16;
unsigned HI16TF = TF | ARMII::MO_HI16;
unsigned PICAddOpc = isARM
? (Opcode == ARM::MOV_ga_pcrel_ldr ? ARM::PICLDR : ARM::PICADD)
: ARM::tPICADD;
MachineInstrBuilder MIB1 = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(LO16Opc), DstReg)
.addGlobalAddress(GV, MO1.getOffset(), TF | LO16TF)
.addImm(LabelId);
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(HI16Opc), DstReg)
.addReg(DstReg)
.addGlobalAddress(GV, MO1.getOffset(), TF | HI16TF)
.addImm(LabelId);
MachineInstrBuilder MIB3 = BuildMI(MBB, MBBI, MI.getDebugLoc(),
TII->get(PICAddOpc))
.addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead))
.addReg(DstReg).addImm(LabelId);
if (isARM) {
MIB3.add(predOps(ARMCC::AL));
if (Opcode == ARM::MOV_ga_pcrel_ldr)
MIB3.cloneMemRefs(MI);
}
TransferImpOps(MI, MIB1, MIB3);
MI.eraseFromParent();
return true;
}
case ARM::MOVi32imm:
case ARM::MOVCCi32imm:
case ARM::t2MOVi32imm:
case ARM::t2MOVCCi32imm:
ExpandMOV32BitImm(MBB, MBBI);
return true;
case ARM::SUBS_PC_LR: {
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::SUBri), ARM::PC)
.addReg(ARM::LR)
.add(MI.getOperand(0))
.add(MI.getOperand(1))
.add(MI.getOperand(2))
.addReg(ARM::CPSR, RegState::Undef);
TransferImpOps(MI, MIB, MIB);
MI.eraseFromParent();
return true;
}
case ARM::VLDMQIA: {
unsigned NewOpc = ARM::VLDMDIA;
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(NewOpc));
unsigned OpIdx = 0;
// Grab the Q register destination.
bool DstIsDead = MI.getOperand(OpIdx).isDead();
Register DstReg = MI.getOperand(OpIdx++).getReg();
// Copy the source register.
MIB.add(MI.getOperand(OpIdx++));
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Add the destination operands (D subregs).
Register D0 = TRI->getSubReg(DstReg, ARM::dsub_0);
Register D1 = TRI->getSubReg(DstReg, ARM::dsub_1);
MIB.addReg(D0, RegState::Define | getDeadRegState(DstIsDead))
.addReg(D1, RegState::Define | getDeadRegState(DstIsDead));
// Add an implicit def for the super-register.
MIB.addReg(DstReg, RegState::ImplicitDefine | getDeadRegState(DstIsDead));
TransferImpOps(MI, MIB, MIB);
MIB.cloneMemRefs(MI);
MI.eraseFromParent();
return true;
}
case ARM::VSTMQIA: {
unsigned NewOpc = ARM::VSTMDIA;
MachineInstrBuilder MIB =
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(NewOpc));
unsigned OpIdx = 0;
// Grab the Q register source.
bool SrcIsKill = MI.getOperand(OpIdx).isKill();
Register SrcReg = MI.getOperand(OpIdx++).getReg();
// Copy the destination register.
MachineOperand Dst(MI.getOperand(OpIdx++));
MIB.add(Dst);
// Copy the predicate operands.
MIB.add(MI.getOperand(OpIdx++));
MIB.add(MI.getOperand(OpIdx++));
// Add the source operands (D subregs).
Register D0 = TRI->getSubReg(SrcReg, ARM::dsub_0);
Register D1 = TRI->getSubReg(SrcReg, ARM::dsub_1);
MIB.addReg(D0, SrcIsKill ? RegState::Kill : 0)
.addReg(D1, SrcIsKill ? RegState::Kill : 0);
if (SrcIsKill) // Add an implicit kill for the Q register.
MIB->addRegisterKilled(SrcReg, TRI, true);
TransferImpOps(MI, MIB, MIB);
MIB.cloneMemRefs(MI);
MI.eraseFromParent();
return true;
}
case ARM::VLD2q8Pseudo:
case ARM::VLD2q16Pseudo:
case ARM::VLD2q32Pseudo:
case ARM::VLD2q8PseudoWB_fixed:
case ARM::VLD2q16PseudoWB_fixed:
case ARM::VLD2q32PseudoWB_fixed:
case ARM::VLD2q8PseudoWB_register:
case ARM::VLD2q16PseudoWB_register:
case ARM::VLD2q32PseudoWB_register:
case ARM::VLD3d8Pseudo:
case ARM::VLD3d16Pseudo:
case ARM::VLD3d32Pseudo:
case ARM::VLD1d8TPseudo:
case ARM::VLD1d16TPseudo:
case ARM::VLD1d32TPseudo:
case ARM::VLD1d64TPseudo:
case ARM::VLD1d64TPseudoWB_fixed:
case ARM::VLD1d64TPseudoWB_register:
case ARM::VLD3d8Pseudo_UPD:
case ARM::VLD3d16Pseudo_UPD:
case ARM::VLD3d32Pseudo_UPD:
case ARM::VLD3q8Pseudo_UPD:
case ARM::VLD3q16Pseudo_UPD:
case ARM::VLD3q32Pseudo_UPD:
case ARM::VLD3q8oddPseudo:
case ARM::VLD3q16oddPseudo:
case ARM::VLD3q32oddPseudo:
case ARM::VLD3q8oddPseudo_UPD:
case ARM::VLD3q16oddPseudo_UPD:
case ARM::VLD3q32oddPseudo_UPD:
case ARM::VLD4d8Pseudo:
case ARM::VLD4d16Pseudo:
case ARM::VLD4d32Pseudo:
case ARM::VLD1d8QPseudo:
case ARM::VLD1d16QPseudo:
case ARM::VLD1d32QPseudo:
case ARM::VLD1d64QPseudo:
case ARM::VLD1d64QPseudoWB_fixed:
case ARM::VLD1d64QPseudoWB_register:
case ARM::VLD1q8HighQPseudo:
case ARM::VLD1q8LowQPseudo_UPD:
case ARM::VLD1q8HighTPseudo:
case ARM::VLD1q8LowTPseudo_UPD:
case ARM::VLD1q16HighQPseudo:
case ARM::VLD1q16LowQPseudo_UPD:
case ARM::VLD1q16HighTPseudo:
case ARM::VLD1q16LowTPseudo_UPD:
case ARM::VLD1q32HighQPseudo:
case ARM::VLD1q32LowQPseudo_UPD:
case ARM::VLD1q32HighTPseudo:
case ARM::VLD1q32LowTPseudo_UPD:
case ARM::VLD1q64HighQPseudo:
case ARM::VLD1q64LowQPseudo_UPD:
case ARM::VLD1q64HighTPseudo:
case ARM::VLD1q64LowTPseudo_UPD:
case ARM::VLD4d8Pseudo_UPD:
case ARM::VLD4d16Pseudo_UPD:
case ARM::VLD4d32Pseudo_UPD:
case ARM::VLD4q8Pseudo_UPD:
case ARM::VLD4q16Pseudo_UPD:
case ARM::VLD4q32Pseudo_UPD:
case ARM::VLD4q8oddPseudo:
case ARM::VLD4q16oddPseudo:
case ARM::VLD4q32oddPseudo:
case ARM::VLD4q8oddPseudo_UPD:
case ARM::VLD4q16oddPseudo_UPD:
case ARM::VLD4q32oddPseudo_UPD:
case ARM::VLD3DUPd8Pseudo:
case ARM::VLD3DUPd16Pseudo:
case ARM::VLD3DUPd32Pseudo:
case ARM::VLD3DUPd8Pseudo_UPD:
case ARM::VLD3DUPd16Pseudo_UPD:
case ARM::VLD3DUPd32Pseudo_UPD:
case ARM::VLD4DUPd8Pseudo:
case ARM::VLD4DUPd16Pseudo:
case ARM::VLD4DUPd32Pseudo:
case ARM::VLD4DUPd8Pseudo_UPD:
case ARM::VLD4DUPd16Pseudo_UPD:
case ARM::VLD4DUPd32Pseudo_UPD:
case ARM::VLD2DUPq8EvenPseudo:
case ARM::VLD2DUPq8OddPseudo:
case ARM::VLD2DUPq16EvenPseudo:
case ARM::VLD2DUPq16OddPseudo:
case ARM::VLD2DUPq32EvenPseudo:
case ARM::VLD2DUPq32OddPseudo:
case ARM::VLD3DUPq8EvenPseudo:
case ARM::VLD3DUPq8OddPseudo:
case ARM::VLD3DUPq16EvenPseudo:
case ARM::VLD3DUPq16OddPseudo:
case ARM::VLD3DUPq32EvenPseudo:
case ARM::VLD3DUPq32OddPseudo:
case ARM::VLD4DUPq8EvenPseudo:
case ARM::VLD4DUPq8OddPseudo:
case ARM::VLD4DUPq16EvenPseudo:
case ARM::VLD4DUPq16OddPseudo:
case ARM::VLD4DUPq32EvenPseudo:
case ARM::VLD4DUPq32OddPseudo:
ExpandVLD(MBBI);
return true;
case ARM::VST2q8Pseudo:
case ARM::VST2q16Pseudo:
case ARM::VST2q32Pseudo:
case ARM::VST2q8PseudoWB_fixed:
case ARM::VST2q16PseudoWB_fixed:
case ARM::VST2q32PseudoWB_fixed:
case ARM::VST2q8PseudoWB_register:
case ARM::VST2q16PseudoWB_register:
case ARM::VST2q32PseudoWB_register:
case ARM::VST3d8Pseudo:
case ARM::VST3d16Pseudo:
case ARM::VST3d32Pseudo:
case ARM::VST1d8TPseudo:
case ARM::VST1d16TPseudo:
case ARM::VST1d32TPseudo:
case ARM::VST1d64TPseudo:
case ARM::VST3d8Pseudo_UPD:
case ARM::VST3d16Pseudo_UPD:
case ARM::VST3d32Pseudo_UPD:
case ARM::VST1d64TPseudoWB_fixed:
case ARM::VST1d64TPseudoWB_register:
case ARM::VST3q8Pseudo_UPD:
case ARM::VST3q16Pseudo_UPD:
case ARM::VST3q32Pseudo_UPD:
case ARM::VST3q8oddPseudo:
case ARM::VST3q16oddPseudo:
case ARM::VST3q32oddPseudo:
case ARM::VST3q8oddPseudo_UPD:
case ARM::VST3q16oddPseudo_UPD:
case ARM::VST3q32oddPseudo_UPD:
case ARM::VST4d8Pseudo:
case ARM::VST4d16Pseudo:
case ARM::VST4d32Pseudo:
case ARM::VST1d8QPseudo:
case ARM::VST1d16QPseudo:
case ARM::VST1d32QPseudo:
case ARM::VST1d64QPseudo:
case ARM::VST4d8Pseudo_UPD:
case ARM::VST4d16Pseudo_UPD:
case ARM::VST4d32Pseudo_UPD:
case ARM::VST1d64QPseudoWB_fixed:
case ARM::VST1d64QPseudoWB_register:
case ARM::VST1q8HighQPseudo:
case ARM::VST1q8LowQPseudo_UPD:
case ARM::VST1q8HighTPseudo:
case ARM::VST1q8LowTPseudo_UPD:
case ARM::VST1q16HighQPseudo:
case ARM::VST1q16LowQPseudo_UPD:
case ARM::VST1q16HighTPseudo:
case ARM::VST1q16LowTPseudo_UPD:
case ARM::VST1q32HighQPseudo:
case ARM::VST1q32LowQPseudo_UPD:
case ARM::VST1q32HighTPseudo:
case ARM::VST1q32LowTPseudo_UPD:
case ARM::VST1q64HighQPseudo:
case ARM::VST1q64LowQPseudo_UPD:
case ARM::VST1q64HighTPseudo:
case ARM::VST1q64LowTPseudo_UPD:
case ARM::VST4q8Pseudo_UPD:
case ARM::VST4q16Pseudo_UPD:
case ARM::VST4q32Pseudo_UPD:
case ARM::VST4q8oddPseudo:
case ARM::VST4q16oddPseudo:
case ARM::VST4q32oddPseudo:
case ARM::VST4q8oddPseudo_UPD:
case ARM::VST4q16oddPseudo_UPD:
case ARM::VST4q32oddPseudo_UPD:
ExpandVST(MBBI);
return true;
case ARM::VLD1LNq8Pseudo:
case ARM::VLD1LNq16Pseudo:
case ARM::VLD1LNq32Pseudo:
case ARM::VLD1LNq8Pseudo_UPD:
case ARM::VLD1LNq16Pseudo_UPD:
case ARM::VLD1LNq32Pseudo_UPD:
case ARM::VLD2LNd8Pseudo:
case ARM::VLD2LNd16Pseudo:
case ARM::VLD2LNd32Pseudo:
case ARM::VLD2LNq16Pseudo:
case ARM::VLD2LNq32Pseudo:
case ARM::VLD2LNd8Pseudo_UPD:
case ARM::VLD2LNd16Pseudo_UPD:
case ARM::VLD2LNd32Pseudo_UPD:
case ARM::VLD2LNq16Pseudo_UPD:
case ARM::VLD2LNq32Pseudo_UPD:
case ARM::VLD3LNd8Pseudo:
case ARM::VLD3LNd16Pseudo:
case ARM::VLD3LNd32Pseudo:
case ARM::VLD3LNq16Pseudo:
case ARM::VLD3LNq32Pseudo:
case ARM::VLD3LNd8Pseudo_UPD:
case ARM::VLD3LNd16Pseudo_UPD:
case ARM::VLD3LNd32Pseudo_UPD:
case ARM::VLD3LNq16Pseudo_UPD:
case ARM::VLD3LNq32Pseudo_UPD:
case ARM::VLD4LNd8Pseudo:
case ARM::VLD4LNd16Pseudo:
case ARM::VLD4LNd32Pseudo:
case ARM::VLD4LNq16Pseudo:
case ARM::VLD4LNq32Pseudo:
case ARM::VLD4LNd8Pseudo_UPD:
case ARM::VLD4LNd16Pseudo_UPD:
case ARM::VLD4LNd32Pseudo_UPD:
case ARM::VLD4LNq16Pseudo_UPD:
case ARM::VLD4LNq32Pseudo_UPD:
case ARM::VST1LNq8Pseudo:
case ARM::VST1LNq16Pseudo:
case ARM::VST1LNq32Pseudo:
case ARM::VST1LNq8Pseudo_UPD:
case ARM::VST1LNq16Pseudo_UPD:
case ARM::VST1LNq32Pseudo_UPD:
case ARM::VST2LNd8Pseudo:
case ARM::VST2LNd16Pseudo:
case ARM::VST2LNd32Pseudo:
case ARM::VST2LNq16Pseudo:
case ARM::VST2LNq32Pseudo:
case ARM::VST2LNd8Pseudo_UPD:
case ARM::VST2LNd16Pseudo_UPD:
case ARM::VST2LNd32Pseudo_UPD:
case ARM::VST2LNq16Pseudo_UPD:
case ARM::VST2LNq32Pseudo_UPD:
case ARM::VST3LNd8Pseudo:
case ARM::VST3LNd16Pseudo:
case ARM::VST3LNd32Pseudo:
case ARM::VST3LNq16Pseudo:
case ARM::VST3LNq32Pseudo:
case ARM::VST3LNd8Pseudo_UPD:
case ARM::VST3LNd16Pseudo_UPD:
case ARM::VST3LNd32Pseudo_UPD:
case ARM::VST3LNq16Pseudo_UPD:
case ARM::VST3LNq32Pseudo_UPD:
case ARM::VST4LNd8Pseudo:
case ARM::VST4LNd16Pseudo:
case ARM::VST4LNd32Pseudo:
case ARM::VST4LNq16Pseudo:
case ARM::VST4LNq32Pseudo:
case ARM::VST4LNd8Pseudo_UPD:
case ARM::VST4LNd16Pseudo_UPD:
case ARM::VST4LNd32Pseudo_UPD:
case ARM::VST4LNq16Pseudo_UPD:
case ARM::VST4LNq32Pseudo_UPD:
ExpandLaneOp(MBBI);
return true;
case ARM::VTBL3Pseudo: ExpandVTBL(MBBI, ARM::VTBL3, false); return true;
case ARM::VTBL4Pseudo: ExpandVTBL(MBBI, ARM::VTBL4, false); return true;
case ARM::VTBX3Pseudo: ExpandVTBL(MBBI, ARM::VTBX3, true); return true;
case ARM::VTBX4Pseudo: ExpandVTBL(MBBI, ARM::VTBX4, true); return true;
case ARM::CMP_SWAP_8:
if (STI->isThumb())
return ExpandCMP_SWAP(MBB, MBBI, ARM::t2LDREXB, ARM::t2STREXB,
ARM::tUXTB, NextMBBI);
else
return ExpandCMP_SWAP(MBB, MBBI, ARM::LDREXB, ARM::STREXB,
ARM::UXTB, NextMBBI);
case ARM::CMP_SWAP_16:
if (STI->isThumb())
return ExpandCMP_SWAP(MBB, MBBI, ARM::t2LDREXH, ARM::t2STREXH,
ARM::tUXTH, NextMBBI);
else
return ExpandCMP_SWAP(MBB, MBBI, ARM::LDREXH, ARM::STREXH,
ARM::UXTH, NextMBBI);
case ARM::CMP_SWAP_32:
if (STI->isThumb())
return ExpandCMP_SWAP(MBB, MBBI, ARM::t2LDREX, ARM::t2STREX, 0,
NextMBBI);
else
return ExpandCMP_SWAP(MBB, MBBI, ARM::LDREX, ARM::STREX, 0, NextMBBI);
case ARM::CMP_SWAP_64:
return ExpandCMP_SWAP_64(MBB, MBBI, NextMBBI);
case ARM::tBL_PUSHLR:
case ARM::BL_PUSHLR: {
const bool Thumb = Opcode == ARM::tBL_PUSHLR;
Register Reg = MI.getOperand(0).getReg();
assert(Reg == ARM::LR && "expect LR register!");
MachineInstrBuilder MIB;
if (Thumb) {
// push {lr}
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::tPUSH))
.add(predOps(ARMCC::AL))
.addReg(Reg);
// bl __gnu_mcount_nc
MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::tBL));
} else {
// stmdb sp!, {lr}
BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::STMDB_UPD))
.addReg(ARM::SP, RegState::Define)
.addReg(ARM::SP)
.add(predOps(ARMCC::AL))
.addReg(Reg);
// bl __gnu_mcount_nc
MIB = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(ARM::BL));
}
MIB.cloneMemRefs(MI);
for (unsigned i = 1; i < MI.getNumOperands(); ++i) MIB.add(MI.getOperand(i));
MI.eraseFromParent();
return true;
}
+ case ARM::LOADDUAL:
+ case ARM::STOREDUAL: {
+ Register PairReg = MI.getOperand(0).getReg();
+
+ MachineInstrBuilder MIB =
+ BuildMI(MBB, MBBI, MI.getDebugLoc(),
+ TII->get(Opcode == ARM::LOADDUAL ? ARM::LDRD : ARM::STRD))
+ .addReg(TRI->getSubReg(PairReg, ARM::gsub_0),
+ Opcode == ARM::LOADDUAL ? RegState::Define : 0)
+ .addReg(TRI->getSubReg(PairReg, ARM::gsub_1),
+ Opcode == ARM::LOADDUAL ? RegState::Define : 0);
+ for (unsigned i = 1; i < MI.getNumOperands(); i++)
+ MIB.add(MI.getOperand(i));
+ MIB.add(predOps(ARMCC::AL));
+ MIB.cloneMemRefs(MI);
+ MI.eraseFromParent();
+ return true;
+ }
}
}
bool ARMExpandPseudo::ExpandMBB(MachineBasicBlock &MBB) {
bool Modified = false;
MachineBasicBlock::iterator MBBI = MBB.begin(), E = MBB.end();
while (MBBI != E) {
MachineBasicBlock::iterator NMBBI = std::next(MBBI);
Modified |= ExpandMI(MBB, MBBI, NMBBI);
MBBI = NMBBI;
}
return Modified;
}
bool ARMExpandPseudo::runOnMachineFunction(MachineFunction &MF) {
STI = &static_cast<const ARMSubtarget &>(MF.getSubtarget());
TII = STI->getInstrInfo();
TRI = STI->getRegisterInfo();
AFI = MF.getInfo<ARMFunctionInfo>();
LLVM_DEBUG(dbgs() << "********** ARM EXPAND PSEUDO INSTRUCTIONS **********\n"
<< "********** Function: " << MF.getName() << '\n');
bool Modified = false;
for (MachineBasicBlock &MBB : MF)
Modified |= ExpandMBB(MBB);
if (VerifyARMPseudo)
MF.verify(this, "After expanding ARM pseudo instructions.");
LLVM_DEBUG(dbgs() << "***************************************************\n");
return Modified;
}
/// createARMExpandPseudoPass - returns an instance of the pseudo instruction
/// expansion pass.
FunctionPass *llvm::createARMExpandPseudoPass() {
return new ARMExpandPseudo();
}
diff --git a/llvm/lib/Target/ARM/ARMISelDAGToDAG.cpp b/llvm/lib/Target/ARM/ARMISelDAGToDAG.cpp
index 1f998de..dcdec77 100644
--- a/llvm/lib/Target/ARM/ARMISelDAGToDAG.cpp
+++ b/llvm/lib/Target/ARM/ARMISelDAGToDAG.cpp
@@ -1,5079 +1,5128 @@
//===-- ARMISelDAGToDAG.cpp - A dag to dag inst selector for ARM ----------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file defines an instruction selector for the ARM target.
//
//===----------------------------------------------------------------------===//
#include "ARM.h"
#include "ARMBaseInstrInfo.h"
#include "ARMTargetMachine.h"
#include "MCTargetDesc/ARMAddressingModes.h"
#include "Utils/ARMBaseInfo.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/CodeGen/MachineFrameInfo.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/SelectionDAG.h"
#include "llvm/CodeGen/SelectionDAGISel.h"
#include "llvm/CodeGen/TargetLowering.h"
#include "llvm/IR/CallingConv.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/IntrinsicsARM.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Target/TargetOptions.h"
using namespace llvm;
#define DEBUG_TYPE "arm-isel"
static cl::opt<bool>
DisableShifterOp("disable-shifter-op", cl::Hidden,
cl::desc("Disable isel of shifter-op"),
cl::init(false));
//===--------------------------------------------------------------------===//
/// ARMDAGToDAGISel - ARM specific code to select ARM machine
/// instructions for SelectionDAG operations.
///
namespace {
class ARMDAGToDAGISel : public SelectionDAGISel {
/// Subtarget - Keep a pointer to the ARMSubtarget around so that we can
/// make the right decision when generating code for different targets.
const ARMSubtarget *Subtarget;
public:
explicit ARMDAGToDAGISel(ARMBaseTargetMachine &tm, CodeGenOpt::Level OptLevel)
: SelectionDAGISel(tm, OptLevel) {}
bool runOnMachineFunction(MachineFunction &MF) override {
// Reset the subtarget each time through.
Subtarget = &MF.getSubtarget<ARMSubtarget>();
SelectionDAGISel::runOnMachineFunction(MF);
return true;
}
StringRef getPassName() const override { return "ARM Instruction Selection"; }
void PreprocessISelDAG() override;
/// getI32Imm - Return a target constant of type i32 with the specified
/// value.
inline SDValue getI32Imm(unsigned Imm, const SDLoc &dl) {
return CurDAG->getTargetConstant(Imm, dl, MVT::i32);
}
void Select(SDNode *N) override;
bool hasNoVMLxHazardUse(SDNode *N) const;
bool isShifterOpProfitable(const SDValue &Shift,
ARM_AM::ShiftOpc ShOpcVal, unsigned ShAmt);
bool SelectRegShifterOperand(SDValue N, SDValue &A,
SDValue &B, SDValue &C,
bool CheckProfitability = true);
bool SelectImmShifterOperand(SDValue N, SDValue &A,
SDValue &B, bool CheckProfitability = true);
bool SelectShiftRegShifterOperand(SDValue N, SDValue &A,
SDValue &B, SDValue &C) {
// Don't apply the profitability check
return SelectRegShifterOperand(N, A, B, C, false);
}
bool SelectShiftImmShifterOperand(SDValue N, SDValue &A,
SDValue &B) {
// Don't apply the profitability check
return SelectImmShifterOperand(N, A, B, false);
}
bool SelectAddLikeOr(SDNode *Parent, SDValue N, SDValue &Out);
bool SelectAddrModeImm12(SDValue N, SDValue &Base, SDValue &OffImm);
bool SelectLdStSOReg(SDValue N, SDValue &Base, SDValue &Offset, SDValue &Opc);
bool SelectCMOVPred(SDValue N, SDValue &Pred, SDValue &Reg) {
const ConstantSDNode *CN = cast<ConstantSDNode>(N);
Pred = CurDAG->getTargetConstant(CN->getZExtValue(), SDLoc(N), MVT::i32);
Reg = CurDAG->getRegister(ARM::CPSR, MVT::i32);
return true;
}
bool SelectAddrMode2OffsetReg(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc);
bool SelectAddrMode2OffsetImm(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc);
bool SelectAddrMode2OffsetImmPre(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc);
bool SelectAddrOffsetNone(SDValue N, SDValue &Base);
bool SelectAddrMode3(SDValue N, SDValue &Base,
SDValue &Offset, SDValue &Opc);
bool SelectAddrMode3Offset(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc);
bool IsAddressingMode5(SDValue N, SDValue &Base, SDValue &Offset, bool FP16);
bool SelectAddrMode5(SDValue N, SDValue &Base, SDValue &Offset);
bool SelectAddrMode5FP16(SDValue N, SDValue &Base, SDValue &Offset);
bool SelectAddrMode6(SDNode *Parent, SDValue N, SDValue &Addr,SDValue &Align);
bool SelectAddrMode6Offset(SDNode *Op, SDValue N, SDValue &Offset);
bool SelectAddrModePC(SDValue N, SDValue &Offset, SDValue &Label);
// Thumb Addressing Modes:
bool SelectThumbAddrModeRR(SDValue N, SDValue &Base, SDValue &Offset);
bool SelectThumbAddrModeRRSext(SDValue N, SDValue &Base, SDValue &Offset);
bool SelectThumbAddrModeImm5S(SDValue N, unsigned Scale, SDValue &Base,
SDValue &OffImm);
bool SelectThumbAddrModeImm5S1(SDValue N, SDValue &Base,
SDValue &OffImm);
bool SelectThumbAddrModeImm5S2(SDValue N, SDValue &Base,
SDValue &OffImm);
bool SelectThumbAddrModeImm5S4(SDValue N, SDValue &Base,
SDValue &OffImm);
bool SelectThumbAddrModeSP(SDValue N, SDValue &Base, SDValue &OffImm);
template <unsigned Shift>
bool SelectTAddrModeImm7(SDValue N, SDValue &Base, SDValue &OffImm);
// Thumb 2 Addressing Modes:
bool SelectT2AddrModeImm12(SDValue N, SDValue &Base, SDValue &OffImm);
+ template <unsigned Shift>
+ bool SelectT2AddrModeImm8(SDValue N, SDValue &Base, SDValue &OffImm);
bool SelectT2AddrModeImm8(SDValue N, SDValue &Base,
SDValue &OffImm);
bool SelectT2AddrModeImm8Offset(SDNode *Op, SDValue N,
SDValue &OffImm);
template <unsigned Shift>
bool SelectT2AddrModeImm7Offset(SDNode *Op, SDValue N, SDValue &OffImm);
bool SelectT2AddrModeImm7Offset(SDNode *Op, SDValue N, SDValue &OffImm,
unsigned Shift);
template <unsigned Shift>
bool SelectT2AddrModeImm7(SDValue N, SDValue &Base, SDValue &OffImm);
bool SelectT2AddrModeSoReg(SDValue N, SDValue &Base,
SDValue &OffReg, SDValue &ShImm);
bool SelectT2AddrModeExclusive(SDValue N, SDValue &Base, SDValue &OffImm);
template<int Min, int Max>
bool SelectImmediateInRange(SDValue N, SDValue &OffImm);
inline bool is_so_imm(unsigned Imm) const {
return ARM_AM::getSOImmVal(Imm) != -1;
}
inline bool is_so_imm_not(unsigned Imm) const {
return ARM_AM::getSOImmVal(~Imm) != -1;
}
inline bool is_t2_so_imm(unsigned Imm) const {
return ARM_AM::getT2SOImmVal(Imm) != -1;
}
inline bool is_t2_so_imm_not(unsigned Imm) const {
return ARM_AM::getT2SOImmVal(~Imm) != -1;
}
// Include the pieces autogenerated from the target description.
