1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
|
//===-- AArch64Subtarget.cpp - AArch64 Subtarget Information ----*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the AArch64 specific subclass of TargetSubtarget.
//
//===----------------------------------------------------------------------===//
#include "AArch64InstrInfo.h"
#include "AArch64Subtarget.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/CodeGen/MachineScheduler.h"
#include "llvm/IR/GlobalValue.h"
#include "llvm/Support/TargetRegistry.h"
using namespace llvm;
#define DEBUG_TYPE "aarch64-subtarget"
#define GET_SUBTARGETINFO_CTOR
#define GET_SUBTARGETINFO_TARGET_DESC
#include "AArch64GenSubtargetInfo.inc"
static cl::opt<bool>
EnableEarlyIfConvert("aarch64-early-ifcvt", cl::desc("Enable the early if "
"converter pass"), cl::init(true), cl::Hidden);
AArch64Subtarget::AArch64Subtarget(const std::string &TT,
const std::string &CPU,
const std::string &FS, bool LittleEndian)
: AArch64GenSubtargetInfo(TT, CPU, FS), ARMProcFamily(Others),
HasFPARMv8(false), HasNEON(false), HasCrypto(false), HasCRC(false),
HasZeroCycleRegMove(false), HasZeroCycleZeroing(false