#include "ARMGenDAGISel.inc"
private:
void transferMemOperands(SDNode *Src, SDNode *Dst);
/// Indexed (pre/post inc/dec) load matching code for ARM.
bool tryARMIndexedLoad(SDNode *N);
bool tryT1IndexedLoad(SDNode *N);
bool tryT2IndexedLoad(SDNode *N);
bool tryMVEIndexedLoad(SDNode *N);
/// SelectVLD - Select NEON load intrinsics. NumVecs should be
/// 1, 2, 3 or 4. The opcode arrays specify the instructions used for
/// loads of D registers and even subregs and odd subregs of Q registers.
/// For NumVecs <= 2, QOpcodes1 is not used.
void SelectVLD(SDNode *N, bool isUpdating, unsigned NumVecs,
const uint16_t *DOpcodes, const uint16_t *QOpcodes0,
const uint16_t *QOpcodes1);
/// SelectVST - Select NEON store intrinsics. NumVecs should
/// be 1, 2, 3 or 4. The opcode arrays specify the instructions used for
/// stores of D registers and even subregs and odd subregs of Q registers.
/// For NumVecs <= 2, QOpcodes1 is not used.
void SelectVST(SDNode *N, bool isUpdating, unsigned NumVecs,
const uint16_t *DOpcodes, const uint16_t *QOpcodes0,
const uint16_t *QOpcodes1);
/// SelectVLDSTLane - Select NEON load/store lane intrinsics. NumVecs should
/// be 2, 3 or 4. The opcode arrays specify the instructions used for
/// load/store of D registers and Q registers.
void SelectVLDSTLane(SDNode *N, bool IsLoad, bool isUpdating,
unsigned NumVecs, const uint16_t *DOpcodes,
const uint16_t *QOpcodes);
/// Helper functions for setting up clusters of MVE predication operands.
template <typename SDValueVector>
void AddMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc,
SDValue PredicateMask);
template <typename SDValueVector>
void AddMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc,
SDValue PredicateMask, SDValue Inactive);
template <typename SDValueVector>
void AddEmptyMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc);
template <typename SDValueVector>
void AddEmptyMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc, EVT InactiveTy);
/// SelectMVE_WB - Select MVE writeback load/store intrinsics.
void SelectMVE_WB(SDNode *N, const uint16_t *Opcodes, bool Predicated);
/// SelectMVE_LongShift - Select MVE 64-bit scalar shift intrinsics.
void SelectMVE_LongShift(SDNode *N, uint16_t Opcode, bool Immediate,
bool HasSaturationOperand);
/// SelectMVE_VADCSBC - Select MVE vector add/sub-with-carry intrinsics.
void SelectMVE_VADCSBC(SDNode *N, uint16_t OpcodeWithCarry,
uint16_t OpcodeWithNoCarry, bool Add, bool Predicated);
/// Select long MVE vector reductions with two vector operands
/// Stride is the number of vector element widths the instruction can operate
/// on:
/// 2 for long non-rounding variants, vml{a,s}ldav[a][x]: [i16, i32]
/// 1 for long rounding variants: vrml{a,s}ldavh[a][x]: [i32]
/// Stride is used when addressing the OpcodesS array which contains multiple
/// opcodes for each element width.
/// TySize is the index into the list of element types listed above
void SelectBaseMVE_VMLLDAV(SDNode *N, bool Predicated,
const uint16_t *OpcodesS, const uint16_t *OpcodesU,
size_t Stride, size_t TySize);
/// Select a 64-bit MVE vector reduction with two vector operands
/// arm_mve_vmlldava_[predicated]
void SelectMVE_VMLLDAV(SDNode *N, bool Predicated, const uint16_t *OpcodesS,
const uint16_t *OpcodesU);
/// Select a 72-bit MVE vector rounding reduction with two vector operands
/// int_arm_mve_vrmlldavha[_predicated]
void SelectMVE_VRMLLDAVH(SDNode *N, bool Predicated, const uint16_t *OpcodesS,
const uint16_t *OpcodesU);
/// SelectMVE_VLD - Select MVE interleaving load intrinsics. NumVecs
/// should be 2 or 4. The opcode array specifies the instructions
/// used for 8, 16 and 32-bit lane sizes respectively, and each
/// pointer points to a set of NumVecs sub-opcodes used for the
/// different stages (e.g. VLD20 versus VLD21) of each load family.
void SelectMVE_VLD(SDNode *N, unsigned NumVecs,
const uint16_t *const *Opcodes);
/// SelectVLDDup - Select NEON load-duplicate intrinsics. NumVecs
/// should be 1, 2, 3 or 4. The opcode array specifies the instructions used
/// for loading D registers.
void SelectVLDDup(SDNode *N, bool IsIntrinsic, bool isUpdating,
unsigned NumVecs, const uint16_t *DOpcodes,
const uint16_t *QOpcodes0 = nullptr,
const uint16_t *QOpcodes1 = nullptr);
/// Try to select SBFX/UBFX instructions for ARM.
bool tryV6T2BitfieldExtractOp(SDNode *N, bool isSigned);
// Select special operations if node forms integer ABS pattern
bool tryABSOp(SDNode *N);
bool tryReadRegister(SDNode *N);
bool tryWriteRegister(SDNode *N);
bool tryInlineAsm(SDNode *N);
void SelectCMPZ(SDNode *N, bool &SwitchEQNEToPLMI);
void SelectCMP_SWAP(SDNode *N);
/// SelectInlineAsmMemoryOperand - Implement addressing mode selection for
/// inline asm expressions.
bool SelectInlineAsmMemoryOperand(const SDValue &Op, unsigned ConstraintID,
std::vector<SDValue> &OutOps) override;
// Form pairs of consecutive R, S, D, or Q registers.
SDNode *createGPRPairNode(EVT VT, SDValue V0, SDValue V1);
SDNode *createSRegPairNode(EVT VT, SDValue V0, SDValue V1);
SDNode *createDRegPairNode(EVT VT, SDValue V0, SDValue V1);
SDNode *createQRegPairNode(EVT VT, SDValue V0, SDValue V1);
// Form sequences of 4 consecutive S, D, or Q registers.
SDNode *createQuadSRegsNode(EVT VT, SDValue V0, SDValue V1, SDValue V2, SDValue V3);
SDNode *createQuadDRegsNode(EVT VT, SDValue V0, SDValue V1, SDValue V2, SDValue V3);
SDNode *createQuadQRegsNode(EVT VT, SDValue V0, SDValue V1, SDValue V2, SDValue V3);
// Get the alignment operand for a NEON VLD or VST instruction.
SDValue GetVLDSTAlign(SDValue Align, const SDLoc &dl, unsigned NumVecs,
bool is64BitVector);
/// Checks if N is a multiplication by a constant where we can extract out a
/// power of two from the constant so that it can be used in a shift, but only
/// if it simplifies the materialization of the constant. Returns true if it
/// is, and assigns to PowerOfTwo the power of two that should be extracted
/// out and to NewMulConst the new constant to be multiplied by.
bool canExtractShiftFromMul(const SDValue &N, unsigned MaxShift,
unsigned &PowerOfTwo, SDValue &NewMulConst) const;
/// Replace N with M in CurDAG, in a way that also ensures that M gets
/// selected when N would have been selected.
void replaceDAGValue(const SDValue &N, SDValue M);
};
}
/// isInt32Immediate - This method tests to see if the node is a 32-bit constant
/// operand. If so Imm will receive the 32-bit value.
static bool isInt32Immediate(SDNode *N, unsigned &Imm) {
if (N->getOpcode() == ISD::Constant && N->getValueType(0) == MVT::i32) {
Imm = cast<ConstantSDNode>(N)->getZExtValue();
return true;
}
return false;
}
// isInt32Immediate - This method tests to see if a constant operand.
// If so Imm will receive the 32 bit value.
static bool isInt32Immediate(SDValue N, unsigned &Imm) {
return isInt32Immediate(N.getNode(), Imm);
}
// isOpcWithIntImmediate - This method tests to see if the node is a specific
// opcode and that it has a immediate integer right operand.
// If so Imm will receive the 32 bit value.
static bool isOpcWithIntImmediate(SDNode *N, unsigned Opc, unsigned& Imm) {
return N->getOpcode() == Opc &&
isInt32Immediate(N->getOperand(1).getNode(), Imm);
}
/// Check whether a particular node is a constant value representable as
/// (N * Scale) where (N in [\p RangeMin, \p RangeMax).
///
/// \param ScaledConstant [out] - On success, the pre-scaled constant value.
static bool isScaledConstantInRange(SDValue Node, int Scale,
int RangeMin, int RangeMax,
int &ScaledConstant) {
assert(Scale > 0 && "Invalid scale!");
// Check that this is a constant.
const ConstantSDNode *C = dyn_cast<ConstantSDNode>(Node);
if (!C)
return false;
ScaledConstant = (int) C->getZExtValue();
if ((ScaledConstant % Scale) != 0)
return false;
ScaledConstant /= Scale;
return ScaledConstant >= RangeMin && ScaledConstant < RangeMax;
}
void ARMDAGToDAGISel::PreprocessISelDAG() {
if (!Subtarget->hasV6T2Ops())
return;
bool isThumb2 = Subtarget->isThumb();
for (SelectionDAG::allnodes_iterator I = CurDAG->allnodes_begin(),
E = CurDAG->allnodes_end(); I != E; ) {
SDNode *N = &*I++; // Preincrement iterator to avoid invalidation issues.
if (N->getOpcode() != ISD::ADD)
continue;
// Look for (add X1, (and (srl X2, c1), c2)) where c2 is constant with
// leading zeros, followed by consecutive set bits, followed by 1 or 2
// trailing zeros, e.g. 1020.
// Transform the expression to
// (add X1, (shl (and (srl X2, c1), (c2>>tz)), tz)) where tz is the number
// of trailing zeros of c2. The left shift would be folded as an shifter
// operand of 'add' and the 'and' and 'srl' would become a bits extraction
// node (UBFX).
SDValue N0 = N->getOperand(0);
SDValue N1 = N->getOperand(1);
unsigned And_imm = 0;
if (!isOpcWithIntImmediate(N1.getNode(), ISD::AND, And_imm)) {
if (isOpcWithIntImmediate(N0.getNode(), ISD::AND, And_imm))
std::swap(N0, N1);
}
if (!And_imm)
continue;
// Check if the AND mask is an immediate of the form: 000.....1111111100
unsigned TZ = countTrailingZeros(And_imm);
if (TZ != 1 && TZ != 2)
// Be conservative here. Shifter operands aren't always free. e.g. On
// Swift, left shifter operand of 1 / 2 for free but others are not.
// e.g.
// ubfx r3, r1, #16, #8
// ldr.w r3, [r0, r3, lsl #2]
// vs.
// mov.w r9, #1020
// and.w r2, r9, r1, lsr #14
// ldr r2, [r0, r2]
continue;
And_imm >>= TZ;
if (And_imm & (And_imm + 1))
continue;
// Look for (and (srl X, c1), c2).
SDValue Srl = N1.getOperand(0);
unsigned Srl_imm = 0;
if (!isOpcWithIntImmediate(Srl.getNode(), ISD::SRL, Srl_imm) ||
(Srl_imm <= 2))
continue;
// Make sure first operand is not a shifter operand which would prevent
// folding of the left shift.
SDValue CPTmp0;
SDValue CPTmp1;
SDValue CPTmp2;
if (isThumb2) {
if (SelectImmShifterOperand(N0, CPTmp0, CPTmp1))
continue;
} else {
if (SelectImmShifterOperand(N0, CPTmp0, CPTmp1) ||
SelectRegShifterOperand(N0, CPTmp0, CPTmp1, CPTmp2))
continue;
}
// Now make the transformation.
Srl = CurDAG->getNode(ISD::SRL, SDLoc(Srl), MVT::i32,
Srl.getOperand(0),
CurDAG->getConstant(Srl_imm + TZ, SDLoc(Srl),
MVT::i32));
N1 = CurDAG->getNode(ISD::AND, SDLoc(N1), MVT::i32,
Srl,
CurDAG->getConstant(And_imm, SDLoc(Srl), MVT::i32));
N1 = CurDAG->getNode(ISD::SHL, SDLoc(N1), MVT::i32,
N1, CurDAG->getConstant(TZ, SDLoc(Srl), MVT::i32));
CurDAG->UpdateNodeOperands(N, N0, N1);
}
}
/// hasNoVMLxHazardUse - Return true if it's desirable to select a FP MLA / MLS
/// node. VFP / NEON fp VMLA / VMLS instructions have special RAW hazards (at
/// least on current ARM implementations) which should be avoidded.
bool ARMDAGToDAGISel::hasNoVMLxHazardUse(SDNode *N) const {
if (OptLevel == CodeGenOpt::None)
return true;
if (!Subtarget->hasVMLxHazards())
return true;
if (!N->hasOneUse())
return false;
SDNode *Use = *N->use_begin();
if (Use->getOpcode() == ISD::CopyToReg)
return true;
if (Use->isMachineOpcode()) {
const ARMBaseInstrInfo *TII = static_cast<const ARMBaseInstrInfo *>(
CurDAG->getSubtarget().getInstrInfo());
const MCInstrDesc &MCID = TII->get(Use->getMachineOpcode());
if (MCID.mayStore())
return true;
unsigned Opcode = MCID.getOpcode();
if (Opcode == ARM::VMOVRS || Opcode == ARM::VMOVRRD)
return true;
// vmlx feeding into another vmlx. We actually want to unfold
// the use later in the MLxExpansion pass. e.g.
// vmla
// vmla (stall 8 cycles)
//
// vmul (5 cycles)
// vadd (5 cycles)
// vmla
// This adds up to about 18 - 19 cycles.
//
// vmla
// vmul (stall 4 cycles)
// vadd adds up to about 14 cycles.
return TII->isFpMLxInstruction(Opcode);
}
return false;
}
bool ARMDAGToDAGISel::isShifterOpProfitable(const SDValue &Shift,
ARM_AM::ShiftOpc ShOpcVal,
unsigned ShAmt) {
if (!Subtarget->isLikeA9() && !Subtarget->isSwift())
return true;
if (Shift.hasOneUse())
return true;
// R << 2 is free.
return ShOpcVal == ARM_AM::lsl &&
(ShAmt == 2 || (Subtarget->isSwift() && ShAmt == 1));
}
bool ARMDAGToDAGISel::canExtractShiftFromMul(const SDValue &N,
unsigned MaxShift,
unsigned &PowerOfTwo,
SDValue &NewMulConst) const {
assert(N.getOpcode() == ISD::MUL);
assert(MaxShift > 0);
// If the multiply is used in more than one place then changing the constant
// will make other uses incorrect, so don't.
if (!N.hasOneUse()) return false;
// Check if the multiply is by a constant
ConstantSDNode *MulConst = dyn_cast<ConstantSDNode>(N.getOperand(1));
if (!MulConst) return false;
// If the constant is used in more than one place then modifying it will mean
// we need to materialize two constants instead of one, which is a bad idea.
if (!MulConst->hasOneUse()) return false;
unsigned MulConstVal = MulConst->getZExtValue();
if (MulConstVal == 0) return false;
// Find the largest power of 2 that MulConstVal is a multiple of
PowerOfTwo = MaxShift;
while ((MulConstVal % (1 << PowerOfTwo)) != 0) {
--PowerOfTwo;
if (PowerOfTwo == 0) return false;
}
// Only optimise if the new cost is better
unsigned NewMulConstVal = MulConstVal / (1 << PowerOfTwo);
NewMulConst = CurDAG->getConstant(NewMulConstVal, SDLoc(N), MVT::i32);
unsigned OldCost = ConstantMaterializationCost(MulConstVal, Subtarget);
unsigned NewCost = ConstantMaterializationCost(NewMulConstVal, Subtarget);
return NewCost < OldCost;
}
void ARMDAGToDAGISel::replaceDAGValue(const SDValue &N, SDValue M) {
CurDAG->RepositionNode(N.getNode()->getIterator(), M.getNode());
ReplaceUses(N, M);
}
bool ARMDAGToDAGISel::SelectImmShifterOperand(SDValue N,
SDValue &BaseReg,
SDValue &Opc,
bool CheckProfitability) {
if (DisableShifterOp)
return false;
// If N is a multiply-by-constant and it's profitable to extract a shift and
// use it in a shifted operand do so.
if (N.getOpcode() == ISD::MUL) {
unsigned PowerOfTwo = 0;
SDValue NewMulConst;
if (canExtractShiftFromMul(N, 31, PowerOfTwo, NewMulConst)) {
HandleSDNode Handle(N);
SDLoc Loc(N);
replaceDAGValue(N.getOperand(1), NewMulConst);
BaseReg = Handle.getValue();
Opc = CurDAG->getTargetConstant(
ARM_AM::getSORegOpc(ARM_AM::lsl, PowerOfTwo), Loc, MVT::i32);
return true;
}
}
ARM_AM::ShiftOpc ShOpcVal = ARM_AM::getShiftOpcForNode(N.getOpcode());
// Don't match base register only case. That is matched to a separate
// lower complexity pattern with explicit register operand.
if (ShOpcVal == ARM_AM::no_shift) return false;
BaseReg = N.getOperand(0);
unsigned ShImmVal = 0;
ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1));
if (!RHS) return false;
ShImmVal = RHS->getZExtValue() & 31;
Opc = CurDAG->getTargetConstant(ARM_AM::getSORegOpc(ShOpcVal, ShImmVal),
SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectRegShifterOperand(SDValue N,
SDValue &BaseReg,
SDValue &ShReg,
SDValue &Opc,
bool CheckProfitability) {
if (DisableShifterOp)
return false;
ARM_AM::ShiftOpc ShOpcVal = ARM_AM::getShiftOpcForNode(N.getOpcode());
// Don't match base register only case. That is matched to a separate
// lower complexity pattern with explicit register operand.
if (ShOpcVal == ARM_AM::no_shift) return false;
BaseReg = N.getOperand(0);
unsigned ShImmVal = 0;
ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1));
if (RHS) return false;
ShReg = N.getOperand(1);
if (CheckProfitability && !isShifterOpProfitable(N, ShOpcVal, ShImmVal))
return false;
Opc = CurDAG->getTargetConstant(ARM_AM::getSORegOpc(ShOpcVal, ShImmVal),
SDLoc(N), MVT::i32);
return true;
}
// Determine whether an ISD::OR's operands are suitable to turn the operation
// into an addition, which often has more compact encodings.
bool ARMDAGToDAGISel::SelectAddLikeOr(SDNode *Parent, SDValue N, SDValue &Out) {
assert(Parent->getOpcode() == ISD::OR && "unexpected parent");
Out = N;
return CurDAG->haveNoCommonBitsSet(N, Parent->getOperand(1));
}
bool ARMDAGToDAGISel::SelectAddrModeImm12(SDValue N,
SDValue &Base,
SDValue &OffImm) {
// Match simple R + imm12 operands.
// Base only.
if (N.getOpcode() != ISD::ADD && N.getOpcode() != ISD::SUB &&
!CurDAG->isBaseWithConstantOffset(N)) {
if (N.getOpcode() == ISD::FrameIndex) {
// Match frame index.
int FI = cast<FrameIndexSDNode>(N)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (N.getOpcode() == ARMISD::Wrapper &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalAddress &&
N.getOperand(0).getOpcode() != ISD::TargetExternalSymbol &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalTLSAddress) {
Base = N.getOperand(0);
} else
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
int RHSC = (int)RHS->getSExtValue();
if (N.getOpcode() == ISD::SUB)
RHSC = -RHSC;
if (RHSC > -0x1000 && RHSC < 0x1000) { // 12 bits
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
OffImm = CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32);
return true;
}
}
// Base only.
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectLdStSOReg(SDValue N, SDValue &Base, SDValue &Offset,
SDValue &Opc) {
if (N.getOpcode() == ISD::MUL &&
((!Subtarget->isLikeA9() && !Subtarget->isSwift()) || N.hasOneUse())) {
if (ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
// X * [3,5,9] -> X + X * [2,4,8] etc.
int RHSC = (int)RHS->getZExtValue();
if (RHSC & 1) {
RHSC = RHSC & ~1;
ARM_AM::AddrOpc AddSub = ARM_AM::add;
if (RHSC < 0) {
AddSub = ARM_AM::sub;
RHSC = - RHSC;
}
if (isPowerOf2_32(RHSC)) {
unsigned ShAmt = Log2_32(RHSC);
Base = Offset = N.getOperand(0);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM2Opc(AddSub, ShAmt,
ARM_AM::lsl),
SDLoc(N), MVT::i32);
return true;
}
}
}
}
if (N.getOpcode() != ISD::ADD && N.getOpcode() != ISD::SUB &&
// ISD::OR that is equivalent to an ISD::ADD.
!CurDAG->isBaseWithConstantOffset(N))
return false;
// Leave simple R +/- imm12 operands for LDRi12
if (N.getOpcode() == ISD::ADD || N.getOpcode() == ISD::OR) {
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), /*Scale=*/1,
-0x1000+1, 0x1000, RHSC)) // 12 bits.
return false;
}
// Otherwise this is R +/- [possibly shifted] R.
ARM_AM::AddrOpc AddSub = N.getOpcode() == ISD::SUB ? ARM_AM::sub:ARM_AM::add;
ARM_AM::ShiftOpc ShOpcVal =
ARM_AM::getShiftOpcForNode(N.getOperand(1).getOpcode());
unsigned ShAmt = 0;
Base = N.getOperand(0);
Offset = N.getOperand(1);
if (ShOpcVal != ARM_AM::no_shift) {
// Check to see if the RHS of the shift is a constant, if not, we can't fold
// it.
if (ConstantSDNode *Sh =
dyn_cast<ConstantSDNode>(N.getOperand(1).getOperand(1))) {
ShAmt = Sh->getZExtValue();
if (isShifterOpProfitable(Offset, ShOpcVal, ShAmt))
Offset = N.getOperand(1).getOperand(0);
else {
ShAmt = 0;
ShOpcVal = ARM_AM::no_shift;
}
} else {
ShOpcVal = ARM_AM::no_shift;
}
}
// Try matching (R shl C) + (R).
if (N.getOpcode() != ISD::SUB && ShOpcVal == ARM_AM::no_shift &&
!(Subtarget->isLikeA9() || Subtarget->isSwift() ||
N.getOperand(0).hasOneUse())) {
ShOpcVal = ARM_AM::getShiftOpcForNode(N.getOperand(0).getOpcode());
if (ShOpcVal != ARM_AM::no_shift) {
// Check to see if the RHS of the shift is a constant, if not, we can't
// fold it.
if (ConstantSDNode *Sh =
dyn_cast<ConstantSDNode>(N.getOperand(0).getOperand(1))) {
ShAmt = Sh->getZExtValue();
if (isShifterOpProfitable(N.getOperand(0), ShOpcVal, ShAmt)) {
Offset = N.getOperand(0).getOperand(0);
Base = N.getOperand(1);
} else {
ShAmt = 0;
ShOpcVal = ARM_AM::no_shift;
}
} else {
ShOpcVal = ARM_AM::no_shift;
}
}
}
// If Offset is a multiply-by-constant and it's profitable to extract a shift
// and use it in a shifted operand do so.
if (Offset.getOpcode() == ISD::MUL && N.hasOneUse()) {
unsigned PowerOfTwo = 0;
SDValue NewMulConst;
if (canExtractShiftFromMul(Offset, 31, PowerOfTwo, NewMulConst)) {
HandleSDNode Handle(Offset);
replaceDAGValue(Offset.getOperand(1), NewMulConst);
Offset = Handle.getValue();
ShAmt = PowerOfTwo;
ShOpcVal = ARM_AM::lsl;
}
}
Opc = CurDAG->getTargetConstant(ARM_AM::getAM2Opc(AddSub, ShAmt, ShOpcVal),
SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode2OffsetReg(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc) {
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM = (Opcode == ISD::LOAD)
? cast<LoadSDNode>(Op)->getAddressingMode()
: cast<StoreSDNode>(Op)->getAddressingMode();
ARM_AM::AddrOpc AddSub = (AM == ISD::PRE_INC || AM == ISD::POST_INC)
? ARM_AM::add : ARM_AM::sub;
int Val;
if (isScaledConstantInRange(N, /*Scale=*/1, 0, 0x1000, Val))
return false;
Offset = N;
ARM_AM::ShiftOpc ShOpcVal = ARM_AM::getShiftOpcForNode(N.getOpcode());
unsigned ShAmt = 0;
if (ShOpcVal != ARM_AM::no_shift) {
// Check to see if the RHS of the shift is a constant, if not, we can't fold
// it.
if (ConstantSDNode *Sh = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
ShAmt = Sh->getZExtValue();
if (isShifterOpProfitable(N, ShOpcVal, ShAmt))
Offset = N.getOperand(0);
else {
ShAmt = 0;
ShOpcVal = ARM_AM::no_shift;
}
} else {
ShOpcVal = ARM_AM::no_shift;
}
}
Opc = CurDAG->getTargetConstant(ARM_AM::getAM2Opc(AddSub, ShAmt, ShOpcVal),
SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode2OffsetImmPre(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc) {
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM = (Opcode == ISD::LOAD)
? cast<LoadSDNode>(Op)->getAddressingMode()
: cast<StoreSDNode>(Op)->getAddressingMode();
ARM_AM::AddrOpc AddSub = (AM == ISD::PRE_INC || AM == ISD::POST_INC)
? ARM_AM::add : ARM_AM::sub;
int Val;
if (isScaledConstantInRange(N, /*Scale=*/1, 0, 0x1000, Val)) { // 12 bits.
if (AddSub == ARM_AM::sub) Val *= -1;
Offset = CurDAG->getRegister(0, MVT::i32);
Opc = CurDAG->getTargetConstant(Val, SDLoc(Op), MVT::i32);
return true;
}
return false;
}
bool ARMDAGToDAGISel::SelectAddrMode2OffsetImm(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc) {
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM = (Opcode == ISD::LOAD)
? cast<LoadSDNode>(Op)->getAddressingMode()
: cast<StoreSDNode>(Op)->getAddressingMode();
ARM_AM::AddrOpc AddSub = (AM == ISD::PRE_INC || AM == ISD::POST_INC)
? ARM_AM::add : ARM_AM::sub;
int Val;
if (isScaledConstantInRange(N, /*Scale=*/1, 0, 0x1000, Val)) { // 12 bits.
Offset = CurDAG->getRegister(0, MVT::i32);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM2Opc(AddSub, Val,
ARM_AM::no_shift),
SDLoc(Op), MVT::i32);
return true;
}
return false;
}
bool ARMDAGToDAGISel::SelectAddrOffsetNone(SDValue N, SDValue &Base) {
Base = N;
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode3(SDValue N,
SDValue &Base, SDValue &Offset,
SDValue &Opc) {
if (N.getOpcode() == ISD::SUB) {
// X - C is canonicalize to X + -C, no need to handle it here.
Base = N.getOperand(0);
Offset = N.getOperand(1);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(ARM_AM::sub, 0), SDLoc(N),
MVT::i32);
return true;
}
if (!CurDAG->isBaseWithConstantOffset(N)) {
Base = N;
if (N.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(N)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
Offset = CurDAG->getRegister(0, MVT::i32);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(ARM_AM::add, 0), SDLoc(N),
MVT::i32);
return true;
}
// If the RHS is +/- imm8, fold into addr mode.
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), /*Scale=*/1,
-256 + 1, 256, RHSC)) { // 8 bits.
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
Offset = CurDAG->getRegister(0, MVT::i32);
ARM_AM::AddrOpc AddSub = ARM_AM::add;
if (RHSC < 0) {
AddSub = ARM_AM::sub;
RHSC = -RHSC;
}
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(AddSub, RHSC), SDLoc(N),
MVT::i32);
return true;
}
Base = N.getOperand(0);
Offset = N.getOperand(1);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(ARM_AM::add, 0), SDLoc(N),
MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode3Offset(SDNode *Op, SDValue N,
SDValue &Offset, SDValue &Opc) {
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM = (Opcode == ISD::LOAD)
? cast<LoadSDNode>(Op)->getAddressingMode()
: cast<StoreSDNode>(Op)->getAddressingMode();
ARM_AM::AddrOpc AddSub = (AM == ISD::PRE_INC || AM == ISD::POST_INC)
? ARM_AM::add : ARM_AM::sub;
int Val;
if (isScaledConstantInRange(N, /*Scale=*/1, 0, 256, Val)) { // 12 bits.
Offset = CurDAG->getRegister(0, MVT::i32);
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(AddSub, Val), SDLoc(Op),
MVT::i32);
return true;
}
Offset = N;
Opc = CurDAG->getTargetConstant(ARM_AM::getAM3Opc(AddSub, 0), SDLoc(Op),
MVT::i32);
return true;
}
bool ARMDAGToDAGISel::IsAddressingMode5(SDValue N, SDValue &Base, SDValue &Offset,
bool FP16) {
if (!CurDAG->isBaseWithConstantOffset(N)) {
Base = N;
if (N.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(N)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
} else if (N.getOpcode() == ARMISD::Wrapper &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalAddress &&
N.getOperand(0).getOpcode() != ISD::TargetExternalSymbol &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalTLSAddress) {
Base = N.getOperand(0);
}
Offset = CurDAG->getTargetConstant(ARM_AM::getAM5Opc(ARM_AM::add, 0),
SDLoc(N), MVT::i32);
return true;
}
// If the RHS is +/- imm8, fold into addr mode.
int RHSC;
const int Scale = FP16 ? 2 : 4;
if (isScaledConstantInRange(N.getOperand(1), Scale, -255, 256, RHSC)) {
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
ARM_AM::AddrOpc AddSub = ARM_AM::add;
if (RHSC < 0) {
AddSub = ARM_AM::sub;
RHSC = -RHSC;
}
if (FP16)
Offset = CurDAG->getTargetConstant(ARM_AM::getAM5FP16Opc(AddSub, RHSC),
SDLoc(N), MVT::i32);
else
Offset = CurDAG->getTargetConstant(ARM_AM::getAM5Opc(AddSub, RHSC),
SDLoc(N), MVT::i32);
return true;
}
Base = N;
if (FP16)
Offset = CurDAG->getTargetConstant(ARM_AM::getAM5FP16Opc(ARM_AM::add, 0),
SDLoc(N), MVT::i32);
else
Offset = CurDAG->getTargetConstant(ARM_AM::getAM5Opc(ARM_AM::add, 0),
SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode5(SDValue N,
SDValue &Base, SDValue &Offset) {
return IsAddressingMode5(N, Base, Offset, /*FP16=*/ false);
}
bool ARMDAGToDAGISel::SelectAddrMode5FP16(SDValue N,
SDValue &Base, SDValue &Offset) {
return IsAddressingMode5(N, Base, Offset, /*FP16=*/ true);
}
bool ARMDAGToDAGISel::SelectAddrMode6(SDNode *Parent, SDValue N, SDValue &Addr,
SDValue &Align) {
Addr = N;
unsigned Alignment = 0;
MemSDNode *MemN = cast<MemSDNode>(Parent);
if (isa<LSBaseSDNode>(MemN) ||
((MemN->getOpcode() == ARMISD::VST1_UPD ||
MemN->getOpcode() == ARMISD::VLD1_UPD) &&
MemN->getConstantOperandVal(MemN->getNumOperands() - 1) == 1)) {
// This case occurs only for VLD1-lane/dup and VST1-lane instructions.
// The maximum alignment is equal to the memory size being referenced.
unsigned MMOAlign = MemN->getAlignment();
unsigned MemSize = MemN->getMemoryVT().getSizeInBits() / 8;
if (MMOAlign >= MemSize && MemSize > 1)
Alignment = MemSize;
} else {
// All other uses of addrmode6 are for intrinsics. For now just record
// the raw alignment value; it will be refined later based on the legal
// alignment operands for the intrinsic.
Alignment = MemN->getAlignment();
}
Align = CurDAG->getTargetConstant(Alignment, SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectAddrMode6Offset(SDNode *Op, SDValue N,
SDValue &Offset) {
LSBaseSDNode *LdSt = cast<LSBaseSDNode>(Op);
ISD::MemIndexedMode AM = LdSt->getAddressingMode();
if (AM != ISD::POST_INC)
return false;
Offset = N;
if (ConstantSDNode *NC = dyn_cast<ConstantSDNode>(N)) {
if (NC->getZExtValue() * 8 == LdSt->getMemoryVT().getSizeInBits())
Offset = CurDAG->getRegister(0, MVT::i32);
}
return true;
}
bool ARMDAGToDAGISel::SelectAddrModePC(SDValue N,
SDValue &Offset, SDValue &Label) {
if (N.getOpcode() == ARMISD::PIC_ADD && N.hasOneUse()) {
Offset = N.getOperand(0);
SDValue N1 = N.getOperand(1);
Label = CurDAG->getTargetConstant(cast<ConstantSDNode>(N1)->getZExtValue(),
SDLoc(N), MVT::i32);
return true;
}
return false;
}
//===----------------------------------------------------------------------===//
// Thumb Addressing Modes
//===----------------------------------------------------------------------===//
static bool shouldUseZeroOffsetLdSt(SDValue N) {
// Negative numbers are difficult to materialise in thumb1. If we are
// selecting the add of a negative, instead try to select ri with a zero
// offset, so create the add node directly which will become a sub.
if (N.getOpcode() != ISD::ADD)
return false;
// Look for an imm which is not legal for ld/st, but is legal for sub.
if (auto C = dyn_cast<ConstantSDNode>(N.getOperand(1)))
return C->getSExtValue() < 0 && C->getSExtValue() >= -255;
return false;
}
bool ARMDAGToDAGISel::SelectThumbAddrModeRRSext(SDValue N, SDValue &Base,
SDValue &Offset) {
if (N.getOpcode() != ISD::ADD && !CurDAG->isBaseWithConstantOffset(N)) {
ConstantSDNode *NC = dyn_cast<ConstantSDNode>(N);
if (!NC || !NC->isNullValue())
return false;
Base = Offset = N;
return true;
}
Base = N.getOperand(0);
Offset = N.getOperand(1);
return true;
}
bool ARMDAGToDAGISel::SelectThumbAddrModeRR(SDValue N, SDValue &Base,
SDValue &Offset) {
if (shouldUseZeroOffsetLdSt(N))
return false; // Select ri instead
return SelectThumbAddrModeRRSext(N, Base, Offset);
}
bool
ARMDAGToDAGISel::SelectThumbAddrModeImm5S(SDValue N, unsigned Scale,
SDValue &Base, SDValue &OffImm) {
if (shouldUseZeroOffsetLdSt(N)) {
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (!CurDAG->isBaseWithConstantOffset(N)) {
if (N.getOpcode() == ISD::ADD) {
return false; // We want to select register offset instead
} else if (N.getOpcode() == ARMISD::Wrapper &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalAddress &&
N.getOperand(0).getOpcode() != ISD::TargetExternalSymbol &&
N.getOperand(0).getOpcode() != ISD::TargetConstantPool &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalTLSAddress) {
Base = N.getOperand(0);
} else {
Base = N;
}
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
// If the RHS is + imm5 * scale, fold into addr mode.
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), Scale, 0, 32, RHSC)) {
Base = N.getOperand(0);
OffImm = CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32);
return true;
}
// Offset is too large, so use register offset instead.
return false;
}
bool
ARMDAGToDAGISel::SelectThumbAddrModeImm5S4(SDValue N, SDValue &Base,
SDValue &OffImm) {
return SelectThumbAddrModeImm5S(N, 4, Base, OffImm);
}
bool
ARMDAGToDAGISel::SelectThumbAddrModeImm5S2(SDValue N, SDValue &Base,
SDValue &OffImm) {
return SelectThumbAddrModeImm5S(N, 2, Base, OffImm);
}
bool
ARMDAGToDAGISel::SelectThumbAddrModeImm5S1(SDValue N, SDValue &Base,
SDValue &OffImm) {
return SelectThumbAddrModeImm5S(N, 1, Base, OffImm);
}
bool ARMDAGToDAGISel::SelectThumbAddrModeSP(SDValue N,
SDValue &Base, SDValue &OffImm) {
if (N.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(N)->getIndex();
// Only multiples of 4 are allowed for the offset, so the frame object
// alignment must be at least 4.
MachineFrameInfo &MFI = MF->getFrameInfo();
if (MFI.getObjectAlignment(FI) < 4)
MFI.setObjectAlignment(FI, 4);
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (!CurDAG->isBaseWithConstantOffset(N))
return false;
if (N.getOperand(0).getOpcode() == ISD::FrameIndex) {
// If the RHS is + imm8 * scale, fold into addr mode.
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), /*Scale=*/4, 0, 256, RHSC)) {
Base = N.getOperand(0);
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
// Make sure the offset is inside the object, or we might fail to
// allocate an emergency spill slot. (An out-of-range access is UB, but
// it could show up anyway.)
MachineFrameInfo &MFI = MF->getFrameInfo();
if (RHSC * 4 < MFI.getObjectSize(FI)) {
// For LHS+RHS to result in an offset that's a multiple of 4 the object
// indexed by the LHS must be 4-byte aligned.
if (!MFI.isFixedObjectIndex(FI) && MFI.getObjectAlignment(FI) < 4)
MFI.setObjectAlignment(FI, 4);
if (MFI.getObjectAlignment(FI) >= 4) {
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
OffImm = CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32);
return true;
}
}
}
}
return false;
}
template <unsigned Shift>
bool ARMDAGToDAGISel::SelectTAddrModeImm7(SDValue N, SDValue &Base,
SDValue &OffImm) {
if (N.getOpcode() == ISD::SUB || CurDAG->isBaseWithConstantOffset(N)) {
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), 1 << Shift, -0x7f, 0x80,
RHSC)) {
Base = N.getOperand(0);
if (N.getOpcode() == ISD::SUB)
RHSC = -RHSC;
OffImm =
CurDAG->getTargetConstant(RHSC * (1 << Shift), SDLoc(N), MVT::i32);
return true;
}
}
// Base only.
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
//===----------------------------------------------------------------------===//
// Thumb 2 Addressing Modes
//===----------------------------------------------------------------------===//
bool ARMDAGToDAGISel::SelectT2AddrModeImm12(SDValue N,
SDValue &Base, SDValue &OffImm) {
// Match simple R + imm12 operands.
// Base only.
if (N.getOpcode() != ISD::ADD && N.getOpcode() != ISD::SUB &&
!CurDAG->isBaseWithConstantOffset(N)) {
if (N.getOpcode() == ISD::FrameIndex) {
// Match frame index.
int FI = cast<FrameIndexSDNode>(N)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (N.getOpcode() == ARMISD::Wrapper &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalAddress &&
N.getOperand(0).getOpcode() != ISD::TargetExternalSymbol &&
N.getOperand(0).getOpcode() != ISD::TargetGlobalTLSAddress) {
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::TargetConstantPool)
return false; // We want to select t2LDRpci instead.
} else
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
if (ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
if (SelectT2AddrModeImm8(N, Base, OffImm))
// Let t2LDRi8 handle (R - imm8).
return false;
int RHSC = (int)RHS->getZExtValue();
if (N.getOpcode() == ISD::SUB)
RHSC = -RHSC;
if (RHSC >= 0 && RHSC < 0x1000) { // 12 bits (unsigned)
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
OffImm = CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32);
return true;
}
}
// Base only.
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
+template <unsigned Shift>
+bool ARMDAGToDAGISel::SelectT2AddrModeImm8(SDValue N, SDValue &Base,
+ SDValue &OffImm) {
+ if (N.getOpcode() == ISD::SUB || CurDAG->isBaseWithConstantOffset(N)) {
+ int RHSC;
+ if (isScaledConstantInRange(N.getOperand(1), 1 << Shift, -255, 256, RHSC)) {
+ Base = N.getOperand(0);
+ if (Base.getOpcode() == ISD::FrameIndex) {
+ int FI = cast<FrameIndexSDNode>(Base)->getIndex();
+ Base = CurDAG->getTargetFrameIndex(
+ FI, TLI->getPointerTy(CurDAG->getDataLayout()));
+ }
+
+ if (N.getOpcode() == ISD::SUB)
+ RHSC = -RHSC;
+ OffImm =
+ CurDAG->getTargetConstant(RHSC * (1 << Shift), SDLoc(N), MVT::i32);
+ return true;
+ }
+ }
+
+ // Base only.
+ Base = N;
+ OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
+ return true;
+}
+
bool ARMDAGToDAGISel::SelectT2AddrModeImm8(SDValue N,
SDValue &Base, SDValue &OffImm) {
// Match simple R - imm8 operands.
if (N.getOpcode() != ISD::ADD && N.getOpcode() != ISD::SUB &&
!CurDAG->isBaseWithConstantOffset(N))
return false;
if (ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
int RHSC = (int)RHS->getSExtValue();
if (N.getOpcode() == ISD::SUB)
RHSC = -RHSC;
if ((RHSC >= -255) && (RHSC < 0)) { // 8 bits (always negative)
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
OffImm = CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32);
return true;
}
}
return false;
}
bool ARMDAGToDAGISel::SelectT2AddrModeImm8Offset(SDNode *Op, SDValue N,
SDValue &OffImm){
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM = (Opcode == ISD::LOAD)
? cast<LoadSDNode>(Op)->getAddressingMode()
: cast<StoreSDNode>(Op)->getAddressingMode();
int RHSC;
if (isScaledConstantInRange(N, /*Scale=*/1, 0, 0x100, RHSC)) { // 8 bits.
OffImm = ((AM == ISD::PRE_INC) || (AM == ISD::POST_INC))
? CurDAG->getTargetConstant(RHSC, SDLoc(N), MVT::i32)
: CurDAG->getTargetConstant(-RHSC, SDLoc(N), MVT::i32);
return true;
}
return false;
}
template <unsigned Shift>
bool ARMDAGToDAGISel::SelectT2AddrModeImm7(SDValue N, SDValue &Base,
SDValue &OffImm) {
if (N.getOpcode() == ISD::SUB || CurDAG->isBaseWithConstantOffset(N)) {
int RHSC;
if (isScaledConstantInRange(N.getOperand(1), 1 << Shift, -0x7f, 0x80,
RHSC)) {
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
if (N.getOpcode() == ISD::SUB)
RHSC = -RHSC;
OffImm =
CurDAG->getTargetConstant(RHSC * (1 << Shift), SDLoc(N), MVT::i32);
return true;
}
}
// Base only.
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
return true;
}
template <unsigned Shift>
bool ARMDAGToDAGISel::SelectT2AddrModeImm7Offset(SDNode *Op, SDValue N,
SDValue &OffImm) {
return SelectT2AddrModeImm7Offset(Op, N, OffImm, Shift);
}
bool ARMDAGToDAGISel::SelectT2AddrModeImm7Offset(SDNode *Op, SDValue N,
SDValue &OffImm,
unsigned Shift) {
unsigned Opcode = Op->getOpcode();
ISD::MemIndexedMode AM;
switch (Opcode) {
case ISD::LOAD:
AM = cast<LoadSDNode>(Op)->getAddressingMode();
break;
case ISD::STORE:
AM = cast<StoreSDNode>(Op)->getAddressingMode();
break;
case ISD::MLOAD:
AM = cast<MaskedLoadSDNode>(Op)->getAddressingMode();
break;
case ISD::MSTORE:
AM = cast<MaskedStoreSDNode>(Op)->getAddressingMode();
break;
default:
llvm_unreachable("Unexpected Opcode for Imm7Offset");
}
int RHSC;
// 7 bit constant, shifted by Shift.
if (isScaledConstantInRange(N, 1 << Shift, 0, 0x80, RHSC)) {
OffImm =
((AM == ISD::PRE_INC) || (AM == ISD::POST_INC))
? CurDAG->getTargetConstant(RHSC * (1 << Shift), SDLoc(N), MVT::i32)
: CurDAG->getTargetConstant(-RHSC * (1 << Shift), SDLoc(N),
MVT::i32);
return true;
}
return false;
}
template <int Min, int Max>
bool ARMDAGToDAGISel::SelectImmediateInRange(SDValue N, SDValue &OffImm) {
int Val;
if (isScaledConstantInRange(N, 1, Min, Max, Val)) {
OffImm = CurDAG->getTargetConstant(Val, SDLoc(N), MVT::i32);
return true;
}
return false;
}
bool ARMDAGToDAGISel::SelectT2AddrModeSoReg(SDValue N,
SDValue &Base,
SDValue &OffReg, SDValue &ShImm) {
// (R - imm8) should be handled by t2LDRi8. The rest are handled by t2LDRi12.
if (N.getOpcode() != ISD::ADD && !CurDAG->isBaseWithConstantOffset(N))
return false;
// Leave (R + imm12) for t2LDRi12, (R - imm8) for t2LDRi8.
if (ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
int RHSC = (int)RHS->getZExtValue();
if (RHSC >= 0 && RHSC < 0x1000) // 12 bits (unsigned)
return false;
else if (RHSC < 0 && RHSC >= -255) // 8 bits
return false;
}
// Look for (R + R) or (R + (R << [1,2,3])).
unsigned ShAmt = 0;
Base = N.getOperand(0);
OffReg = N.getOperand(1);
// Swap if it is ((R << c) + R).
ARM_AM::ShiftOpc ShOpcVal = ARM_AM::getShiftOpcForNode(OffReg.getOpcode());
if (ShOpcVal != ARM_AM::lsl) {
ShOpcVal = ARM_AM::getShiftOpcForNode(Base.getOpcode());
if (ShOpcVal == ARM_AM::lsl)
std::swap(Base, OffReg);
}
if (ShOpcVal == ARM_AM::lsl) {
// Check to see if the RHS of the shift is a constant, if not, we can't fold
// it.
if (ConstantSDNode *Sh = dyn_cast<ConstantSDNode>(OffReg.getOperand(1))) {
ShAmt = Sh->getZExtValue();
if (ShAmt < 4 && isShifterOpProfitable(OffReg, ShOpcVal, ShAmt))
OffReg = OffReg.getOperand(0);
else {
ShAmt = 0;
}
}
}
// If OffReg is a multiply-by-constant and it's profitable to extract a shift
// and use it in a shifted operand do so.
if (OffReg.getOpcode() == ISD::MUL && N.hasOneUse()) {
unsigned PowerOfTwo = 0;
SDValue NewMulConst;
if (canExtractShiftFromMul(OffReg, 3, PowerOfTwo, NewMulConst)) {
HandleSDNode Handle(OffReg);
replaceDAGValue(OffReg.getOperand(1), NewMulConst);
OffReg = Handle.getValue();
ShAmt = PowerOfTwo;
}
}
ShImm = CurDAG->getTargetConstant(ShAmt, SDLoc(N), MVT::i32);
return true;
}
bool ARMDAGToDAGISel::SelectT2AddrModeExclusive(SDValue N, SDValue &Base,
SDValue &OffImm) {
// This *must* succeed since it's used for the irreplaceable ldrex and strex
// instructions.
Base = N;
OffImm = CurDAG->getTargetConstant(0, SDLoc(N), MVT::i32);
if (N.getOpcode() != ISD::ADD || !CurDAG->isBaseWithConstantOffset(N))
return true;
ConstantSDNode *RHS = dyn_cast<ConstantSDNode>(N.getOperand(1));
if (!RHS)
return true;
uint32_t RHSC = (int)RHS->getZExtValue();
if (RHSC > 1020 || RHSC % 4 != 0)
return true;
Base = N.getOperand(0);
if (Base.getOpcode() == ISD::FrameIndex) {
int FI = cast<FrameIndexSDNode>(Base)->getIndex();
Base = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
}
OffImm = CurDAG->getTargetConstant(RHSC/4, SDLoc(N), MVT::i32);
return true;
}
//===--------------------------------------------------------------------===//
/// getAL - Returns a ARMCC::AL immediate node.
static inline SDValue getAL(SelectionDAG *CurDAG, const SDLoc &dl) {
return CurDAG->getTargetConstant((uint64_t)ARMCC::AL, dl, MVT::i32);
}
void ARMDAGToDAGISel::transferMemOperands(SDNode *N, SDNode *Result) {
MachineMemOperand *MemOp = cast<MemSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(Result), {MemOp});
}
bool ARMDAGToDAGISel::tryARMIndexedLoad(SDNode *N) {
LoadSDNode *LD = cast<LoadSDNode>(N);
ISD::MemIndexedMode AM = LD->getAddressingMode();
if (AM == ISD::UNINDEXED)
return false;
EVT LoadedVT = LD->getMemoryVT();
SDValue Offset, AMOpc;
bool isPre = (AM == ISD::PRE_INC) || (AM == ISD::PRE_DEC);
unsigned Opcode = 0;
bool Match = false;
if (LoadedVT == MVT::i32 && isPre &&
SelectAddrMode2OffsetImmPre(N, LD->getOffset(), Offset, AMOpc)) {
Opcode = ARM::LDR_PRE_IMM;
Match = true;
} else if (LoadedVT == MVT::i32 && !isPre &&
SelectAddrMode2OffsetImm(N, LD->getOffset(), Offset, AMOpc)) {
Opcode = ARM::LDR_POST_IMM;
Match = true;
} else if (LoadedVT == MVT::i32 &&
SelectAddrMode2OffsetReg(N, LD->getOffset(), Offset, AMOpc)) {
Opcode = isPre ? ARM::LDR_PRE_REG : ARM::LDR_POST_REG;
Match = true;
} else if (LoadedVT == MVT::i16 &&
SelectAddrMode3Offset(N, LD->getOffset(), Offset, AMOpc)) {
Match = true;
Opcode = (LD->getExtensionType() == ISD::SEXTLOAD)
? (isPre ? ARM::LDRSH_PRE : ARM::LDRSH_POST)
: (isPre ? ARM::LDRH_PRE : ARM::LDRH_POST);
} else if (LoadedVT == MVT::i8 || LoadedVT == MVT::i1) {
if (LD->getExtensionType() == ISD::SEXTLOAD) {
if (SelectAddrMode3Offset(N, LD->getOffset(), Offset, AMOpc)) {
Match = true;
Opcode = isPre ? ARM::LDRSB_PRE : ARM::LDRSB_POST;
}
} else {
if (isPre &&
SelectAddrMode2OffsetImmPre(N, LD->getOffset(), Offset, AMOpc)) {
Match = true;
Opcode = ARM::LDRB_PRE_IMM;
} else if (!isPre &&
SelectAddrMode2OffsetImm(N, LD->getOffset(), Offset, AMOpc)) {
Match = true;
Opcode = ARM::LDRB_POST_IMM;
} else if (SelectAddrMode2OffsetReg(N, LD->getOffset(), Offset, AMOpc)) {
Match = true;
Opcode = isPre ? ARM::LDRB_PRE_REG : ARM::LDRB_POST_REG;
}
}
}
if (Match) {
if (Opcode == ARM::LDR_PRE_IMM || Opcode == ARM::LDRB_PRE_IMM) {
SDValue Chain = LD->getChain();
SDValue Base = LD->getBasePtr();
SDValue Ops[]= { Base, AMOpc, getAL(CurDAG, SDLoc(N)),
CurDAG->getRegister(0, MVT::i32), Chain };
SDNode *New = CurDAG->getMachineNode(Opcode, SDLoc(N), MVT::i32, MVT::i32,
MVT::Other, Ops);
transferMemOperands(N, New);
ReplaceNode(N, New);
return true;
} else {
SDValue Chain = LD->getChain();
SDValue Base = LD->getBasePtr();
SDValue Ops[]= { Base, Offset, AMOpc, getAL(CurDAG, SDLoc(N)),
CurDAG->getRegister(0, MVT::i32), Chain };
SDNode *New = CurDAG->getMachineNode(Opcode, SDLoc(N), MVT::i32, MVT::i32,
MVT::Other, Ops);
transferMemOperands(N, New);
ReplaceNode(N, New);
return true;
}
}
return false;
}
bool ARMDAGToDAGISel::tryT1IndexedLoad(SDNode *N) {
LoadSDNode *LD = cast<LoadSDNode>(N);
EVT LoadedVT = LD->getMemoryVT();
ISD::MemIndexedMode AM = LD->getAddressingMode();
if (AM != ISD::POST_INC || LD->getExtensionType() != ISD::NON_EXTLOAD ||
LoadedVT.getSimpleVT().SimpleTy != MVT::i32)
return false;
auto *COffs = dyn_cast<ConstantSDNode>(LD->getOffset());
if (!COffs || COffs->getZExtValue() != 4)
return false;
// A T1 post-indexed load is just a single register LDM: LDM r0!, {r1}.
// The encoding of LDM is not how the rest of ISel expects a post-inc load to
// look however, so we use a pseudo here and switch it for a tLDMIA_UPD after
// ISel.
SDValue Chain = LD->getChain();
SDValue Base = LD->getBasePtr();
SDValue Ops[]= { Base, getAL(CurDAG, SDLoc(N)),
CurDAG->getRegister(0, MVT::i32), Chain };
SDNode *New = CurDAG->getMachineNode(ARM::tLDR_postidx, SDLoc(N), MVT::i32,
MVT::i32, MVT::Other, Ops);
transferMemOperands(N, New);
ReplaceNode(N, New);
return true;
}
bool ARMDAGToDAGISel::tryT2IndexedLoad(SDNode *N) {
LoadSDNode *LD = cast<LoadSDNode>(N);
ISD::MemIndexedMode AM = LD->getAddressingMode();
if (AM == ISD::UNINDEXED)
return false;
EVT LoadedVT = LD->getMemoryVT();
bool isSExtLd = LD->getExtensionType() == ISD::SEXTLOAD;
SDValue Offset;
bool isPre = (AM == ISD::PRE_INC) || (AM == ISD::PRE_DEC);
unsigned Opcode = 0;
bool Match = false;
if (SelectT2AddrModeImm8Offset(N, LD->getOffset(), Offset)) {
switch (LoadedVT.getSimpleVT().SimpleTy) {
case MVT::i32:
Opcode = isPre ? ARM::t2LDR_PRE : ARM::t2LDR_POST;
break;
case MVT::i16:
if (isSExtLd)
Opcode = isPre ? ARM::t2LDRSH_PRE : ARM::t2LDRSH_POST;
else
Opcode = isPre ? ARM::t2LDRH_PRE : ARM::t2LDRH_POST;
break;
case MVT::i8:
case MVT::i1:
if (isSExtLd)
Opcode = isPre ? ARM::t2LDRSB_PRE : ARM::t2LDRSB_POST;
else
Opcode = isPre ? ARM::t2LDRB_PRE : ARM::t2LDRB_POST;
break;
default:
return false;
}
Match = true;
}
if (Match) {
SDValue Chain = LD->getChain();
SDValue Base = LD->getBasePtr();
SDValue Ops[]= { Base, Offset, getAL(CurDAG, SDLoc(N)),
CurDAG->getRegister(0, MVT::i32), Chain };
SDNode *New = CurDAG->getMachineNode(Opcode, SDLoc(N), MVT::i32, MVT::i32,
MVT::Other, Ops);
transferMemOperands(N, New);
ReplaceNode(N, New);
return true;
}
return false;
}
bool ARMDAGToDAGISel::tryMVEIndexedLoad(SDNode *N) {
EVT LoadedVT;
unsigned Opcode = 0;
bool isSExtLd, isPre;
unsigned Align;
ARMVCC::VPTCodes Pred;
SDValue PredReg;
SDValue Chain, Base, Offset;
if (LoadSDNode *LD = dyn_cast<LoadSDNode>(N)) {
ISD::MemIndexedMode AM = LD->getAddressingMode();
if (AM == ISD::UNINDEXED)
return false;
LoadedVT = LD->getMemoryVT();
if (!LoadedVT.isVector())
return false;
Chain = LD->getChain();
Base = LD->getBasePtr();
Offset = LD->getOffset();
Align = LD->getAlignment();
isSExtLd = LD->getExtensionType() == ISD::SEXTLOAD;
isPre = (AM == ISD::PRE_INC) || (AM == ISD::PRE_DEC);
Pred = ARMVCC::None;
PredReg = CurDAG->getRegister(0, MVT::i32);
} else if (MaskedLoadSDNode *LD = dyn_cast<MaskedLoadSDNode>(N)) {
ISD::MemIndexedMode AM = LD->getAddressingMode();
if (AM == ISD::UNINDEXED)
return false;
LoadedVT = LD->getMemoryVT();
if (!LoadedVT.isVector())
return false;
Chain = LD->getChain();
Base = LD->getBasePtr();
Offset = LD->getOffset();
Align = LD->getAlignment();
isSExtLd = LD->getExtensionType() == ISD::SEXTLOAD;
isPre = (AM == ISD::PRE_INC) || (AM == ISD::PRE_DEC);
Pred = ARMVCC::Then;
PredReg = LD->getMask();
} else
llvm_unreachable("Expected a Load or a Masked Load!");
// We allow LE non-masked loads to change the type (for example use a vldrb.8
// as opposed to a vldrw.32). This can allow extra addressing modes or
// alignments for what is otherwise an equivalent instruction.
bool CanChangeType = Subtarget->isLittle() && !isa<MaskedLoadSDNode>(N);
SDValue NewOffset;
if (Align >= 2 && LoadedVT == MVT::v4i16 &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 1)) {
if (isSExtLd)
Opcode = isPre ? ARM::MVE_VLDRHS32_pre : ARM::MVE_VLDRHS32_post;
else
Opcode = isPre ? ARM::MVE_VLDRHU32_pre : ARM::MVE_VLDRHU32_post;
} else if (LoadedVT == MVT::v8i8 &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 0)) {
if (isSExtLd)
Opcode = isPre ? ARM::MVE_VLDRBS16_pre : ARM::MVE_VLDRBS16_post;
else
Opcode = isPre ? ARM::MVE_VLDRBU16_pre : ARM::MVE_VLDRBU16_post;
} else if (LoadedVT == MVT::v4i8 &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 0)) {
if (isSExtLd)
Opcode = isPre ? ARM::MVE_VLDRBS32_pre : ARM::MVE_VLDRBS32_post;
else
Opcode = isPre ? ARM::MVE_VLDRBU32_pre : ARM::MVE_VLDRBU32_post;
} else if (Align >= 4 &&
(CanChangeType || LoadedVT == MVT::v4i32 ||
LoadedVT == MVT::v4f32) &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 2))
Opcode = isPre ? ARM::MVE_VLDRWU32_pre : ARM::MVE_VLDRWU32_post;
else if (Align >= 2 &&
(CanChangeType || LoadedVT == MVT::v8i16 ||
LoadedVT == MVT::v8f16) &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 1))
Opcode = isPre ? ARM::MVE_VLDRHU16_pre : ARM::MVE_VLDRHU16_post;
else if ((CanChangeType || LoadedVT == MVT::v16i8) &&
SelectT2AddrModeImm7Offset(N, Offset, NewOffset, 0))
Opcode = isPre ? ARM::MVE_VLDRBU8_pre : ARM::MVE_VLDRBU8_post;
else
return false;
SDValue Ops[] = {Base, NewOffset,
CurDAG->getTargetConstant(Pred, SDLoc(N), MVT::i32), PredReg,
Chain};
SDNode *New = CurDAG->getMachineNode(Opcode, SDLoc(N), N->getValueType(0),
MVT::i32, MVT::Other, Ops);
transferMemOperands(N, New);
ReplaceUses(SDValue(N, 0), SDValue(New, 1));
ReplaceUses(SDValue(N, 1), SDValue(New, 0));
ReplaceUses(SDValue(N, 2), SDValue(New, 2));
CurDAG->RemoveDeadNode(N);
return true;
}
/// Form a GPRPair pseudo register from a pair of GPR regs.
SDNode *ARMDAGToDAGISel::createGPRPairNode(EVT VT, SDValue V0, SDValue V1) {
SDLoc dl(V0.getNode());
SDValue RegClass =
CurDAG->getTargetConstant(ARM::GPRPairRegClassID, dl, MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::gsub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::gsub_1, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form a D register from a pair of S registers.
SDNode *ARMDAGToDAGISel::createSRegPairNode(EVT VT, SDValue V0, SDValue V1) {
SDLoc dl(V0.getNode());
SDValue RegClass =
CurDAG->getTargetConstant(ARM::DPR_VFP2RegClassID, dl, MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::ssub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::ssub_1, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form a quad register from a pair of D registers.
SDNode *ARMDAGToDAGISel::createDRegPairNode(EVT VT, SDValue V0, SDValue V1) {
SDLoc dl(V0.getNode());
SDValue RegClass = CurDAG->getTargetConstant(ARM::QPRRegClassID, dl,
MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::dsub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::dsub_1, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form 4 consecutive D registers from a pair of Q registers.
SDNode *ARMDAGToDAGISel::createQRegPairNode(EVT VT, SDValue V0, SDValue V1) {
SDLoc dl(V0.getNode());
SDValue RegClass = CurDAG->getTargetConstant(ARM::QQPRRegClassID, dl,
MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::qsub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::qsub_1, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form 4 consecutive S registers.
SDNode *ARMDAGToDAGISel::createQuadSRegsNode(EVT VT, SDValue V0, SDValue V1,
SDValue V2, SDValue V3) {
SDLoc dl(V0.getNode());
SDValue RegClass =
CurDAG->getTargetConstant(ARM::QPR_VFP2RegClassID, dl, MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::ssub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::ssub_1, dl, MVT::i32);
SDValue SubReg2 = CurDAG->getTargetConstant(ARM::ssub_2, dl, MVT::i32);
SDValue SubReg3 = CurDAG->getTargetConstant(ARM::ssub_3, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1,
V2, SubReg2, V3, SubReg3 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form 4 consecutive D registers.
SDNode *ARMDAGToDAGISel::createQuadDRegsNode(EVT VT, SDValue V0, SDValue V1,
SDValue V2, SDValue V3) {
SDLoc dl(V0.getNode());
SDValue RegClass = CurDAG->getTargetConstant(ARM::QQPRRegClassID, dl,
MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::dsub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::dsub_1, dl, MVT::i32);
SDValue SubReg2 = CurDAG->getTargetConstant(ARM::dsub_2, dl, MVT::i32);
SDValue SubReg3 = CurDAG->getTargetConstant(ARM::dsub_3, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1,
V2, SubReg2, V3, SubReg3 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// Form 4 consecutive Q registers.
SDNode *ARMDAGToDAGISel::createQuadQRegsNode(EVT VT, SDValue V0, SDValue V1,
SDValue V2, SDValue V3) {
SDLoc dl(V0.getNode());
SDValue RegClass = CurDAG->getTargetConstant(ARM::QQQQPRRegClassID, dl,
MVT::i32);
SDValue SubReg0 = CurDAG->getTargetConstant(ARM::qsub_0, dl, MVT::i32);
SDValue SubReg1 = CurDAG->getTargetConstant(ARM::qsub_1, dl, MVT::i32);
SDValue SubReg2 = CurDAG->getTargetConstant(ARM::qsub_2, dl, MVT::i32);
SDValue SubReg3 = CurDAG->getTargetConstant(ARM::qsub_3, dl, MVT::i32);
const SDValue Ops[] = { RegClass, V0, SubReg0, V1, SubReg1,
V2, SubReg2, V3, SubReg3 };
return CurDAG->getMachineNode(TargetOpcode::REG_SEQUENCE, dl, VT, Ops);
}
/// GetVLDSTAlign - Get the alignment (in bytes) for the alignment operand
/// of a NEON VLD or VST instruction. The supported values depend on the
/// number of registers being loaded.
SDValue ARMDAGToDAGISel::GetVLDSTAlign(SDValue Align, const SDLoc &dl,
unsigned NumVecs, bool is64BitVector) {
unsigned NumRegs = NumVecs;
if (!is64BitVector && NumVecs < 3)
NumRegs *= 2;
unsigned Alignment = cast<ConstantSDNode>(Align)->getZExtValue();
if (Alignment >= 32 && NumRegs == 4)
Alignment = 32;
else if (Alignment >= 16 && (NumRegs == 2 || NumRegs == 4))
Alignment = 16;
else if (Alignment >= 8)
Alignment = 8;
else
Alignment = 0;
return CurDAG->getTargetConstant(Alignment, dl, MVT::i32);
}
static bool isVLDfixed(unsigned Opc)
{
switch (Opc) {
default: return false;
case ARM::VLD1d8wb_fixed : return true;
case ARM::VLD1d16wb_fixed : return true;
case ARM::VLD1d64Qwb_fixed : return true;
case ARM::VLD1d32wb_fixed : return true;
case ARM::VLD1d64wb_fixed : return true;
case ARM::VLD1d64TPseudoWB_fixed : return true;
case ARM::VLD1d64QPseudoWB_fixed : return true;
case ARM::VLD1q8wb_fixed : return true;
case ARM::VLD1q16wb_fixed : return true;
case ARM::VLD1q32wb_fixed : return true;
case ARM::VLD1q64wb_fixed : return true;
case ARM::VLD1DUPd8wb_fixed : return true;
case ARM::VLD1DUPd16wb_fixed : return true;
case ARM::VLD1DUPd32wb_fixed : return true;
case ARM::VLD1DUPq8wb_fixed : return true;
case ARM::VLD1DUPq16wb_fixed : return true;
case ARM::VLD1DUPq32wb_fixed : return true;
case ARM::VLD2d8wb_fixed : return true;
case ARM::VLD2d16wb_fixed : return true;
case ARM::VLD2d32wb_fixed : return true;
case ARM::VLD2q8PseudoWB_fixed : return true;
case ARM::VLD2q16PseudoWB_fixed : return true;
case ARM::VLD2q32PseudoWB_fixed : return true;
case ARM::VLD2DUPd8wb_fixed : return true;
case ARM::VLD2DUPd16wb_fixed : return true;
case ARM::VLD2DUPd32wb_fixed : return true;
}
}
static bool isVSTfixed(unsigned Opc)
{
switch (Opc) {
default: return false;
case ARM::VST1d8wb_fixed : return true;
case ARM::VST1d16wb_fixed : return true;
case ARM::VST1d32wb_fixed : return true;
case ARM::VST1d64wb_fixed : return true;
case ARM::VST1q8wb_fixed : return true;
case ARM::VST1q16wb_fixed : return true;
case ARM::VST1q32wb_fixed : return true;
case ARM::VST1q64wb_fixed : return true;
case ARM::VST1d64TPseudoWB_fixed : return true;
case ARM::VST1d64QPseudoWB_fixed : return true;
case ARM::VST2d8wb_fixed : return true;
case ARM::VST2d16wb_fixed : return true;
case ARM::VST2d32wb_fixed : return true;
case ARM::VST2q8PseudoWB_fixed : return true;
case ARM::VST2q16PseudoWB_fixed : return true;
case ARM::VST2q32PseudoWB_fixed : return true;
}
}
// Get the register stride update opcode of a VLD/VST instruction that
// is otherwise equivalent to the given fixed stride updating instruction.
static unsigned getVLDSTRegisterUpdateOpcode(unsigned Opc) {
assert((isVLDfixed(Opc) || isVSTfixed(Opc))
&& "Incorrect fixed stride updating instruction.");
switch (Opc) {
default: break;
case ARM::VLD1d8wb_fixed: return ARM::VLD1d8wb_register;
case ARM::VLD1d16wb_fixed: return ARM::VLD1d16wb_register;
case ARM::VLD1d32wb_fixed: return ARM::VLD1d32wb_register;
case ARM::VLD1d64wb_fixed: return ARM::VLD1d64wb_register;
case ARM::VLD1q8wb_fixed: return ARM::VLD1q8wb_register;
case ARM::VLD1q16wb_fixed: return ARM::VLD1q16wb_register;
case ARM::VLD1q32wb_fixed: return ARM::VLD1q32wb_register;
case ARM::VLD1q64wb_fixed: return ARM::VLD1q64wb_register;
case ARM::VLD1d64Twb_fixed: return ARM::VLD1d64Twb_register;
case ARM::VLD1d64Qwb_fixed: return ARM::VLD1d64Qwb_register;
case ARM::VLD1d64TPseudoWB_fixed: return ARM::VLD1d64TPseudoWB_register;
case ARM::VLD1d64QPseudoWB_fixed: return ARM::VLD1d64QPseudoWB_register;
case ARM::VLD1DUPd8wb_fixed : return ARM::VLD1DUPd8wb_register;
case ARM::VLD1DUPd16wb_fixed : return ARM::VLD1DUPd16wb_register;
case ARM::VLD1DUPd32wb_fixed : return ARM::VLD1DUPd32wb_register;
case ARM::VLD1DUPq8wb_fixed : return ARM::VLD1DUPq8wb_register;
case ARM::VLD1DUPq16wb_fixed : return ARM::VLD1DUPq16wb_register;
case ARM::VLD1DUPq32wb_fixed : return ARM::VLD1DUPq32wb_register;
case ARM::VST1d8wb_fixed: return ARM::VST1d8wb_register;
case ARM::VST1d16wb_fixed: return ARM::VST1d16wb_register;
case ARM::VST1d32wb_fixed: return ARM::VST1d32wb_register;
case ARM::VST1d64wb_fixed: return ARM::VST1d64wb_register;
case ARM::VST1q8wb_fixed: return ARM::VST1q8wb_register;
case ARM::VST1q16wb_fixed: return ARM::VST1q16wb_register;
case ARM::VST1q32wb_fixed: return ARM::VST1q32wb_register;
case ARM::VST1q64wb_fixed: return ARM::VST1q64wb_register;
case ARM::VST1d64TPseudoWB_fixed: return ARM::VST1d64TPseudoWB_register;
case ARM::VST1d64QPseudoWB_fixed: return ARM::VST1d64QPseudoWB_register;
case ARM::VLD2d8wb_fixed: return ARM::VLD2d8wb_register;
case ARM::VLD2d16wb_fixed: return ARM::VLD2d16wb_register;
case ARM::VLD2d32wb_fixed: return ARM::VLD2d32wb_register;
case ARM::VLD2q8PseudoWB_fixed: return ARM::VLD2q8PseudoWB_register;
case ARM::VLD2q16PseudoWB_fixed: return ARM::VLD2q16PseudoWB_register;
case ARM::VLD2q32PseudoWB_fixed: return ARM::VLD2q32PseudoWB_register;
case ARM::VST2d8wb_fixed: return ARM::VST2d8wb_register;
case ARM::VST2d16wb_fixed: return ARM::VST2d16wb_register;
case ARM::VST2d32wb_fixed: return ARM::VST2d32wb_register;
case ARM::VST2q8PseudoWB_fixed: return ARM::VST2q8PseudoWB_register;
case ARM::VST2q16PseudoWB_fixed: return ARM::VST2q16PseudoWB_register;
case ARM::VST2q32PseudoWB_fixed: return ARM::VST2q32PseudoWB_register;
case ARM::VLD2DUPd8wb_fixed: return ARM::VLD2DUPd8wb_register;
case ARM::VLD2DUPd16wb_fixed: return ARM::VLD2DUPd16wb_register;
case ARM::VLD2DUPd32wb_fixed: return ARM::VLD2DUPd32wb_register;
}
return Opc; // If not one we handle, return it unchanged.
}
/// Returns true if the given increment is a Constant known to be equal to the
/// access size performed by a NEON load/store. This means the "[rN]!" form can
/// be used.
static bool isPerfectIncrement(SDValue Inc, EVT VecTy, unsigned NumVecs) {
auto C = dyn_cast<ConstantSDNode>(Inc);
return C && C->getZExtValue() == VecTy.getSizeInBits() / 8 * NumVecs;
}
void ARMDAGToDAGISel::SelectVLD(SDNode *N, bool isUpdating, unsigned NumVecs,
const uint16_t *DOpcodes,
const uint16_t *QOpcodes0,
const uint16_t *QOpcodes1) {
assert(NumVecs >= 1 && NumVecs <= 4 && "VLD NumVecs out-of-range");
SDLoc dl(N);
SDValue MemAddr, Align;
bool IsIntrinsic = !isUpdating; // By coincidence, all supported updating
// nodes are not intrinsics.
unsigned AddrOpIdx = IsIntrinsic ? 2 : 1;
if (!SelectAddrMode6(N, N->getOperand(AddrOpIdx), MemAddr, Align))
return;
SDValue Chain = N->getOperand(0);
EVT VT = N->getValueType(0);
bool is64BitVector = VT.is64BitVector();
Align = GetVLDSTAlign(Align, dl, NumVecs, is64BitVector);
unsigned OpcodeIndex;
switch (VT.getSimpleVT().SimpleTy) {
default: llvm_unreachable("unhandled vld type");
// Double-register operations:
case MVT::v8i8: OpcodeIndex = 0; break;
case MVT::v4f16:
case MVT::v4i16: OpcodeIndex = 1; break;
case MVT::v2f32:
case MVT::v2i32: OpcodeIndex = 2; break;
case MVT::v1i64: OpcodeIndex = 3; break;
// Quad-register operations:
case MVT::v16i8: OpcodeIndex = 0; break;
case MVT::v8f16:
case MVT::v8i16: OpcodeIndex = 1; break;
case MVT::v4f32:
case MVT::v4i32: OpcodeIndex = 2; break;
case MVT::v2f64:
case MVT::v2i64: OpcodeIndex = 3; break;
}
EVT ResTy;
if (NumVecs == 1)
ResTy = VT;
else {
unsigned ResTyElts = (NumVecs == 3) ? 4 : NumVecs;
if (!is64BitVector)
ResTyElts *= 2;
ResTy = EVT::getVectorVT(*CurDAG->getContext(), MVT::i64, ResTyElts);
}
std::vector<EVT> ResTys;
ResTys.push_back(ResTy);
if (isUpdating)
ResTys.push_back(MVT::i32);
ResTys.push_back(MVT::Other);
SDValue Pred = getAL(CurDAG, dl);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
SDNode *VLd;
SmallVector<SDValue, 7> Ops;
// Double registers and VLD1/VLD2 quad registers are directly supported.
if (is64BitVector || NumVecs <= 2) {
unsigned Opc = (is64BitVector ? DOpcodes[OpcodeIndex] :
QOpcodes0[OpcodeIndex]);
Ops.push_back(MemAddr);
Ops.push_back(Align);
if (isUpdating) {
SDValue Inc = N->getOperand(AddrOpIdx + 1);
bool IsImmUpdate = isPerfectIncrement(Inc, VT, NumVecs);
if (!IsImmUpdate) {
// We use a VLD1 for v1i64 even if the pseudo says vld2/3/4, so
// check for the opcode rather than the number of vector elements.
if (isVLDfixed(Opc))
Opc = getVLDSTRegisterUpdateOpcode(Opc);
Ops.push_back(Inc);
// VLD1/VLD2 fixed increment does not need Reg0 so only include it in
// the operands if not such an opcode.
} else if (!isVLDfixed(Opc))
Ops.push_back(Reg0);
}
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
VLd = CurDAG->getMachineNode(Opc, dl, ResTys, Ops);
} else {
// Otherwise, quad registers are loaded with two separate instructions,
// where one loads the even registers and the other loads the odd registers.
EVT AddrTy = MemAddr.getValueType();
// Load the even subregs. This is always an updating load, so that it
// provides the address to the second load for the odd subregs.
SDValue ImplDef =
SDValue(CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, dl, ResTy), 0);
const SDValue OpsA[] = { MemAddr, Align, Reg0, ImplDef, Pred, Reg0, Chain };
SDNode *VLdA = CurDAG->getMachineNode(QOpcodes0[OpcodeIndex], dl,
ResTy, AddrTy, MVT::Other, OpsA);
Chain = SDValue(VLdA, 2);
// Load the odd subregs.
Ops.push_back(SDValue(VLdA, 1));
Ops.push_back(Align);
if (isUpdating) {
SDValue Inc = N->getOperand(AddrOpIdx + 1);
assert(isa<ConstantSDNode>(Inc.getNode()) &&
"only constant post-increment update allowed for VLD3/4");
(void)Inc;
Ops.push_back(Reg0);
}
Ops.push_back(SDValue(VLdA, 0));
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
VLd = CurDAG->getMachineNode(QOpcodes1[OpcodeIndex], dl, ResTys, Ops);
}
// Transfer memoperands.
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VLd), {MemOp});
if (NumVecs == 1) {
ReplaceNode(N, VLd);
return;
}
// Extract out the subregisters.
SDValue SuperReg = SDValue(VLd, 0);
static_assert(ARM::dsub_7 == ARM::dsub_0 + 7 &&
ARM::qsub_3 == ARM::qsub_0 + 3,
"Unexpected subreg numbering");
unsigned Sub0 = (is64BitVector ? ARM::dsub_0 : ARM::qsub_0);
for (unsigned Vec = 0; Vec < NumVecs; ++Vec)
ReplaceUses(SDValue(N, Vec),
CurDAG->getTargetExtractSubreg(Sub0 + Vec, dl, VT, SuperReg));
ReplaceUses(SDValue(N, NumVecs), SDValue(VLd, 1));
if (isUpdating)
ReplaceUses(SDValue(N, NumVecs + 1), SDValue(VLd, 2));
CurDAG->RemoveDeadNode(N);
}
void ARMDAGToDAGISel::SelectVST(SDNode *N, bool isUpdating, unsigned NumVecs,
const uint16_t *DOpcodes,
const uint16_t *QOpcodes0,
const uint16_t *QOpcodes1) {
assert(NumVecs >= 1 && NumVecs <= 4 && "VST NumVecs out-of-range");
SDLoc dl(N);
SDValue MemAddr, Align;
bool IsIntrinsic = !isUpdating; // By coincidence, all supported updating
// nodes are not intrinsics.
unsigned AddrOpIdx = IsIntrinsic ? 2 : 1;
unsigned Vec0Idx = 3; // AddrOpIdx + (isUpdating ? 2 : 1)
if (!SelectAddrMode6(N, N->getOperand(AddrOpIdx), MemAddr, Align))
return;
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
SDValue Chain = N->getOperand(0);
EVT VT = N->getOperand(Vec0Idx).getValueType();
bool is64BitVector = VT.is64BitVector();
Align = GetVLDSTAlign(Align, dl, NumVecs, is64BitVector);
unsigned OpcodeIndex;
switch (VT.getSimpleVT().SimpleTy) {
default: llvm_unreachable("unhandled vst type");
// Double-register operations:
case MVT::v8i8: OpcodeIndex = 0; break;
case MVT::v4f16:
case MVT::v4i16: OpcodeIndex = 1; break;
case MVT::v2f32:
case MVT::v2i32: OpcodeIndex = 2; break;
case MVT::v1i64: OpcodeIndex = 3; break;
// Quad-register operations:
case MVT::v16i8: OpcodeIndex = 0; break;
case MVT::v8f16:
case MVT::v8i16: OpcodeIndex = 1; break;
case MVT::v4f32:
case MVT::v4i32: OpcodeIndex = 2; break;
case MVT::v2f64:
case MVT::v2i64: OpcodeIndex = 3; break;
}
std::vector<EVT> ResTys;
if (isUpdating)
ResTys.push_back(MVT::i32);
ResTys.push_back(MVT::Other);
SDValue Pred = getAL(CurDAG, dl);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
SmallVector<SDValue, 7> Ops;
// Double registers and VST1/VST2 quad registers are directly supported.
if (is64BitVector || NumVecs <= 2) {
SDValue SrcReg;
if (NumVecs == 1) {
SrcReg = N->getOperand(Vec0Idx);
} else if (is64BitVector) {
// Form a REG_SEQUENCE to force register allocation.
SDValue V0 = N->getOperand(Vec0Idx + 0);
SDValue V1 = N->getOperand(Vec0Idx + 1);
if (NumVecs == 2)
SrcReg = SDValue(createDRegPairNode(MVT::v2i64, V0, V1), 0);
else {
SDValue V2 = N->getOperand(Vec0Idx + 2);
// If it's a vst3, form a quad D-register and leave the last part as
// an undef.
SDValue V3 = (NumVecs == 3)
? SDValue(CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF,dl,VT), 0)
: N->getOperand(Vec0Idx + 3);
SrcReg = SDValue(createQuadDRegsNode(MVT::v4i64, V0, V1, V2, V3), 0);
}
} else {
// Form a QQ register.
SDValue Q0 = N->getOperand(Vec0Idx);
SDValue Q1 = N->getOperand(Vec0Idx + 1);
SrcReg = SDValue(createQRegPairNode(MVT::v4i64, Q0, Q1), 0);
}
unsigned Opc = (is64BitVector ? DOpcodes[OpcodeIndex] :
QOpcodes0[OpcodeIndex]);
Ops.push_back(MemAddr);
Ops.push_back(Align);
if (isUpdating) {
SDValue Inc = N->getOperand(AddrOpIdx + 1);
bool IsImmUpdate = isPerfectIncrement(Inc, VT, NumVecs);
if (!IsImmUpdate) {
// We use a VST1 for v1i64 even if the pseudo says VST2/3/4, so
// check for the opcode rather than the number of vector elements.
if (isVSTfixed(Opc))
Opc = getVLDSTRegisterUpdateOpcode(Opc);
Ops.push_back(Inc);
}
// VST1/VST2 fixed increment does not need Reg0 so only include it in
// the operands if not such an opcode.
else if (!isVSTfixed(Opc))
Ops.push_back(Reg0);
}
Ops.push_back(SrcReg);
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
SDNode *VSt = CurDAG->getMachineNode(Opc, dl, ResTys, Ops);
// Transfer memoperands.
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VSt), {MemOp});
ReplaceNode(N, VSt);
return;
}
// Otherwise, quad registers are stored with two separate instructions,
// where one stores the even registers and the other stores the odd registers.
// Form the QQQQ REG_SEQUENCE.
SDValue V0 = N->getOperand(Vec0Idx + 0);
SDValue V1 = N->getOperand(Vec0Idx + 1);
SDValue V2 = N->getOperand(Vec0Idx + 2);
SDValue V3 = (NumVecs == 3)
? SDValue(CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, dl, VT), 0)
: N->getOperand(Vec0Idx + 3);
SDValue RegSeq = SDValue(createQuadQRegsNode(MVT::v8i64, V0, V1, V2, V3), 0);
// Store the even D registers. This is always an updating store, so that it
// provides the address to the second store for the odd subregs.
const SDValue OpsA[] = { MemAddr, Align, Reg0, RegSeq, Pred, Reg0, Chain };
SDNode *VStA = CurDAG->getMachineNode(QOpcodes0[OpcodeIndex], dl,
MemAddr.getValueType(),
MVT::Other, OpsA);
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VStA), {MemOp});
Chain = SDValue(VStA, 1);
// Store the odd D registers.
Ops.push_back(SDValue(VStA, 0));
Ops.push_back(Align);
if (isUpdating) {
SDValue Inc = N->getOperand(AddrOpIdx + 1);
assert(isa<ConstantSDNode>(Inc.getNode()) &&
"only constant post-increment update allowed for VST3/4");
(void)Inc;
Ops.push_back(Reg0);
}
Ops.push_back(RegSeq);
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
SDNode *VStB = CurDAG->getMachineNode(QOpcodes1[OpcodeIndex], dl, ResTys,
Ops);
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VStB), {MemOp});
ReplaceNode(N, VStB);
}
void ARMDAGToDAGISel::SelectVLDSTLane(SDNode *N, bool IsLoad, bool isUpdating,
unsigned NumVecs,
const uint16_t *DOpcodes,
const uint16_t *QOpcodes) {
assert(NumVecs >=2 && NumVecs <= 4 && "VLDSTLane NumVecs out-of-range");
SDLoc dl(N);
SDValue MemAddr, Align;
bool IsIntrinsic = !isUpdating; // By coincidence, all supported updating
// nodes are not intrinsics.
unsigned AddrOpIdx = IsIntrinsic ? 2 : 1;
unsigned Vec0Idx = 3; // AddrOpIdx + (isUpdating ? 2 : 1)
if (!SelectAddrMode6(N, N->getOperand(AddrOpIdx), MemAddr, Align))
return;
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
SDValue Chain = N->getOperand(0);
unsigned Lane =
cast<ConstantSDNode>(N->getOperand(Vec0Idx + NumVecs))->getZExtValue();
EVT VT = N->getOperand(Vec0Idx).getValueType();
bool is64BitVector = VT.is64BitVector();
unsigned Alignment = 0;
if (NumVecs != 3) {
Alignment = cast<ConstantSDNode>(Align)->getZExtValue();
unsigned NumBytes = NumVecs * VT.getScalarSizeInBits() / 8;
if (Alignment > NumBytes)
Alignment = NumBytes;
if (Alignment < 8 && Alignment < NumBytes)
Alignment = 0;
// Alignment must be a power of two; make sure of that.
Alignment = (Alignment & -Alignment);
if (Alignment == 1)
Alignment = 0;
}
Align = CurDAG->getTargetConstant(Alignment, dl, MVT::i32);
unsigned OpcodeIndex;
switch (VT.getSimpleVT().SimpleTy) {
default: llvm_unreachable("unhandled vld/vst lane type");
// Double-register operations:
case MVT::v8i8: OpcodeIndex = 0; break;
case MVT::v4f16:
case MVT::v4i16: OpcodeIndex = 1; break;
case MVT::v2f32:
case MVT::v2i32: OpcodeIndex = 2; break;
// Quad-register operations:
case MVT::v8f16:
case MVT::v8i16: OpcodeIndex = 0; break;
case MVT::v4f32:
case MVT::v4i32: OpcodeIndex = 1; break;
}
std::vector<EVT> ResTys;
if (IsLoad) {
unsigned ResTyElts = (NumVecs == 3) ? 4 : NumVecs;
if (!is64BitVector)
ResTyElts *= 2;
ResTys.push_back(EVT::getVectorVT(*CurDAG->getContext(),
MVT::i64, ResTyElts));
}
if (isUpdating)
ResTys.push_back(MVT::i32);
ResTys.push_back(MVT::Other);
SDValue Pred = getAL(CurDAG, dl);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
SmallVector<SDValue, 8> Ops;
Ops.push_back(MemAddr);
Ops.push_back(Align);
if (isUpdating) {
SDValue Inc = N->getOperand(AddrOpIdx + 1);
bool IsImmUpdate =
isPerfectIncrement(Inc, VT.getVectorElementType(), NumVecs);
Ops.push_back(IsImmUpdate ? Reg0 : Inc);
}
SDValue SuperReg;
SDValue V0 = N->getOperand(Vec0Idx + 0);
SDValue V1 = N->getOperand(Vec0Idx + 1);
if (NumVecs == 2) {
if (is64BitVector)
SuperReg = SDValue(createDRegPairNode(MVT::v2i64, V0, V1), 0);
else
SuperReg = SDValue(createQRegPairNode(MVT::v4i64, V0, V1), 0);
} else {
SDValue V2 = N->getOperand(Vec0Idx + 2);
SDValue V3 = (NumVecs == 3)
? SDValue(CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, dl, VT), 0)
: N->getOperand(Vec0Idx + 3);
if (is64BitVector)
SuperReg = SDValue(createQuadDRegsNode(MVT::v4i64, V0, V1, V2, V3), 0);
else
SuperReg = SDValue(createQuadQRegsNode(MVT::v8i64, V0, V1, V2, V3), 0);
}
Ops.push_back(SuperReg);
Ops.push_back(getI32Imm(Lane, dl));
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
unsigned Opc = (is64BitVector ? DOpcodes[OpcodeIndex] :
QOpcodes[OpcodeIndex]);
SDNode *VLdLn = CurDAG->getMachineNode(Opc, dl, ResTys, Ops);
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VLdLn), {MemOp});
if (!IsLoad) {
ReplaceNode(N, VLdLn);
return;
}
// Extract the subregisters.
SuperReg = SDValue(VLdLn, 0);
static_assert(ARM::dsub_7 == ARM::dsub_0 + 7 &&
ARM::qsub_3 == ARM::qsub_0 + 3,
"Unexpected subreg numbering");
unsigned Sub0 = is64BitVector ? ARM::dsub_0 : ARM::qsub_0;
for (unsigned Vec = 0; Vec < NumVecs; ++Vec)
ReplaceUses(SDValue(N, Vec),
CurDAG->getTargetExtractSubreg(Sub0 + Vec, dl, VT, SuperReg));
ReplaceUses(SDValue(N, NumVecs), SDValue(VLdLn, 1));
if (isUpdating)
ReplaceUses(SDValue(N, NumVecs + 1), SDValue(VLdLn, 2));
CurDAG->RemoveDeadNode(N);
}
template <typename SDValueVector>
void ARMDAGToDAGISel::AddMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc,
SDValue PredicateMask) {
Ops.push_back(CurDAG->getTargetConstant(ARMVCC::Then, Loc, MVT::i32));
Ops.push_back(PredicateMask);
}
template <typename SDValueVector>
void ARMDAGToDAGISel::AddMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc,
SDValue PredicateMask,
SDValue Inactive) {
Ops.push_back(CurDAG->getTargetConstant(ARMVCC::Then, Loc, MVT::i32));
Ops.push_back(PredicateMask);
Ops.push_back(Inactive);
}
template <typename SDValueVector>
void ARMDAGToDAGISel::AddEmptyMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc) {
Ops.push_back(CurDAG->getTargetConstant(ARMVCC::None, Loc, MVT::i32));
Ops.push_back(CurDAG->getRegister(0, MVT::i32));
}
template <typename SDValueVector>
void ARMDAGToDAGISel::AddEmptyMVEPredicateToOps(SDValueVector &Ops, SDLoc Loc,
EVT InactiveTy) {
Ops.push_back(CurDAG->getTargetConstant(ARMVCC::None, Loc, MVT::i32));
Ops.push_back(CurDAG->getRegister(0, MVT::i32));
Ops.push_back(SDValue(
CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, Loc, InactiveTy), 0));
}
void ARMDAGToDAGISel::SelectMVE_WB(SDNode *N, const uint16_t *Opcodes,
bool Predicated) {
SDLoc Loc(N);
SmallVector<SDValue, 8> Ops;
uint16_t Opcode;
switch (N->getValueType(1).getVectorElementType().getSizeInBits()) {
case 32:
Opcode = Opcodes[0];
break;
case 64:
Opcode = Opcodes[1];
break;
default:
llvm_unreachable("bad vector element size in SelectMVE_WB");
}
Ops.push_back(N->getOperand(2)); // vector of base addresses
int32_t ImmValue = cast<ConstantSDNode>(N->getOperand(3))->getZExtValue();
Ops.push_back(getI32Imm(ImmValue, Loc)); // immediate offset
if (Predicated)
AddMVEPredicateToOps(Ops, Loc, N->getOperand(4));
else
AddEmptyMVEPredicateToOps(Ops, Loc);
Ops.push_back(N->getOperand(0)); // chain
CurDAG->SelectNodeTo(N, Opcode, N->getVTList(), makeArrayRef(Ops));
}
void ARMDAGToDAGISel::SelectMVE_LongShift(SDNode *N, uint16_t Opcode,
bool Immediate,
bool HasSaturationOperand) {
SDLoc Loc(N);
SmallVector<SDValue, 8> Ops;
// Two 32-bit halves of the value to be shifted
Ops.push_back(N->getOperand(1));
Ops.push_back(N->getOperand(2));
// The shift count
if (Immediate) {
int32_t ImmValue = cast<ConstantSDNode>(N->getOperand(3))->getZExtValue();
Ops.push_back(getI32Imm(ImmValue, Loc)); // immediate shift count
} else {
Ops.push_back(N->getOperand(3));
}
// The immediate saturation operand, if any
if (HasSaturationOperand) {
int32_t SatOp = cast<ConstantSDNode>(N->getOperand(4))->getZExtValue();
int SatBit = (SatOp == 64 ? 0 : 1);
Ops.push_back(getI32Imm(SatBit, Loc));
}
// MVE scalar shifts are IT-predicable, so include the standard
// predicate arguments.
Ops.push_back(getAL(CurDAG, Loc));
Ops.push_back(CurDAG->getRegister(0, MVT::i32));
CurDAG->SelectNodeTo(N, Opcode, N->getVTList(), makeArrayRef(Ops));
}
void ARMDAGToDAGISel::SelectMVE_VADCSBC(SDNode *N, uint16_t OpcodeWithCarry,
uint16_t OpcodeWithNoCarry,
bool Add, bool Predicated) {
SDLoc Loc(N);
SmallVector<SDValue, 8> Ops;
uint16_t Opcode;
unsigned FirstInputOp = Predicated ? 2 : 1;
// Two input vectors and the input carry flag
Ops.push_back(N->getOperand(FirstInputOp));
Ops.push_back(N->getOperand(FirstInputOp + 1));
SDValue CarryIn = N->getOperand(FirstInputOp + 2);
ConstantSDNode *CarryInConstant = dyn_cast<ConstantSDNode>(CarryIn);
uint32_t CarryMask = 1 << 29;
uint32_t CarryExpected = Add ? 0 : CarryMask;
if (CarryInConstant &&
(CarryInConstant->getZExtValue() & CarryMask) == CarryExpected) {
Opcode = OpcodeWithNoCarry;
} else {
Ops.push_back(CarryIn);
Opcode = OpcodeWithCarry;
}
if (Predicated)
AddMVEPredicateToOps(Ops, Loc,
N->getOperand(FirstInputOp + 3), // predicate
N->getOperand(FirstInputOp - 1)); // inactive
else
AddEmptyMVEPredicateToOps(Ops, Loc, N->getValueType(0));
CurDAG->SelectNodeTo(N, Opcode, N->getVTList(), makeArrayRef(Ops));
}
static bool SDValueToConstBool(SDValue SDVal) {
assert(isa<ConstantSDNode>(SDVal) && "expected a compile-time constant");
ConstantSDNode *SDValConstant = dyn_cast<ConstantSDNode>(SDVal);
uint64_t Value = SDValConstant->getZExtValue();
assert((Value == 0 || Value == 1) && "expected value 0 or 1");
return Value;
}
void ARMDAGToDAGISel::SelectBaseMVE_VMLLDAV(SDNode *N, bool Predicated,
const uint16_t *OpcodesS,
const uint16_t *OpcodesU,
size_t Stride, size_t TySize) {
assert(TySize < Stride && "Invalid TySize");
bool IsUnsigned = SDValueToConstBool(N->getOperand(1));
bool IsSub = SDValueToConstBool(N->getOperand(2));
bool IsExchange = SDValueToConstBool(N->getOperand(3));
if (IsUnsigned) {
assert(!IsSub &&
"Unsigned versions of vmlsldav[a]/vrmlsldavh[a] do not exist");
assert(!IsExchange &&
"Unsigned versions of vmlaldav[a]x/vrmlaldavh[a]x do not exist");
}
auto OpIsZero = [N](size_t OpNo) {
if (ConstantSDNode *OpConst = dyn_cast<ConstantSDNode>(N->getOperand(OpNo)))
if (OpConst->getZExtValue() == 0)
return true;
return false;
};
// If the input accumulator value is not zero, select an instruction with
// accumulator, otherwise select an instruction without accumulator
bool IsAccum = !(OpIsZero(4) && OpIsZero(5));
const uint16_t *Opcodes = IsUnsigned ? OpcodesU : OpcodesS;
if (IsSub)
Opcodes += 4 * Stride;
if (IsExchange)
Opcodes += 2 * Stride;
if (IsAccum)
Opcodes += Stride;
uint16_t Opcode = Opcodes[TySize];
SDLoc Loc(N);
SmallVector<SDValue, 8> Ops;
// Push the accumulator operands, if they are used
if (IsAccum) {
Ops.push_back(N->getOperand(4));
Ops.push_back(N->getOperand(5));
}
// Push the two vector operands
Ops.push_back(N->getOperand(6));
Ops.push_back(N->getOperand(7));
if (Predicated)
AddMVEPredicateToOps(Ops, Loc, N->getOperand(8));
else
AddEmptyMVEPredicateToOps(Ops, Loc);
CurDAG->SelectNodeTo(N, Opcode, N->getVTList(), makeArrayRef(Ops));
}
void ARMDAGToDAGISel::SelectMVE_VMLLDAV(SDNode *N, bool Predicated,
const uint16_t *OpcodesS,
const uint16_t *OpcodesU) {
EVT VecTy = N->getOperand(6).getValueType();
size_t SizeIndex;
switch (VecTy.getVectorElementType().getSizeInBits()) {
case 16:
SizeIndex = 0;
break;
case 32:
SizeIndex = 1;
break;
default:
llvm_unreachable("bad vector element size");
}
SelectBaseMVE_VMLLDAV(N, Predicated, OpcodesS, OpcodesU, 2, SizeIndex);
}
void ARMDAGToDAGISel::SelectMVE_VRMLLDAVH(SDNode *N, bool Predicated,
const uint16_t *OpcodesS,
const uint16_t *OpcodesU) {
assert(
N->getOperand(6).getValueType().getVectorElementType().getSizeInBits() ==
32 &&
"bad vector element size");
SelectBaseMVE_VMLLDAV(N, Predicated, OpcodesS, OpcodesU, 1, 0);
}
void ARMDAGToDAGISel::SelectMVE_VLD(SDNode *N, unsigned NumVecs,
const uint16_t *const *Opcodes) {
EVT VT = N->getValueType(0);
SDLoc Loc(N);
const uint16_t *OurOpcodes;
switch (VT.getVectorElementType().getSizeInBits()) {
case 8:
OurOpcodes = Opcodes[0];
break;
case 16:
OurOpcodes = Opcodes[1];
break;
case 32:
OurOpcodes = Opcodes[2];
break;
default:
llvm_unreachable("bad vector element size in SelectMVE_VLD");
}
EVT DataTy = EVT::getVectorVT(*CurDAG->getContext(), MVT::i64, NumVecs * 2);
EVT ResultTys[] = {DataTy, MVT::Other};
auto Data = SDValue(
CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, Loc, DataTy), 0);
SDValue Chain = N->getOperand(0);
for (unsigned Stage = 0; Stage < NumVecs; ++Stage) {
SDValue Ops[] = {Data, N->getOperand(2), Chain};
auto LoadInst =
CurDAG->getMachineNode(OurOpcodes[Stage], Loc, ResultTys, Ops);
Data = SDValue(LoadInst, 0);
Chain = SDValue(LoadInst, 1);
}
for (unsigned i = 0; i < NumVecs; i++)
ReplaceUses(SDValue(N, i),
CurDAG->getTargetExtractSubreg(ARM::qsub_0 + i, Loc, VT, Data));
ReplaceUses(SDValue(N, NumVecs), Chain);
CurDAG->RemoveDeadNode(N);
}
void ARMDAGToDAGISel::SelectVLDDup(SDNode *N, bool IsIntrinsic,
bool isUpdating, unsigned NumVecs,
const uint16_t *DOpcodes,
const uint16_t *QOpcodes0,
const uint16_t *QOpcodes1) {
assert(NumVecs >= 1 && NumVecs <= 4 && "VLDDup NumVecs out-of-range");
SDLoc dl(N);
SDValue MemAddr, Align;
unsigned AddrOpIdx = IsIntrinsic ? 2 : 1;
if (!SelectAddrMode6(N, N->getOperand(AddrOpIdx), MemAddr, Align))
return;
SDValue Chain = N->getOperand(0);
EVT VT = N->getValueType(0);
bool is64BitVector = VT.is64BitVector();
unsigned Alignment = 0;
if (NumVecs != 3) {
Alignment = cast<ConstantSDNode>(Align)->getZExtValue();
unsigned NumBytes = NumVecs * VT.getScalarSizeInBits() / 8;
if (Alignment > NumBytes)
Alignment = NumBytes;
if (Alignment < 8 && Alignment < NumBytes)
Alignment = 0;
// Alignment must be a power of two; make sure of that.
Alignment = (Alignment & -Alignment);
if (Alignment == 1)
Alignment = 0;
}
Align = CurDAG->getTargetConstant(Alignment, dl, MVT::i32);
unsigned OpcodeIndex;
switch (VT.getSimpleVT().SimpleTy) {
default: llvm_unreachable("unhandled vld-dup type");
case MVT::v8i8:
case MVT::v16i8: OpcodeIndex = 0; break;
case MVT::v4i16:
case MVT::v8i16:
case MVT::v4f16:
case MVT::v8f16:
OpcodeIndex = 1; break;
case MVT::v2f32:
case MVT::v2i32:
case MVT::v4f32:
case MVT::v4i32: OpcodeIndex = 2; break;
case MVT::v1f64:
case MVT::v1i64: OpcodeIndex = 3; break;
}
unsigned ResTyElts = (NumVecs == 3) ? 4 : NumVecs;
if (!is64BitVector)
ResTyElts *= 2;
EVT ResTy = EVT::getVectorVT(*CurDAG->getContext(), MVT::i64, ResTyElts);
std::vector<EVT> ResTys;
ResTys.push_back(ResTy);
if (isUpdating)
ResTys.push_back(MVT::i32);
ResTys.push_back(MVT::Other);
SDValue Pred = getAL(CurDAG, dl);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
SDNode *VLdDup;
if (is64BitVector || NumVecs == 1) {
SmallVector<SDValue, 6> Ops;
Ops.push_back(MemAddr);
Ops.push_back(Align);
unsigned Opc = is64BitVector ? DOpcodes[OpcodeIndex] :
QOpcodes0[OpcodeIndex];
if (isUpdating) {
// fixed-stride update instructions don't have an explicit writeback
// operand. It's implicit in the opcode itself.
SDValue Inc = N->getOperand(2);
bool IsImmUpdate =
isPerfectIncrement(Inc, VT.getVectorElementType(), NumVecs);
if (NumVecs <= 2 && !IsImmUpdate)
Opc = getVLDSTRegisterUpdateOpcode(Opc);
if (!IsImmUpdate)
Ops.push_back(Inc);
// FIXME: VLD3 and VLD4 haven't been updated to that form yet.
else if (NumVecs > 2)
Ops.push_back(Reg0);
}
Ops.push_back(Pred);
Ops.push_back(Reg0);
Ops.push_back(Chain);
VLdDup = CurDAG->getMachineNode(Opc, dl, ResTys, Ops);
} else if (NumVecs == 2) {
const SDValue OpsA[] = { MemAddr, Align, Pred, Reg0, Chain };
SDNode *VLdA = CurDAG->getMachineNode(QOpcodes0[OpcodeIndex],
dl, ResTys, OpsA);
Chain = SDValue(VLdA, 1);
const SDValue OpsB[] = { MemAddr, Align, Pred, Reg0, Chain };
VLdDup = CurDAG->getMachineNode(QOpcodes1[OpcodeIndex], dl, ResTys, OpsB);
} else {
SDValue ImplDef =
SDValue(CurDAG->getMachineNode(TargetOpcode::IMPLICIT_DEF, dl, ResTy), 0);
const SDValue OpsA[] = { MemAddr, Align, ImplDef, Pred, Reg0, Chain };
SDNode *VLdA = CurDAG->getMachineNode(QOpcodes0[OpcodeIndex],
dl, ResTys, OpsA);
SDValue SuperReg = SDValue(VLdA, 0);
Chain = SDValue(VLdA, 1);
const SDValue OpsB[] = { MemAddr, Align, SuperReg, Pred, Reg0, Chain };
VLdDup = CurDAG->getMachineNode(QOpcodes1[OpcodeIndex], dl, ResTys, OpsB);
}
// Transfer memoperands.
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(VLdDup), {MemOp});
// Extract the subregisters.
if (NumVecs == 1) {
ReplaceUses(SDValue(N, 0), SDValue(VLdDup, 0));
} else {
SDValue SuperReg = SDValue(VLdDup, 0);
static_assert(ARM::dsub_7 == ARM::dsub_0 + 7, "Unexpected subreg numbering");
unsigned SubIdx = is64BitVector ? ARM::dsub_0 : ARM::qsub_0;
for (unsigned Vec = 0; Vec != NumVecs; ++Vec) {
ReplaceUses(SDValue(N, Vec),
CurDAG->getTargetExtractSubreg(SubIdx+Vec, dl, VT, SuperReg));
}
}
ReplaceUses(SDValue(N, NumVecs), SDValue(VLdDup, 1));
if (isUpdating)
ReplaceUses(SDValue(N, NumVecs + 1), SDValue(VLdDup, 2));
CurDAG->RemoveDeadNode(N);
}
bool ARMDAGToDAGISel::tryV6T2BitfieldExtractOp(SDNode *N, bool isSigned) {
if (!Subtarget->hasV6T2Ops())
return false;
unsigned Opc = isSigned
? (Subtarget->isThumb() ? ARM::t2SBFX : ARM::SBFX)
: (Subtarget->isThumb() ? ARM::t2UBFX : ARM::UBFX);
SDLoc dl(N);
// For unsigned extracts, check for a shift right and mask
unsigned And_imm = 0;
if (N->getOpcode() == ISD::AND) {
if (isOpcWithIntImmediate(N, ISD::AND, And_imm)) {
// The immediate is a mask of the low bits iff imm & (imm+1) == 0
if (And_imm & (And_imm + 1))
return false;
unsigned Srl_imm = 0;
if (isOpcWithIntImmediate(N->getOperand(0).getNode(), ISD::SRL,
Srl_imm)) {
assert(Srl_imm > 0 && Srl_imm < 32 && "bad amount in shift node!");
// Mask off the unnecessary bits of the AND immediate; normally
// DAGCombine will do this, but that might not happen if
// targetShrinkDemandedConstant chooses a different immediate.
And_imm &= -1U >> Srl_imm;
// Note: The width operand is encoded as width-1.
unsigned Width = countTrailingOnes(And_imm) - 1;
unsigned LSB = Srl_imm;
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
if ((LSB + Width + 1) == N->getValueType(0).getSizeInBits()) {
// It's cheaper to use a right shift to extract the top bits.
if (Subtarget->isThumb()) {
Opc = isSigned ? ARM::t2ASRri : ARM::t2LSRri;
SDValue Ops[] = { N->getOperand(0).getOperand(0),
CurDAG->getTargetConstant(LSB, dl, MVT::i32),
getAL(CurDAG, dl), Reg0, Reg0 };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return true;
}
// ARM models shift instructions as MOVsi with shifter operand.
ARM_AM::ShiftOpc ShOpcVal = ARM_AM::getShiftOpcForNode(ISD::SRL);
SDValue ShOpc =
CurDAG->getTargetConstant(ARM_AM::getSORegOpc(ShOpcVal, LSB), dl,
MVT::i32);
SDValue Ops[] = { N->getOperand(0).getOperand(0), ShOpc,
getAL(CurDAG, dl), Reg0, Reg0 };
CurDAG->SelectNodeTo(N, ARM::MOVsi, MVT::i32, Ops);
return true;
}
assert(LSB + Width + 1 <= 32 && "Shouldn't create an invalid ubfx");
SDValue Ops[] = { N->getOperand(0).getOperand(0),
CurDAG->getTargetConstant(LSB, dl, MVT::i32),
CurDAG->getTargetConstant(Width, dl, MVT::i32),
getAL(CurDAG, dl), Reg0 };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return true;
}
}
return false;
}
// Otherwise, we're looking for a shift of a shift
unsigned Shl_imm = 0;
if (isOpcWithIntImmediate(N->getOperand(0).getNode(), ISD::SHL, Shl_imm)) {
assert(Shl_imm > 0 && Shl_imm < 32 && "bad amount in shift node!");
unsigned Srl_imm = 0;
if (isInt32Immediate(N->getOperand(1), Srl_imm)) {
assert(Srl_imm > 0 && Srl_imm < 32 && "bad amount in shift node!");
// Note: The width operand is encoded as width-1.
unsigned Width = 32 - Srl_imm - 1;
int LSB = Srl_imm - Shl_imm;
if (LSB < 0)
return false;
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
assert(LSB + Width + 1 <= 32 && "Shouldn't create an invalid ubfx");
SDValue Ops[] = { N->getOperand(0).getOperand(0),
CurDAG->getTargetConstant(LSB, dl, MVT::i32),
CurDAG->getTargetConstant(Width, dl, MVT::i32),
getAL(CurDAG, dl), Reg0 };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return true;
}
}
// Or we are looking for a shift of an and, with a mask operand
if (isOpcWithIntImmediate(N->getOperand(0).getNode(), ISD::AND, And_imm) &&
isShiftedMask_32(And_imm)) {
unsigned Srl_imm = 0;
unsigned LSB = countTrailingZeros(And_imm);
// Shift must be the same as the ands lsb
if (isInt32Immediate(N->getOperand(1), Srl_imm) && Srl_imm == LSB) {
assert(Srl_imm > 0 && Srl_imm < 32 && "bad amount in shift node!");
unsigned MSB = 31 - countLeadingZeros(And_imm);
// Note: The width operand is encoded as width-1.
unsigned Width = MSB - LSB;
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
assert(Srl_imm + Width + 1 <= 32 && "Shouldn't create an invalid ubfx");
SDValue Ops[] = { N->getOperand(0).getOperand(0),
CurDAG->getTargetConstant(Srl_imm, dl, MVT::i32),
CurDAG->getTargetConstant(Width, dl, MVT::i32),
getAL(CurDAG, dl), Reg0 };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return true;
}
}
if (N->getOpcode() == ISD::SIGN_EXTEND_INREG) {
unsigned Width = cast<VTSDNode>(N->getOperand(1))->getVT().getSizeInBits();
unsigned LSB = 0;
if (!isOpcWithIntImmediate(N->getOperand(0).getNode(), ISD::SRL, LSB) &&
!isOpcWithIntImmediate(N->getOperand(0).getNode(), ISD::SRA, LSB))
return false;
if (LSB + Width > 32)
return false;
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
assert(LSB + Width <= 32 && "Shouldn't create an invalid ubfx");
SDValue Ops[] = { N->getOperand(0).getOperand(0),
CurDAG->getTargetConstant(LSB, dl, MVT::i32),
CurDAG->getTargetConstant(Width - 1, dl, MVT::i32),
getAL(CurDAG, dl), Reg0 };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return true;
}
return false;
}
/// Target-specific DAG combining for ISD::XOR.
/// Target-independent combining lowers SELECT_CC nodes of the form
/// select_cc setg[ge] X, 0, X, -X
/// select_cc setgt X, -1, X, -X
/// select_cc setl[te] X, 0, -X, X
/// select_cc setlt X, 1, -X, X
/// which represent Integer ABS into:
/// Y = sra (X, size(X)-1); xor (add (X, Y), Y)
/// ARM instruction selection detects the latter and matches it to
/// ARM::ABS or ARM::t2ABS machine node.
bool ARMDAGToDAGISel::tryABSOp(SDNode *N){
SDValue XORSrc0 = N->getOperand(0);
SDValue XORSrc1 = N->getOperand(1);
EVT VT = N->getValueType(0);
if (Subtarget->isThumb1Only())
return false;
if (XORSrc0.getOpcode() != ISD::ADD || XORSrc1.getOpcode() != ISD::SRA)
return false;
SDValue ADDSrc0 = XORSrc0.getOperand(0);
SDValue ADDSrc1 = XORSrc0.getOperand(1);
SDValue SRASrc0 = XORSrc1.getOperand(0);
SDValue SRASrc1 = XORSrc1.getOperand(1);
ConstantSDNode *SRAConstant = dyn_cast<ConstantSDNode>(SRASrc1);
EVT XType = SRASrc0.getValueType();
unsigned Size = XType.getSizeInBits() - 1;
if (ADDSrc1 == XORSrc1 && ADDSrc0 == SRASrc0 &&
XType.isInteger() && SRAConstant != nullptr &&
Size == SRAConstant->getZExtValue()) {
unsigned Opcode = Subtarget->isThumb2() ? ARM::t2ABS : ARM::ABS;
CurDAG->SelectNodeTo(N, Opcode, VT, ADDSrc0);
return true;
}
return false;
}
/// We've got special pseudo-instructions for these
void ARMDAGToDAGISel::SelectCMP_SWAP(SDNode *N) {
unsigned Opcode;
EVT MemTy = cast<MemSDNode>(N)->getMemoryVT();
if (MemTy == MVT::i8)
Opcode = ARM::CMP_SWAP_8;
else if (MemTy == MVT::i16)
Opcode = ARM::CMP_SWAP_16;
else if (MemTy == MVT::i32)
Opcode = ARM::CMP_SWAP_32;
else
llvm_unreachable("Unknown AtomicCmpSwap type");
SDValue Ops[] = {N->getOperand(1), N->getOperand(2), N->getOperand(3),
N->getOperand(0)};
SDNode *CmpSwap = CurDAG->getMachineNode(
Opcode, SDLoc(N),
CurDAG->getVTList(MVT::i32, MVT::i32, MVT::Other), Ops);
MachineMemOperand *MemOp = cast<MemSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(CmpSwap), {MemOp});
ReplaceUses(SDValue(N, 0), SDValue(CmpSwap, 0));
ReplaceUses(SDValue(N, 1), SDValue(CmpSwap, 2));
CurDAG->RemoveDeadNode(N);
}
static Optional<std::pair<unsigned, unsigned>>
getContiguousRangeOfSetBits(const APInt &A) {
unsigned FirstOne = A.getBitWidth() - A.countLeadingZeros() - 1;
unsigned LastOne = A.countTrailingZeros();
if (A.countPopulation() != (FirstOne - LastOne + 1))
return Optional<std::pair<unsigned,unsigned>>();
return std::make_pair(FirstOne, LastOne);
}
void ARMDAGToDAGISel::SelectCMPZ(SDNode *N, bool &SwitchEQNEToPLMI) {
assert(N->getOpcode() == ARMISD::CMPZ);
SwitchEQNEToPLMI = false;
if (!Subtarget->isThumb())
// FIXME: Work out whether it is profitable to do this in A32 mode - LSL and
// LSR don't exist as standalone instructions - they need the barrel shifter.
return;
// select (cmpz (and X, C), #0) -> (LSLS X) or (LSRS X) or (LSRS (LSLS X))
SDValue And = N->getOperand(0);
if (!And->hasOneUse())
return;
SDValue Zero = N->getOperand(1);
if (!isa<ConstantSDNode>(Zero) || !cast<ConstantSDNode>(Zero)->isNullValue() ||
And->getOpcode() != ISD::AND)
return;
SDValue X = And.getOperand(0);
auto C = dyn_cast<ConstantSDNode>(And.getOperand(1));
if (!C)
return;
auto Range = getContiguousRangeOfSetBits(C->getAPIntValue());
if (!Range)
return;
// There are several ways to lower this:
SDNode *NewN;
SDLoc dl(N);
auto EmitShift = [&](unsigned Opc, SDValue Src, unsigned Imm) -> SDNode* {
if (Subtarget->isThumb2()) {
Opc = (Opc == ARM::tLSLri) ? ARM::t2LSLri : ARM::t2LSRri;
SDValue Ops[] = { Src, CurDAG->getTargetConstant(Imm, dl, MVT::i32),
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32) };
return CurDAG->getMachineNode(Opc, dl, MVT::i32, Ops);
} else {
SDValue Ops[] = {CurDAG->getRegister(ARM::CPSR, MVT::i32), Src,
CurDAG->getTargetConstant(Imm, dl, MVT::i32),
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32)};
return CurDAG->getMachineNode(Opc, dl, MVT::i32, Ops);
}
};
if (Range->second == 0) {
// 1. Mask includes the LSB -> Simply shift the top N bits off
NewN = EmitShift(ARM::tLSLri, X, 31 - Range->first);
ReplaceNode(And.getNode(), NewN);
} else if (Range->first == 31) {
// 2. Mask includes the MSB -> Simply shift the bottom N bits off
NewN = EmitShift(ARM::tLSRri, X, Range->second);
ReplaceNode(And.getNode(), NewN);
} else if (Range->first == Range->second) {
// 3. Only one bit is set. We can shift this into the sign bit and use a
// PL/MI comparison.
NewN = EmitShift(ARM::tLSLri, X, 31 - Range->first);
ReplaceNode(And.getNode(), NewN);
SwitchEQNEToPLMI = true;
} else if (!Subtarget->hasV6T2Ops()) {
// 4. Do a double shift to clear bottom and top bits, but only in
// thumb-1 mode as in thumb-2 we can use UBFX.
NewN = EmitShift(ARM::tLSLri, X, 31 - Range->first);
NewN = EmitShift(ARM::tLSRri, SDValue(NewN, 0),
Range->second + (31 - Range->first));
ReplaceNode(And.getNode(), NewN);
}
}
void ARMDAGToDAGISel::Select(SDNode *N) {
SDLoc dl(N);
if (N->isMachineOpcode()) {
N->setNodeId(-1);
return; // Already selected.
}
switch (N->getOpcode()) {
default: break;
case ISD::STORE: {
// For Thumb1, match an sp-relative store in C++. This is a little
// unfortunate, but I don't think I can make the chain check work
// otherwise. (The chain of the store has to be the same as the chain
// of the CopyFromReg, or else we can't replace the CopyFromReg with
// a direct reference to "SP".)
//
// This is only necessary on Thumb1 because Thumb1 sp-relative stores use
// a different addressing mode from other four-byte stores.
//
// This pattern usually comes up with call arguments.
StoreSDNode *ST = cast<StoreSDNode>(N);
SDValue Ptr = ST->getBasePtr();
if (Subtarget->isThumb1Only() && ST->isUnindexed()) {
int RHSC = 0;
if (Ptr.getOpcode() == ISD::ADD &&
isScaledConstantInRange(Ptr.getOperand(1), /*Scale=*/4, 0, 256, RHSC))
Ptr = Ptr.getOperand(0);
if (Ptr.getOpcode() == ISD::CopyFromReg &&
cast<RegisterSDNode>(Ptr.getOperand(1))->getReg() == ARM::SP &&
Ptr.getOperand(0) == ST->getChain()) {
SDValue Ops[] = {ST->getValue(),
CurDAG->getRegister(ARM::SP, MVT::i32),
CurDAG->getTargetConstant(RHSC, dl, MVT::i32),
getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
ST->getChain()};
MachineSDNode *ResNode =
CurDAG->getMachineNode(ARM::tSTRspi, dl, MVT::Other, Ops);
MachineMemOperand *MemOp = ST->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(ResNode), {MemOp});
ReplaceNode(N, ResNode);
return;
}
}
break;
}
case ISD::WRITE_REGISTER:
if (tryWriteRegister(N))
return;
break;
case ISD::READ_REGISTER:
if (tryReadRegister(N))
return;
break;
case ISD::INLINEASM:
case ISD::INLINEASM_BR:
if (tryInlineAsm(N))
return;
break;
case ISD::XOR:
// Select special operations if XOR node forms integer ABS pattern
if (tryABSOp(N))
return;
// Other cases are autogenerated.
break;
case ISD::Constant: {
unsigned Val = cast<ConstantSDNode>(N)->getZExtValue();
// If we can't materialize the constant we need to use a literal pool
if (ConstantMaterializationCost(Val, Subtarget) > 2) {
SDValue CPIdx = CurDAG->getTargetConstantPool(
ConstantInt::get(Type::getInt32Ty(*CurDAG->getContext()), Val),
TLI->getPointerTy(CurDAG->getDataLayout()));
SDNode *ResNode;
if (Subtarget->isThumb()) {
SDValue Ops[] = {
CPIdx,
getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
CurDAG->getEntryNode()
};
ResNode = CurDAG->getMachineNode(ARM::tLDRpci, dl, MVT::i32, MVT::Other,
Ops);
} else {
SDValue Ops[] = {
CPIdx,
CurDAG->getTargetConstant(0, dl, MVT::i32),
getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
CurDAG->getEntryNode()
};
ResNode = CurDAG->getMachineNode(ARM::LDRcp, dl, MVT::i32, MVT::Other,
Ops);
}
// Annotate the Node with memory operand information so that MachineInstr
// queries work properly. This e.g. gives the register allocation the
// required information for rematerialization.
MachineFunction& MF = CurDAG->getMachineFunction();
MachineMemOperand *MemOp =
MF.getMachineMemOperand(MachinePointerInfo::getConstantPool(MF),
MachineMemOperand::MOLoad, 4, 4);
CurDAG->setNodeMemRefs(cast<MachineSDNode>(ResNode), {MemOp});
ReplaceNode(N, ResNode);
return;
}
// Other cases are autogenerated.
break;
}
case ISD::FrameIndex: {
// Selects to ADDri FI, 0 which in turn will become ADDri SP, imm.
int FI = cast<FrameIndexSDNode>(N)->getIndex();
SDValue TFI = CurDAG->getTargetFrameIndex(
FI, TLI->getPointerTy(CurDAG->getDataLayout()));
if (Subtarget->isThumb1Only()) {
// Set the alignment of the frame object to 4, to avoid having to generate
// more than one ADD
MachineFrameInfo &MFI = MF->getFrameInfo();
if (MFI.getObjectAlignment(FI) < 4)
MFI.setObjectAlignment(FI, 4);
CurDAG->SelectNodeTo(N, ARM::tADDframe, MVT::i32, TFI,
CurDAG->getTargetConstant(0, dl, MVT::i32));
return;
} else {
unsigned Opc = ((Subtarget->isThumb() && Subtarget->hasThumb2()) ?
ARM::t2ADDri : ARM::ADDri);
SDValue Ops[] = { TFI, CurDAG->getTargetConstant(0, dl, MVT::i32),
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32) };
CurDAG->SelectNodeTo(N, Opc, MVT::i32, Ops);
return;
}
}
case ISD::SRL:
if (tryV6T2BitfieldExtractOp(N, false))
return;
break;
case ISD::SIGN_EXTEND_INREG:
case ISD::SRA:
if (tryV6T2BitfieldExtractOp(N, true))
return;
break;
case ISD::MUL:
if (Subtarget->isThumb1Only())
break;
if (ConstantSDNode *C = dyn_cast<ConstantSDNode>(N->getOperand(1))) {
unsigned RHSV = C->getZExtValue();
if (!RHSV) break;
if (isPowerOf2_32(RHSV-1)) { // 2^n+1?
unsigned ShImm = Log2_32(RHSV-1);
if (ShImm >= 32)
break;
SDValue V = N->getOperand(0);
ShImm = ARM_AM::getSORegOpc(ARM_AM::lsl, ShImm);
SDValue ShImmOp = CurDAG->getTargetConstant(ShImm, dl, MVT::i32);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
if (Subtarget->isThumb()) {
SDValue Ops[] = { V, V, ShImmOp, getAL(CurDAG, dl), Reg0, Reg0 };
CurDAG->SelectNodeTo(N, ARM::t2ADDrs, MVT::i32, Ops);
return;
} else {
SDValue Ops[] = { V, V, Reg0, ShImmOp, getAL(CurDAG, dl), Reg0,
Reg0 };
CurDAG->SelectNodeTo(N, ARM::ADDrsi, MVT::i32, Ops);
return;
}
}
if (isPowerOf2_32(RHSV+1)) { // 2^n-1?
unsigned ShImm = Log2_32(RHSV+1);
if (ShImm >= 32)
break;
SDValue V = N->getOperand(0);
ShImm = ARM_AM::getSORegOpc(ARM_AM::lsl, ShImm);
SDValue ShImmOp = CurDAG->getTargetConstant(ShImm, dl, MVT::i32);
SDValue Reg0 = CurDAG->getRegister(0, MVT::i32);
if (Subtarget->isThumb()) {
SDValue Ops[] = { V, V, ShImmOp, getAL(CurDAG, dl), Reg0, Reg0 };
CurDAG->SelectNodeTo(N, ARM::t2RSBrs, MVT::i32, Ops);
return;
} else {
SDValue Ops[] = { V, V, Reg0, ShImmOp, getAL(CurDAG, dl), Reg0,
Reg0 };
CurDAG->SelectNodeTo(N, ARM::RSBrsi, MVT::i32, Ops);
return;
}
}
}
break;
case ISD::AND: {
// Check for unsigned bitfield extract
if (tryV6T2BitfieldExtractOp(N, false))
return;
// If an immediate is used in an AND node, it is possible that the immediate
// can be more optimally materialized when negated. If this is the case we
// can negate the immediate and use a BIC instead.
auto *N1C = dyn_cast<ConstantSDNode>(N->getOperand(1));
if (N1C && N1C->hasOneUse() && Subtarget->isThumb()) {
uint32_t Imm = (uint32_t) N1C->getZExtValue();
// In Thumb2 mode, an AND can take a 12-bit immediate. If this
// immediate can be negated and fit in the immediate operand of
// a t2BIC, don't do any manual transform here as this can be
// handled by the generic ISel machinery.
bool PreferImmediateEncoding =
Subtarget->hasThumb2() && (is_t2_so_imm(Imm) || is_t2_so_imm_not(Imm));
if (!PreferImmediateEncoding &&
ConstantMaterializationCost(Imm, Subtarget) >
ConstantMaterializationCost(~Imm, Subtarget)) {
// The current immediate costs more to materialize than a negated
// immediate, so negate the immediate and use a BIC.
SDValue NewImm =
CurDAG->getConstant(~N1C->getZExtValue(), dl, MVT::i32);
// If the new constant didn't exist before, reposition it in the topological
// ordering so it is just before N. Otherwise, don't touch its location.
if (NewImm->getNodeId() == -1)
CurDAG->RepositionNode(N->getIterator(), NewImm.getNode());
if (!Subtarget->hasThumb2()) {
SDValue Ops[] = {CurDAG->getRegister(ARM::CPSR, MVT::i32),
N->getOperand(0), NewImm, getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32)};
ReplaceNode(N, CurDAG->getMachineNode(ARM::tBIC, dl, MVT::i32, Ops));
return;
} else {
SDValue Ops[] = {N->getOperand(0), NewImm, getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32)};
ReplaceNode(N,
CurDAG->getMachineNode(ARM::t2BICrr, dl, MVT::i32, Ops));
return;
}
}
}
// (and (or x, c2), c1) and top 16-bits of c1 and c2 match, lower 16-bits
// of c1 are 0xffff, and lower 16-bit of c2 are 0. That is, the top 16-bits
// are entirely contributed by c2 and lower 16-bits are entirely contributed
// by x. That's equal to (or (and x, 0xffff), (and c1, 0xffff0000)).
// Select it to: "movt x, ((c1 & 0xffff) >> 16)
EVT VT = N->getValueType(0);
if (VT != MVT::i32)
break;
unsigned Opc = (Subtarget->isThumb() && Subtarget->hasThumb2())
? ARM::t2MOVTi16
: (Subtarget->hasV6T2Ops() ? ARM::MOVTi16 : 0);
if (!Opc)
break;
SDValue N0 = N->getOperand(0), N1 = N->getOperand(1);
N1C = dyn_cast<ConstantSDNode>(N1);
if (!N1C)
break;
if (N0.getOpcode() == ISD::OR && N0.getNode()->hasOneUse()) {
SDValue N2 = N0.getOperand(1);
ConstantSDNode *N2C = dyn_cast<ConstantSDNode>(N2);
if (!N2C)
break;
unsigned N1CVal = N1C->getZExtValue();
unsigned N2CVal = N2C->getZExtValue();
if ((N1CVal & 0xffff0000U) == (N2CVal & 0xffff0000U) &&
(N1CVal & 0xffffU) == 0xffffU &&
(N2CVal & 0xffffU) == 0x0U) {
SDValue Imm16 = CurDAG->getTargetConstant((N2CVal & 0xFFFF0000U) >> 16,
dl, MVT::i32);
SDValue Ops[] = { N0.getOperand(0), Imm16,
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32) };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, VT, Ops));
return;
}
}
break;
}
case ARMISD::UMAAL: {
unsigned Opc = Subtarget->isThumb() ? ARM::t2UMAAL : ARM::UMAAL;
SDValue Ops[] = { N->getOperand(0), N->getOperand(1),
N->getOperand(2), N->getOperand(3),
getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32) };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, MVT::i32, MVT::i32, Ops));
return;
}
case ARMISD::UMLAL:{
if (Subtarget->isThumb()) {
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), N->getOperand(2),
N->getOperand(3), getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32)};
ReplaceNode(
N, CurDAG->getMachineNode(ARM::t2UMLAL, dl, MVT::i32, MVT::i32, Ops));
return;
}else{
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), N->getOperand(2),
N->getOperand(3), getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32) };
ReplaceNode(N, CurDAG->getMachineNode(
Subtarget->hasV6Ops() ? ARM::UMLAL : ARM::UMLALv5, dl,
MVT::i32, MVT::i32, Ops));
return;
}
}
case ARMISD::SMLAL:{
if (Subtarget->isThumb()) {
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), N->getOperand(2),
N->getOperand(3), getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32)};
ReplaceNode(
N, CurDAG->getMachineNode(ARM::t2SMLAL, dl, MVT::i32, MVT::i32, Ops));
return;
}else{
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), N->getOperand(2),
N->getOperand(3), getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32) };
ReplaceNode(N, CurDAG->getMachineNode(
Subtarget->hasV6Ops() ? ARM::SMLAL : ARM::SMLALv5, dl,
MVT::i32, MVT::i32, Ops));
return;
}
}
case ARMISD::SUBE: {
if (!Subtarget->hasV6Ops() || !Subtarget->hasDSP())
break;
// Look for a pattern to match SMMLS
// (sube a, (smul_loHi a, b), (subc 0, (smul_LOhi(a, b))))
if (N->getOperand(1).getOpcode() != ISD::SMUL_LOHI ||
N->getOperand(2).getOpcode() != ARMISD::SUBC ||
!SDValue(N, 1).use_empty())
break;
if (Subtarget->isThumb())
assert(Subtarget->hasThumb2() &&
"This pattern should not be generated for Thumb");
SDValue SmulLoHi = N->getOperand(1);
SDValue Subc = N->getOperand(2);
auto *Zero = dyn_cast<ConstantSDNode>(Subc.getOperand(0));
if (!Zero || Zero->getZExtValue() != 0 ||
Subc.getOperand(1) != SmulLoHi.getValue(0) ||
N->getOperand(1) != SmulLoHi.getValue(1) ||
N->getOperand(2) != Subc.getValue(1))
break;
unsigned Opc = Subtarget->isThumb2() ? ARM::t2SMMLS : ARM::SMMLS;
SDValue Ops[] = { SmulLoHi.getOperand(0), SmulLoHi.getOperand(1),
N->getOperand(0), getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32) };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, MVT::i32, Ops));
return;
}
case ISD::LOAD: {
if (Subtarget->hasMVEIntegerOps() && tryMVEIndexedLoad(N))
return;
if (Subtarget->isThumb() && Subtarget->hasThumb2()) {
if (tryT2IndexedLoad(N))
return;
} else if (Subtarget->isThumb()) {
if (tryT1IndexedLoad(N))
return;
} else if (tryARMIndexedLoad(N))
return;
// Other cases are autogenerated.
break;
}
case ISD::MLOAD:
if (Subtarget->hasMVEIntegerOps() && tryMVEIndexedLoad(N))
return;
// Other cases are autogenerated.
break;
case ARMISD::WLS:
case ARMISD::LE: {
SDValue Ops[] = { N->getOperand(1),
N->getOperand(2),
N->getOperand(0) };
unsigned Opc = N->getOpcode() == ARMISD::WLS ?
ARM::t2WhileLoopStart : ARM::t2LoopEnd;
SDNode *New = CurDAG->getMachineNode(Opc, dl, MVT::Other, Ops);
ReplaceUses(N, New);
CurDAG->RemoveDeadNode(N);
return;
}
+ case ARMISD::LDRD: {
+ if (Subtarget->isThumb2())
+ break; // TableGen handles isel in this case.
+ SDValue Base, RegOffset, ImmOffset;
+ const SDValue &Chain = N->getOperand(0);
+ const SDValue &Addr = N->getOperand(1);
+ SelectAddrMode3(Addr, Base, RegOffset, ImmOffset);
+ SDValue Ops[] = {Base, RegOffset, ImmOffset, Chain};
+ SDNode *New = CurDAG->getMachineNode(ARM::LOADDUAL, dl,
+ {MVT::Untyped, MVT::Other}, Ops);
+ SDValue Lo = CurDAG->getTargetExtractSubreg(ARM::gsub_0, dl, MVT::i32,
+ SDValue(New, 0));
+ SDValue Hi = CurDAG->getTargetExtractSubreg(ARM::gsub_1, dl, MVT::i32,
+ SDValue(New, 0));
+ ReplaceUses(SDValue(N, 0), Lo);
+ ReplaceUses(SDValue(N, 1), Hi);
+ ReplaceUses(SDValue(N, 2), SDValue(New, 1));
+ CurDAG->RemoveDeadNode(N);
+ return;
+ }
case ARMISD::LOOP_DEC: {
SDValue Ops[] = { N->getOperand(1),
N->getOperand(2),
N->getOperand(0) };
SDNode *Dec =
CurDAG->getMachineNode(ARM::t2LoopDec, dl,
CurDAG->getVTList(MVT::i32, MVT::Other), Ops);
ReplaceUses(N, Dec);
CurDAG->RemoveDeadNode(N);
return;
}
case ARMISD::BRCOND: {
// Pattern: (ARMbrcond:void (bb:Other):$dst, (imm:i32):$cc)
// Emits: (Bcc:void (bb:Other):$dst, (imm:i32):$cc)
// Pattern complexity = 6 cost = 1 size = 0
// Pattern: (ARMbrcond:void (bb:Other):$dst, (imm:i32):$cc)
// Emits: (tBcc:void (bb:Other):$dst, (imm:i32):$cc)
// Pattern complexity = 6 cost = 1 size = 0
// Pattern: (ARMbrcond:void (bb:Other):$dst, (imm:i32):$cc)
// Emits: (t2Bcc:void (bb:Other):$dst, (imm:i32):$cc)
// Pattern complexity = 6 cost = 1 size = 0
unsigned Opc = Subtarget->isThumb() ?
((Subtarget->hasThumb2()) ? ARM::t2Bcc : ARM::tBcc) : ARM::Bcc;
SDValue Chain = N->getOperand(0);
SDValue N1 = N->getOperand(1);
SDValue N2 = N->getOperand(2);
SDValue N3 = N->getOperand(3);
SDValue InFlag = N->getOperand(4);
assert(N1.getOpcode() == ISD::BasicBlock);
assert(N2.getOpcode() == ISD::Constant);
assert(N3.getOpcode() == ISD::Register);
unsigned CC = (unsigned) cast<ConstantSDNode>(N2)->getZExtValue();
if (InFlag.getOpcode() == ARMISD::CMPZ) {
if (InFlag.getOperand(0).getOpcode() == ISD::INTRINSIC_W_CHAIN) {
SDValue Int = InFlag.getOperand(0);
uint64_t ID = cast<ConstantSDNode>(Int->getOperand(1))->getZExtValue();
// Handle low-overhead loops.
if (ID == Intrinsic::loop_decrement_reg) {
SDValue Elements = Int.getOperand(2);
SDValue Size = CurDAG->getTargetConstant(
cast<ConstantSDNode>(Int.getOperand(3))->getZExtValue(), dl,
MVT::i32);
SDValue Args[] = { Elements, Size, Int.getOperand(0) };
SDNode *LoopDec =
CurDAG->getMachineNode(ARM::t2LoopDec, dl,
CurDAG->getVTList(MVT::i32, MVT::Other),
Args);
ReplaceUses(Int.getNode(), LoopDec);
SDValue EndArgs[] = { SDValue(LoopDec, 0), N1, Chain };
SDNode *LoopEnd =
CurDAG->getMachineNode(ARM::t2LoopEnd, dl, MVT::Other, EndArgs);
ReplaceUses(N, LoopEnd);
CurDAG->RemoveDeadNode(N);
CurDAG->RemoveDeadNode(InFlag.getNode());
CurDAG->RemoveDeadNode(Int.getNode());
return;
}
}
bool SwitchEQNEToPLMI;
SelectCMPZ(InFlag.getNode(), SwitchEQNEToPLMI);
InFlag = N->getOperand(4);
if (SwitchEQNEToPLMI) {
switch ((ARMCC::CondCodes)CC) {
default: llvm_unreachable("CMPZ must be either NE or EQ!");
case ARMCC::NE:
CC = (unsigned)ARMCC::MI;
break;
case ARMCC::EQ:
CC = (unsigned)ARMCC::PL;
break;
}
}
}
SDValue Tmp2 = CurDAG->getTargetConstant(CC, dl, MVT::i32);
SDValue Ops[] = { N1, Tmp2, N3, Chain, InFlag };
SDNode *ResNode = CurDAG->getMachineNode(Opc, dl, MVT::Other,
MVT::Glue, Ops);
Chain = SDValue(ResNode, 0);
if (N->getNumValues() == 2) {
InFlag = SDValue(ResNode, 1);
ReplaceUses(SDValue(N, 1), InFlag);
}
ReplaceUses(SDValue(N, 0),
SDValue(Chain.getNode(), Chain.getResNo()));
CurDAG->RemoveDeadNode(N);
return;
}
case ARMISD::CMPZ: {
// select (CMPZ X, #-C) -> (CMPZ (ADDS X, #C), #0)
// This allows us to avoid materializing the expensive negative constant.
// The CMPZ #0 is useless and will be peepholed away but we need to keep it
// for its glue output.
SDValue X = N->getOperand(0);
auto *C = dyn_cast<ConstantSDNode>(N->getOperand(1).getNode());
if (C && C->getSExtValue() < 0 && Subtarget->isThumb()) {
int64_t Addend = -C->getSExtValue();
SDNode *Add = nullptr;
// ADDS can be better than CMN if the immediate fits in a
// 16-bit ADDS, which means either [0,256) for tADDi8 or [0,8) for tADDi3.
// Outside that range we can just use a CMN which is 32-bit but has a
// 12-bit immediate range.
if (Addend < 1<<8) {
if (Subtarget->isThumb2()) {
SDValue Ops[] = { X, CurDAG->getTargetConstant(Addend, dl, MVT::i32),
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32),
CurDAG->getRegister(0, MVT::i32) };
Add = CurDAG->getMachineNode(ARM::t2ADDri, dl, MVT::i32, Ops);
} else {
unsigned Opc = (Addend < 1<<3) ? ARM::tADDi3 : ARM::tADDi8;
SDValue Ops[] = {CurDAG->getRegister(ARM::CPSR, MVT::i32), X,
CurDAG->getTargetConstant(Addend, dl, MVT::i32),
getAL(CurDAG, dl), CurDAG->getRegister(0, MVT::i32)};
Add = CurDAG->getMachineNode(Opc, dl, MVT::i32, Ops);
}
}
if (Add) {
SDValue Ops2[] = {SDValue(Add, 0), CurDAG->getConstant(0, dl, MVT::i32)};
CurDAG->MorphNodeTo(N, ARMISD::CMPZ, CurDAG->getVTList(MVT::Glue), Ops2);
}
}
// Other cases are autogenerated.
break;
}
case ARMISD::CMOV: {
SDValue InFlag = N->getOperand(4);
if (InFlag.getOpcode() == ARMISD::CMPZ) {
bool SwitchEQNEToPLMI;
SelectCMPZ(InFlag.getNode(), SwitchEQNEToPLMI);
if (SwitchEQNEToPLMI) {
SDValue ARMcc = N->getOperand(2);
ARMCC::CondCodes CC =
(ARMCC::CondCodes)cast<ConstantSDNode>(ARMcc)->getZExtValue();
switch (CC) {
default: llvm_unreachable("CMPZ must be either NE or EQ!");
case ARMCC::NE:
CC = ARMCC::MI;
break;
case ARMCC::EQ:
CC = ARMCC::PL;
break;
}
SDValue NewARMcc = CurDAG->getConstant((unsigned)CC, dl, MVT::i32);
SDValue Ops[] = {N->getOperand(0), N->getOperand(1), NewARMcc,
N->getOperand(3), N->getOperand(4)};
CurDAG->MorphNodeTo(N, ARMISD::CMOV, N->getVTList(), Ops);
}
}
// Other cases are autogenerated.
break;
}
case ARMISD::VZIP: {
unsigned Opc = 0;
EVT VT = N->getValueType(0);
switch (VT.getSimpleVT().SimpleTy) {
default: return;
case MVT::v8i8: Opc = ARM::VZIPd8; break;
case MVT::v4f16:
case MVT::v4i16: Opc = ARM::VZIPd16; break;
case MVT::v2f32:
// vzip.32 Dd, Dm is a pseudo-instruction expanded to vtrn.32 Dd, Dm.
case MVT::v2i32: Opc = ARM::VTRNd32; break;
case MVT::v16i8: Opc = ARM::VZIPq8; break;
case MVT::v8f16:
case MVT::v8i16: Opc = ARM::VZIPq16; break;
case MVT::v4f32:
case MVT::v4i32: Opc = ARM::VZIPq32; break;
}
SDValue Pred = getAL(CurDAG, dl);
SDValue PredReg = CurDAG->getRegister(0, MVT::i32);
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), Pred, PredReg };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, VT, VT, Ops));
return;
}
case ARMISD::VUZP: {
unsigned Opc = 0;
EVT VT = N->getValueType(0);
switch (VT.getSimpleVT().SimpleTy) {
default: return;
case MVT::v8i8: Opc = ARM::VUZPd8; break;
case MVT::v4f16:
case MVT::v4i16: Opc = ARM::VUZPd16; break;
case MVT::v2f32:
// vuzp.32 Dd, Dm is a pseudo-instruction expanded to vtrn.32 Dd, Dm.
case MVT::v2i32: Opc = ARM::VTRNd32; break;
case MVT::v16i8: Opc = ARM::VUZPq8; break;
case MVT::v8f16:
case MVT::v8i16: Opc = ARM::VUZPq16; break;
case MVT::v4f32:
case MVT::v4i32: Opc = ARM::VUZPq32; break;
}
SDValue Pred = getAL(CurDAG, dl);
SDValue PredReg = CurDAG->getRegister(0, MVT::i32);
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), Pred, PredReg };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, VT, VT, Ops));
return;
}
case ARMISD::VTRN: {
unsigned Opc = 0;
EVT VT = N->getValueType(0);
switch (VT.getSimpleVT().SimpleTy) {
default: return;
case MVT::v8i8: Opc = ARM::VTRNd8; break;
case MVT::v4f16:
case MVT::v4i16: Opc = ARM::VTRNd16; break;
case MVT::v2f32:
case MVT::v2i32: Opc = ARM::VTRNd32; break;
case MVT::v16i8: Opc = ARM::VTRNq8; break;
case MVT::v8f16:
case MVT::v8i16: Opc = ARM::VTRNq16; break;
case MVT::v4f32:
case MVT::v4i32: Opc = ARM::VTRNq32; break;
}
SDValue Pred = getAL(CurDAG, dl);
SDValue PredReg = CurDAG->getRegister(0, MVT::i32);
SDValue Ops[] = { N->getOperand(0), N->getOperand(1), Pred, PredReg };
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, VT, VT, Ops));
return;
}
case ARMISD::BUILD_VECTOR: {
EVT VecVT = N->getValueType(0);
EVT EltVT = VecVT.getVectorElementType();
unsigned NumElts = VecVT.getVectorNumElements();
if (EltVT == MVT::f64) {
assert(NumElts == 2 && "unexpected type for BUILD_VECTOR");
ReplaceNode(
N, createDRegPairNode(VecVT, N->getOperand(0), N->getOperand(1)));
return;
}
assert(EltVT == MVT::f32 && "unexpected type for BUILD_VECTOR");
if (NumElts == 2) {
ReplaceNode(
N, createSRegPairNode(VecVT, N->getOperand(0), N->getOperand(1)));
return;
}
assert(NumElts == 4 && "unexpected type for BUILD_VECTOR");
ReplaceNode(N,
createQuadSRegsNode(VecVT, N->getOperand(0), N->getOperand(1),
N->getOperand(2), N->getOperand(3)));
return;
}
case ARMISD::VLD1DUP: {
static const uint16_t DOpcodes[] = { ARM::VLD1DUPd8, ARM::VLD1DUPd16,
ARM::VLD1DUPd32 };
static const uint16_t QOpcodes[] = { ARM::VLD1DUPq8, ARM::VLD1DUPq16,
ARM::VLD1DUPq32 };
SelectVLDDup(N, /* IsIntrinsic= */ false, false, 1, DOpcodes, QOpcodes);
return;
}
case ARMISD::VLD2DUP: {
static const uint16_t Opcodes[] = { ARM::VLD2DUPd8, ARM::VLD2DUPd16,
ARM::VLD2DUPd32 };
SelectVLDDup(N, /* IsIntrinsic= */ false, false, 2, Opcodes);
return;
}
case ARMISD::VLD3DUP: {
static const uint16_t Opcodes[] = { ARM::VLD3DUPd8Pseudo,
ARM::VLD3DUPd16Pseudo,
ARM::VLD3DUPd32Pseudo };
SelectVLDDup(N, /* IsIntrinsic= */ false, false, 3, Opcodes);
return;
}
case ARMISD::VLD4DUP: {
static const uint16_t Opcodes[] = { ARM::VLD4DUPd8Pseudo,
ARM::VLD4DUPd16Pseudo,
ARM::VLD4DUPd32Pseudo };
SelectVLDDup(N, /* IsIntrinsic= */ false, false, 4, Opcodes);
return;
}
case ARMISD::VLD1DUP_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD1DUPd8wb_fixed,
ARM::VLD1DUPd16wb_fixed,
ARM::VLD1DUPd32wb_fixed };
static const uint16_t QOpcodes[] = { ARM::VLD1DUPq8wb_fixed,
ARM::VLD1DUPq16wb_fixed,
ARM::VLD1DUPq32wb_fixed };
SelectVLDDup(N, /* IsIntrinsic= */ false, true, 1, DOpcodes, QOpcodes);
return;
}
case ARMISD::VLD2DUP_UPD: {
static const uint16_t Opcodes[] = { ARM::VLD2DUPd8wb_fixed,
ARM::VLD2DUPd16wb_fixed,
ARM::VLD2DUPd32wb_fixed };
SelectVLDDup(N, /* IsIntrinsic= */ false, true, 2, Opcodes);
return;
}
case ARMISD::VLD3DUP_UPD: {
static const uint16_t Opcodes[] = { ARM::VLD3DUPd8Pseudo_UPD,
ARM::VLD3DUPd16Pseudo_UPD,
ARM::VLD3DUPd32Pseudo_UPD };
SelectVLDDup(N, /* IsIntrinsic= */ false, true, 3, Opcodes);
return;
}
case ARMISD::VLD4DUP_UPD: {
static const uint16_t Opcodes[] = { ARM::VLD4DUPd8Pseudo_UPD,
ARM::VLD4DUPd16Pseudo_UPD,
ARM::VLD4DUPd32Pseudo_UPD };
SelectVLDDup(N, /* IsIntrinsic= */ false, true, 4, Opcodes);
return;
}
case ARMISD::VLD1_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD1d8wb_fixed,
ARM::VLD1d16wb_fixed,
ARM::VLD1d32wb_fixed,
ARM::VLD1d64wb_fixed };
static const uint16_t QOpcodes[] = { ARM::VLD1q8wb_fixed,
ARM::VLD1q16wb_fixed,
ARM::VLD1q32wb_fixed,
ARM::VLD1q64wb_fixed };
SelectVLD(N, true, 1, DOpcodes, QOpcodes, nullptr);
return;
}
case ARMISD::VLD2_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD2d8wb_fixed,
ARM::VLD2d16wb_fixed,
ARM::VLD2d32wb_fixed,
ARM::VLD1q64wb_fixed};
static const uint16_t QOpcodes[] = { ARM::VLD2q8PseudoWB_fixed,
ARM::VLD2q16PseudoWB_fixed,
ARM::VLD2q32PseudoWB_fixed };
SelectVLD(N, true, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case ARMISD::VLD3_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD3d8Pseudo_UPD,
ARM::VLD3d16Pseudo_UPD,
ARM::VLD3d32Pseudo_UPD,
ARM::VLD1d64TPseudoWB_fixed};
static const uint16_t QOpcodes0[] = { ARM::VLD3q8Pseudo_UPD,
ARM::VLD3q16Pseudo_UPD,
ARM::VLD3q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD3q8oddPseudo_UPD,
ARM::VLD3q16oddPseudo_UPD,
ARM::VLD3q32oddPseudo_UPD };
SelectVLD(N, true, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case ARMISD::VLD4_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD4d8Pseudo_UPD,
ARM::VLD4d16Pseudo_UPD,
ARM::VLD4d32Pseudo_UPD,
ARM::VLD1d64QPseudoWB_fixed};
static const uint16_t QOpcodes0[] = { ARM::VLD4q8Pseudo_UPD,
ARM::VLD4q16Pseudo_UPD,
ARM::VLD4q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD4q8oddPseudo_UPD,
ARM::VLD4q16oddPseudo_UPD,
ARM::VLD4q32oddPseudo_UPD };
SelectVLD(N, true, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case ARMISD::VLD2LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD2LNd8Pseudo_UPD,
ARM::VLD2LNd16Pseudo_UPD,
ARM::VLD2LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VLD2LNq16Pseudo_UPD,
ARM::VLD2LNq32Pseudo_UPD };
SelectVLDSTLane(N, true, true, 2, DOpcodes, QOpcodes);
return;
}
case ARMISD::VLD3LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD3LNd8Pseudo_UPD,
ARM::VLD3LNd16Pseudo_UPD,
ARM::VLD3LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VLD3LNq16Pseudo_UPD,
ARM::VLD3LNq32Pseudo_UPD };
SelectVLDSTLane(N, true, true, 3, DOpcodes, QOpcodes);
return;
}
case ARMISD::VLD4LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VLD4LNd8Pseudo_UPD,
ARM::VLD4LNd16Pseudo_UPD,
ARM::VLD4LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VLD4LNq16Pseudo_UPD,
ARM::VLD4LNq32Pseudo_UPD };
SelectVLDSTLane(N, true, true, 4, DOpcodes, QOpcodes);
return;
}
case ARMISD::VST1_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST1d8wb_fixed,
ARM::VST1d16wb_fixed,
ARM::VST1d32wb_fixed,
ARM::VST1d64wb_fixed };
static const uint16_t QOpcodes[] = { ARM::VST1q8wb_fixed,
ARM::VST1q16wb_fixed,
ARM::VST1q32wb_fixed,
ARM::VST1q64wb_fixed };
SelectVST(N, true, 1, DOpcodes, QOpcodes, nullptr);
return;
}
case ARMISD::VST2_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST2d8wb_fixed,
ARM::VST2d16wb_fixed,
ARM::VST2d32wb_fixed,
ARM::VST1q64wb_fixed};
static const uint16_t QOpcodes[] = { ARM::VST2q8PseudoWB_fixed,
ARM::VST2q16PseudoWB_fixed,
ARM::VST2q32PseudoWB_fixed };
SelectVST(N, true, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case ARMISD::VST3_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST3d8Pseudo_UPD,
ARM::VST3d16Pseudo_UPD,
ARM::VST3d32Pseudo_UPD,
ARM::VST1d64TPseudoWB_fixed};
static const uint16_t QOpcodes0[] = { ARM::VST3q8Pseudo_UPD,
ARM::VST3q16Pseudo_UPD,
ARM::VST3q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST3q8oddPseudo_UPD,
ARM::VST3q16oddPseudo_UPD,
ARM::VST3q32oddPseudo_UPD };
SelectVST(N, true, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case ARMISD::VST4_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST4d8Pseudo_UPD,
ARM::VST4d16Pseudo_UPD,
ARM::VST4d32Pseudo_UPD,
ARM::VST1d64QPseudoWB_fixed};
static const uint16_t QOpcodes0[] = { ARM::VST4q8Pseudo_UPD,
ARM::VST4q16Pseudo_UPD,
ARM::VST4q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST4q8oddPseudo_UPD,
ARM::VST4q16oddPseudo_UPD,
ARM::VST4q32oddPseudo_UPD };
SelectVST(N, true, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case ARMISD::VST2LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST2LNd8Pseudo_UPD,
ARM::VST2LNd16Pseudo_UPD,
ARM::VST2LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VST2LNq16Pseudo_UPD,
ARM::VST2LNq32Pseudo_UPD };
SelectVLDSTLane(N, false, true, 2, DOpcodes, QOpcodes);
return;
}
case ARMISD::VST3LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST3LNd8Pseudo_UPD,
ARM::VST3LNd16Pseudo_UPD,
ARM::VST3LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VST3LNq16Pseudo_UPD,
ARM::VST3LNq32Pseudo_UPD };
SelectVLDSTLane(N, false, true, 3, DOpcodes, QOpcodes);
return;
}
case ARMISD::VST4LN_UPD: {
static const uint16_t DOpcodes[] = { ARM::VST4LNd8Pseudo_UPD,
ARM::VST4LNd16Pseudo_UPD,
ARM::VST4LNd32Pseudo_UPD };
static const uint16_t QOpcodes[] = { ARM::VST4LNq16Pseudo_UPD,
ARM::VST4LNq32Pseudo_UPD };
SelectVLDSTLane(N, false, true, 4, DOpcodes, QOpcodes);
return;
}
case ISD::INTRINSIC_VOID:
case ISD::INTRINSIC_W_CHAIN: {
unsigned IntNo = cast<ConstantSDNode>(N->getOperand(1))->getZExtValue();
switch (IntNo) {
default:
break;
case Intrinsic::arm_mrrc:
case Intrinsic::arm_mrrc2: {
SDLoc dl(N);
SDValue Chain = N->getOperand(0);
unsigned Opc;
if (Subtarget->isThumb())
Opc = (IntNo == Intrinsic::arm_mrrc ? ARM::t2MRRC : ARM::t2MRRC2);
else
Opc = (IntNo == Intrinsic::arm_mrrc ? ARM::MRRC : ARM::MRRC2);
SmallVector<SDValue, 5> Ops;
Ops.push_back(getI32Imm(cast<ConstantSDNode>(N->getOperand(2))->getZExtValue(), dl)); /* coproc */
Ops.push_back(getI32Imm(cast<ConstantSDNode>(N->getOperand(3))->getZExtValue(), dl)); /* opc */
Ops.push_back(getI32Imm(cast<ConstantSDNode>(N->getOperand(4))->getZExtValue(), dl)); /* CRm */
// The mrrc2 instruction in ARM doesn't allow predicates, the top 4 bits of the encoded
// instruction will always be '1111' but it is possible in assembly language to specify
// AL as a predicate to mrrc2 but it doesn't make any difference to the encoded instruction.
if (Opc != ARM::MRRC2) {
Ops.push_back(getAL(CurDAG, dl));
Ops.push_back(CurDAG->getRegister(0, MVT::i32));
}
Ops.push_back(Chain);
// Writes to two registers.
const EVT RetType[] = {MVT::i32, MVT::i32, MVT::Other};
ReplaceNode(N, CurDAG->getMachineNode(Opc, dl, RetType, Ops));
return;
}
case Intrinsic::arm_ldaexd:
case Intrinsic::arm_ldrexd: {
SDLoc dl(N);
SDValue Chain = N->getOperand(0);
SDValue MemAddr = N->getOperand(2);
bool isThumb = Subtarget->isThumb() && Subtarget->hasV8MBaselineOps();
bool IsAcquire = IntNo == Intrinsic::arm_ldaexd;
unsigned NewOpc = isThumb ? (IsAcquire ? ARM::t2LDAEXD : ARM::t2LDREXD)
: (IsAcquire ? ARM::LDAEXD : ARM::LDREXD);
// arm_ldrexd returns a i64 value in {i32, i32}
std::vector<EVT> ResTys;
if (isThumb) {
ResTys.push_back(MVT::i32);
ResTys.push_back(MVT::i32);
} else
ResTys.push_back(MVT::Untyped);
ResTys.push_back(MVT::Other);
// Place arguments in the right order.
SDValue Ops[] = {MemAddr, getAL(CurDAG, dl),
CurDAG->getRegister(0, MVT::i32), Chain};
SDNode *Ld = CurDAG->getMachineNode(NewOpc, dl, ResTys, Ops);
// Transfer memoperands.
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(Ld), {MemOp});
// Remap uses.
SDValue OutChain = isThumb ? SDValue(Ld, 2) : SDValue(Ld, 1);
if (!SDValue(N, 0).use_empty()) {
SDValue Result;
if (isThumb)
Result = SDValue(Ld, 0);
else {
SDValue SubRegIdx =
CurDAG->getTargetConstant(ARM::gsub_0, dl, MVT::i32);
SDNode *ResNode = CurDAG->getMachineNode(TargetOpcode::EXTRACT_SUBREG,
dl, MVT::i32, SDValue(Ld, 0), SubRegIdx);
Result = SDValue(ResNode,0);
}
ReplaceUses(SDValue(N, 0), Result);
}
if (!SDValue(N, 1).use_empty()) {
SDValue Result;
if (isThumb)
Result = SDValue(Ld, 1);
else {
SDValue SubRegIdx =
CurDAG->getTargetConstant(ARM::gsub_1, dl, MVT::i32);
SDNode *ResNode = CurDAG->getMachineNode(TargetOpcode::EXTRACT_SUBREG,
dl, MVT::i32, SDValue(Ld, 0), SubRegIdx);
Result = SDValue(ResNode,0);
}
ReplaceUses(SDValue(N, 1), Result);
}
ReplaceUses(SDValue(N, 2), OutChain);
CurDAG->RemoveDeadNode(N);
return;
}
case Intrinsic::arm_stlexd:
case Intrinsic::arm_strexd: {
SDLoc dl(N);
SDValue Chain = N->getOperand(0);
SDValue Val0 = N->getOperand(2);
SDValue Val1 = N->getOperand(3);
SDValue MemAddr = N->getOperand(4);
// Store exclusive double return a i32 value which is the return status
// of the issued store.
const EVT ResTys[] = {MVT::i32, MVT::Other};
bool isThumb = Subtarget->isThumb() && Subtarget->hasThumb2();
// Place arguments in the right order.
SmallVector<SDValue, 7> Ops;
if (isThumb) {
Ops.push_back(Val0);
Ops.push_back(Val1);
} else
// arm_strexd uses GPRPair.
Ops.push_back(SDValue(createGPRPairNode(MVT::Untyped, Val0, Val1), 0));
Ops.push_back(MemAddr);
Ops.push_back(getAL(CurDAG, dl));
Ops.push_back(CurDAG->getRegister(0, MVT::i32));
Ops.push_back(Chain);
bool IsRelease = IntNo == Intrinsic::arm_stlexd;
unsigned NewOpc = isThumb ? (IsRelease ? ARM::t2STLEXD : ARM::t2STREXD)
: (IsRelease ? ARM::STLEXD : ARM::STREXD);
SDNode *St = CurDAG->getMachineNode(NewOpc, dl, ResTys, Ops);
// Transfer memoperands.
MachineMemOperand *MemOp = cast<MemIntrinsicSDNode>(N)->getMemOperand();
CurDAG->setNodeMemRefs(cast<MachineSDNode>(St), {MemOp});
ReplaceNode(N, St);
return;
}
case Intrinsic::arm_neon_vld1: {
static const uint16_t DOpcodes[] = { ARM::VLD1d8, ARM::VLD1d16,
ARM::VLD1d32, ARM::VLD1d64 };
static const uint16_t QOpcodes[] = { ARM::VLD1q8, ARM::VLD1q16,
ARM::VLD1q32, ARM::VLD1q64};
SelectVLD(N, false, 1, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vld1x2: {
static const uint16_t DOpcodes[] = { ARM::VLD1q8, ARM::VLD1q16,
ARM::VLD1q32, ARM::VLD1q64 };
static const uint16_t QOpcodes[] = { ARM::VLD1d8QPseudo,
ARM::VLD1d16QPseudo,
ARM::VLD1d32QPseudo,
ARM::VLD1d64QPseudo };
SelectVLD(N, false, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vld1x3: {
static const uint16_t DOpcodes[] = { ARM::VLD1d8TPseudo,
ARM::VLD1d16TPseudo,
ARM::VLD1d32TPseudo,
ARM::VLD1d64TPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD1q8LowTPseudo_UPD,
ARM::VLD1q16LowTPseudo_UPD,
ARM::VLD1q32LowTPseudo_UPD,
ARM::VLD1q64LowTPseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD1q8HighTPseudo,
ARM::VLD1q16HighTPseudo,
ARM::VLD1q32HighTPseudo,
ARM::VLD1q64HighTPseudo };
SelectVLD(N, false, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld1x4: {
static const uint16_t DOpcodes[] = { ARM::VLD1d8QPseudo,
ARM::VLD1d16QPseudo,
ARM::VLD1d32QPseudo,
ARM::VLD1d64QPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD1q8LowQPseudo_UPD,
ARM::VLD1q16LowQPseudo_UPD,
ARM::VLD1q32LowQPseudo_UPD,
ARM::VLD1q64LowQPseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD1q8HighQPseudo,
ARM::VLD1q16HighQPseudo,
ARM::VLD1q32HighQPseudo,
ARM::VLD1q64HighQPseudo };
SelectVLD(N, false, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld2: {
static const uint16_t DOpcodes[] = { ARM::VLD2d8, ARM::VLD2d16,
ARM::VLD2d32, ARM::VLD1q64 };
static const uint16_t QOpcodes[] = { ARM::VLD2q8Pseudo, ARM::VLD2q16Pseudo,
ARM::VLD2q32Pseudo };
SelectVLD(N, false, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vld3: {
static const uint16_t DOpcodes[] = { ARM::VLD3d8Pseudo,
ARM::VLD3d16Pseudo,
ARM::VLD3d32Pseudo,
ARM::VLD1d64TPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD3q8Pseudo_UPD,
ARM::VLD3q16Pseudo_UPD,
ARM::VLD3q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD3q8oddPseudo,
ARM::VLD3q16oddPseudo,
ARM::VLD3q32oddPseudo };
SelectVLD(N, false, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld4: {
static const uint16_t DOpcodes[] = { ARM::VLD4d8Pseudo,
ARM::VLD4d16Pseudo,
ARM::VLD4d32Pseudo,
ARM::VLD1d64QPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD4q8Pseudo_UPD,
ARM::VLD4q16Pseudo_UPD,
ARM::VLD4q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VLD4q8oddPseudo,
ARM::VLD4q16oddPseudo,
ARM::VLD4q32oddPseudo };
SelectVLD(N, false, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld2dup: {
static const uint16_t DOpcodes[] = { ARM::VLD2DUPd8, ARM::VLD2DUPd16,
ARM::VLD2DUPd32, ARM::VLD1q64 };
static const uint16_t QOpcodes0[] = { ARM::VLD2DUPq8EvenPseudo,
ARM::VLD2DUPq16EvenPseudo,
ARM::VLD2DUPq32EvenPseudo };
static const uint16_t QOpcodes1[] = { ARM::VLD2DUPq8OddPseudo,
ARM::VLD2DUPq16OddPseudo,
ARM::VLD2DUPq32OddPseudo };
SelectVLDDup(N, /* IsIntrinsic= */ true, false, 2,
DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld3dup: {
static const uint16_t DOpcodes[] = { ARM::VLD3DUPd8Pseudo,
ARM::VLD3DUPd16Pseudo,
ARM::VLD3DUPd32Pseudo,
ARM::VLD1d64TPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD3DUPq8EvenPseudo,
ARM::VLD3DUPq16EvenPseudo,
ARM::VLD3DUPq32EvenPseudo };
static const uint16_t QOpcodes1[] = { ARM::VLD3DUPq8OddPseudo,
ARM::VLD3DUPq16OddPseudo,
ARM::VLD3DUPq32OddPseudo };
SelectVLDDup(N, /* IsIntrinsic= */ true, false, 3,
DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld4dup: {
static const uint16_t DOpcodes[] = { ARM::VLD4DUPd8Pseudo,
ARM::VLD4DUPd16Pseudo,
ARM::VLD4DUPd32Pseudo,
ARM::VLD1d64QPseudo };
static const uint16_t QOpcodes0[] = { ARM::VLD4DUPq8EvenPseudo,
ARM::VLD4DUPq16EvenPseudo,
ARM::VLD4DUPq32EvenPseudo };
static const uint16_t QOpcodes1[] = { ARM::VLD4DUPq8OddPseudo,
ARM::VLD4DUPq16OddPseudo,
ARM::VLD4DUPq32OddPseudo };
SelectVLDDup(N, /* IsIntrinsic= */ true, false, 4,
DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vld2lane: {
static const uint16_t DOpcodes[] = { ARM::VLD2LNd8Pseudo,
ARM::VLD2LNd16Pseudo,
ARM::VLD2LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VLD2LNq16Pseudo,
ARM::VLD2LNq32Pseudo };
SelectVLDSTLane(N, true, false, 2, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_neon_vld3lane: {
static const uint16_t DOpcodes[] = { ARM::VLD3LNd8Pseudo,
ARM::VLD3LNd16Pseudo,
ARM::VLD3LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VLD3LNq16Pseudo,
ARM::VLD3LNq32Pseudo };
SelectVLDSTLane(N, true, false, 3, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_neon_vld4lane: {
static const uint16_t DOpcodes[] = { ARM::VLD4LNd8Pseudo,
ARM::VLD4LNd16Pseudo,
ARM::VLD4LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VLD4LNq16Pseudo,
ARM::VLD4LNq32Pseudo };
SelectVLDSTLane(N, true, false, 4, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_neon_vst1: {
static const uint16_t DOpcodes[] = { ARM::VST1d8, ARM::VST1d16,
ARM::VST1d32, ARM::VST1d64 };
static const uint16_t QOpcodes[] = { ARM::VST1q8, ARM::VST1q16,
ARM::VST1q32, ARM::VST1q64 };
SelectVST(N, false, 1, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vst1x2: {
static const uint16_t DOpcodes[] = { ARM::VST1q8, ARM::VST1q16,
ARM::VST1q32, ARM::VST1q64 };
static const uint16_t QOpcodes[] = { ARM::VST1d8QPseudo,
ARM::VST1d16QPseudo,
ARM::VST1d32QPseudo,
ARM::VST1d64QPseudo };
SelectVST(N, false, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vst1x3: {
static const uint16_t DOpcodes[] = { ARM::VST1d8TPseudo,
ARM::VST1d16TPseudo,
ARM::VST1d32TPseudo,
ARM::VST1d64TPseudo };
static const uint16_t QOpcodes0[] = { ARM::VST1q8LowTPseudo_UPD,
ARM::VST1q16LowTPseudo_UPD,
ARM::VST1q32LowTPseudo_UPD,
ARM::VST1q64LowTPseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST1q8HighTPseudo,
ARM::VST1q16HighTPseudo,
ARM::VST1q32HighTPseudo,
ARM::VST1q64HighTPseudo };
SelectVST(N, false, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vst1x4: {
static const uint16_t DOpcodes[] = { ARM::VST1d8QPseudo,
ARM::VST1d16QPseudo,
ARM::VST1d32QPseudo,
ARM::VST1d64QPseudo };
static const uint16_t QOpcodes0[] = { ARM::VST1q8LowQPseudo_UPD,
ARM::VST1q16LowQPseudo_UPD,
ARM::VST1q32LowQPseudo_UPD,
ARM::VST1q64LowQPseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST1q8HighQPseudo,
ARM::VST1q16HighQPseudo,
ARM::VST1q32HighQPseudo,
ARM::VST1q64HighQPseudo };
SelectVST(N, false, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vst2: {
static const uint16_t DOpcodes[] = { ARM::VST2d8, ARM::VST2d16,
ARM::VST2d32, ARM::VST1q64 };
static const uint16_t QOpcodes[] = { ARM::VST2q8Pseudo, ARM::VST2q16Pseudo,
ARM::VST2q32Pseudo };
SelectVST(N, false, 2, DOpcodes, QOpcodes, nullptr);
return;
}
case Intrinsic::arm_neon_vst3: {
static const uint16_t DOpcodes[] = { ARM::VST3d8Pseudo,
ARM::VST3d16Pseudo,
ARM::VST3d32Pseudo,
ARM::VST1d64TPseudo };
static const uint16_t QOpcodes0[] = { ARM::VST3q8Pseudo_UPD,
ARM::VST3q16Pseudo_UPD,
ARM::VST3q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST3q8oddPseudo,
ARM::VST3q16oddPseudo,
ARM::VST3q32oddPseudo };
SelectVST(N, false, 3, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vst4: {
static const uint16_t DOpcodes[] = { ARM::VST4d8Pseudo,
ARM::VST4d16Pseudo,
ARM::VST4d32Pseudo,
ARM::VST1d64QPseudo };
static const uint16_t QOpcodes0[] = { ARM::VST4q8Pseudo_UPD,
ARM::VST4q16Pseudo_UPD,
ARM::VST4q32Pseudo_UPD };
static const uint16_t QOpcodes1[] = { ARM::VST4q8oddPseudo,
ARM::VST4q16oddPseudo,
ARM::VST4q32oddPseudo };
SelectVST(N, false, 4, DOpcodes, QOpcodes0, QOpcodes1);
return;
}
case Intrinsic::arm_neon_vst2lane: {
static const uint16_t DOpcodes[] = { ARM::VST2LNd8Pseudo,
ARM::VST2LNd16Pseudo,
ARM::VST2LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VST2LNq16Pseudo,
ARM::VST2LNq32Pseudo };
SelectVLDSTLane(N, false, false, 2, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_neon_vst3lane: {
static const uint16_t DOpcodes[] = { ARM::VST3LNd8Pseudo,
ARM::VST3LNd16Pseudo,
ARM::VST3LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VST3LNq16Pseudo,
ARM::VST3LNq32Pseudo };
SelectVLDSTLane(N, false, false, 3, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_neon_vst4lane: {
static const uint16_t DOpcodes[] = { ARM::VST4LNd8Pseudo,
ARM::VST4LNd16Pseudo,
ARM::VST4LNd32Pseudo };
static const uint16_t QOpcodes[] = { ARM::VST4LNq16Pseudo,
ARM::VST4LNq32Pseudo };
SelectVLDSTLane(N, false, false, 4, DOpcodes, QOpcodes);
return;
}
case Intrinsic::arm_mve_vldr_gather_base_wb:
case Intrinsic::arm_mve_vldr_gather_base_wb_predicated: {
static const uint16_t Opcodes[] = {ARM::MVE_VLDRWU32_qi_pre,
ARM::MVE_VLDRDU64_qi_pre};
SelectMVE_WB(N, Opcodes,
IntNo == Intrinsic::arm_mve_vldr_gather_base_wb_predicated);
return;
}
case Intrinsic::arm_mve_vld2q: {
static const uint16_t Opcodes8[] = {ARM::MVE_VLD20_8, ARM::MVE_VLD21_8};
static const uint16_t Opcodes16[] = {ARM::MVE_VLD20_16,
ARM::MVE_VLD21_16};
static const uint16_t Opcodes32[] = {ARM::MVE_VLD20_32,
ARM::MVE_VLD21_32};
static const uint16_t *const Opcodes[] = {Opcodes8, Opcodes16, Opcodes32};
SelectMVE_VLD(N, 2, Opcodes);
return;
}
case Intrinsic::arm_mve_vld4q: {
static const uint16_t Opcodes8[] = {ARM::MVE_VLD40_8, ARM::MVE_VLD41_8,
ARM::MVE_VLD42_8, ARM::MVE_VLD43_8};
static const uint16_t Opcodes16[] = {ARM::MVE_VLD40_16, ARM::MVE_VLD41_16,
ARM::MVE_VLD42_16,
ARM::MVE_VLD43_16};
static const uint16_t Opcodes32[] = {ARM::MVE_VLD40_32, ARM::MVE_VLD41_32,
ARM::MVE_VLD42_32,
ARM::MVE_VLD43_32};
static const uint16_t *const Opcodes[] = {Opcodes8, Opcodes16, Opcodes32};
SelectMVE_VLD(N, 4, Opcodes);
return;
}
}
break;
}
case ISD::INTRINSIC_WO_CHAIN: {
unsigned IntNo = cast<ConstantSDNode>(N->getOperand(0))->getZExtValue();
switch (IntNo) {
default:
break;
case Intrinsic::arm_mve_urshrl:
SelectMVE_LongShift(N, ARM::MVE_URSHRL, true, false);
return;
case Intrinsic::arm_mve_uqshll:
SelectMVE_LongShift(N, ARM::MVE_UQSHLL, true, false);
return;
case Intrinsic::arm_mve_srshrl:
SelectMVE_LongShift(N, ARM::MVE_SRSHRL, true, false);
return;
case Intrinsic::arm_mve_sqshll:
SelectMVE_LongShift(N, ARM::MVE_SQSHLL, true, false);
return;
case Intrinsic::arm_mve_uqrshll:
SelectMVE_LongShift(N, ARM::MVE_UQRSHLL, false, true);
return;
case Intrinsic::arm_mve_sqrshrl:
SelectMVE_LongShift(N, ARM::MVE_SQRSHRL, false, true);
return;
case Intrinsic::arm_mve_lsll:
SelectMVE_LongShift(N, ARM::MVE_LSLLr, false, false);
return;
case Intrinsic::arm_mve_asrl:
SelectMVE_LongShift(N, ARM::MVE_ASRLr, false, false);
return;
case Intrinsic::arm_mve_vadc:
case Intrinsic::arm_mve_vadc_predicated:
SelectMVE_VADCSBC(N, ARM::MVE_VADC, ARM::MVE_VADCI, true,
IntNo == Intrinsic::arm_mve_vadc_predicated);
return;
case Intrinsic::arm_mve_vmlldava:
case Intrinsic::arm_mve_vmlldava_predicated: {
static const uint16_t OpcodesU[] = {
ARM::MVE_VMLALDAVu16, ARM::MVE_VMLALDAVu32,
ARM::MVE_VMLALDAVau16, ARM::MVE_VMLALDAVau32,
};
static const uint16_t OpcodesS[] = {
ARM::MVE_VMLALDAVs16, ARM::MVE_VMLALDAVs32,
ARM::MVE_VMLALDAVas16, ARM::MVE_VMLALDAVas32,
ARM::MVE_VMLALDAVxs16, ARM::MVE_VMLALDAVxs32,
ARM::MVE_VMLALDAVaxs16, ARM::MVE_VMLALDAVaxs32,
ARM::MVE_VMLSLDAVs16, ARM::MVE_VMLSLDAVs32,
ARM::MVE_VMLSLDAVas16, ARM::MVE_VMLSLDAVas32,
ARM::MVE_VMLSLDAVxs16, ARM::MVE_VMLSLDAVxs32,
ARM::MVE_VMLSLDAVaxs16, ARM::MVE_VMLSLDAVaxs32,
};
SelectMVE_VMLLDAV(N, IntNo == Intrinsic::arm_mve_vmlldava_predicated,
OpcodesS, OpcodesU);
return;
}
case Intrinsic::arm_mve_vrmlldavha:
case Intrinsic::arm_mve_vrmlldavha_predicated: {
static const uint16_t OpcodesU[] = {
ARM::MVE_VRMLALDAVHu32, ARM::MVE_VRMLALDAVHau32,
};
static const uint16_t OpcodesS[] = {
ARM::MVE_VRMLALDAVHs32, ARM::MVE_VRMLALDAVHas32,
ARM::MVE_VRMLALDAVHxs32, ARM::MVE_VRMLALDAVHaxs32,
ARM::MVE_VRMLSLDAVHs32, ARM::MVE_VRMLSLDAVHas32,
ARM::MVE_VRMLSLDAVHxs32, ARM::MVE_VRMLSLDAVHaxs32,
};
SelectMVE_VRMLLDAVH(N, IntNo == Intrinsic::arm_mve_vrmlldavha_predicated,
OpcodesS, OpcodesU);
return;
}
}
break;
}
case ISD::ATOMIC_CMP_SWAP:
SelectCMP_SWAP(N);
return;
}
SelectCode(N);
}
// Inspect a register string of the form
// cp<coprocessor>:<opc1>:c<CRn>:c<CRm>:<opc2> (32bit) or
// cp<coprocessor>:<opc1>:c<CRm> (64bit) inspect the fields of the string
// and obtain the integer operands from them, adding these operands to the
// provided vector.
static void getIntOperandsFromRegisterString(StringRef RegString,
SelectionDAG *CurDAG,
const SDLoc &DL,
std::vector<SDValue> &Ops) {
SmallVector<StringRef, 5> Fields;
RegString.split(Fields, ':');
if (Fields.size() > 1) {
bool AllIntFields = true;
for (StringRef Field : Fields) {
// Need to trim out leading 'cp' characters and get the integer field.
unsigned IntField;
AllIntFields &= !Field.trim("CPcp").getAsInteger(10, IntField);
Ops.push_back(CurDAG->getTargetConstant(IntField, DL, MVT::i32));
}
assert(AllIntFields &&
"Unexpected non-integer value in special register string.");
}
}
// Maps a Banked Register string to its mask value. The mask value returned is
// for use in the MRSbanked / MSRbanked instruction nodes as the Banked Register
// mask operand, which expresses which register is to be used, e.g. r8, and in
// which mode it is to be used, e.g. usr. Returns -1 to signify that the string
// was invalid.
static inline int getBankedRegisterMask(StringRef RegString) {
auto TheReg = ARMBankedReg::lookupBankedRegByName(RegString.lower());
if (!TheReg)
return -1;
return TheReg->Encoding;
}
// The flags here are common to those allowed for apsr in the A class cores and
// those allowed for the special registers in the M class cores. Returns a
// value representing which flags were present, -1 if invalid.
static inline int getMClassFlagsMask(StringRef Flags) {
return StringSwitch<int>(Flags)
.Case("", 0x2) // no flags means nzcvq for psr registers, and 0x2 is
// correct when flags are not permitted
.Case("g", 0x1)
.Case("nzcvq", 0x2)
.Case("nzcvqg", 0x3)
.Default(-1);
}
// Maps MClass special registers string to its value for use in the
// t2MRS_M/t2MSR_M instruction nodes as the SYSm value operand.
// Returns -1 to signify that the string was invalid.
static int getMClassRegisterMask(StringRef Reg, const ARMSubtarget *Subtarget) {
auto TheReg = ARMSysReg::lookupMClassSysRegByName(Reg);
const FeatureBitset &FeatureBits = Subtarget->getFeatureBits();
if (!TheReg || !TheReg->hasRequiredFeatures(FeatureBits))
return -1;
return (int)(TheReg->Encoding & 0xFFF); // SYSm value
}
static int getARClassRegisterMask(StringRef Reg, StringRef Flags) {
// The mask operand contains the special register (R Bit) in bit 4, whether
// the register is spsr (R bit is 1) or one of cpsr/apsr (R bit is 0), and
// bits 3-0 contains the fields to be accessed in the special register, set by
// the flags provided with the register.
int Mask = 0;
if (Reg == "apsr") {
// The flags permitted for apsr are the same flags that are allowed in
// M class registers. We get the flag value and then shift the flags into
// the correct place to combine with the mask.
Mask = getMClassFlagsMask(Flags);
if (Mask == -1)
return -1;
return Mask << 2;
}
if (Reg != "cpsr" && Reg != "spsr") {
return -1;
}
// This is the same as if the flags were "fc"
if (Flags.empty() || Flags == "all")
return Mask | 0x9;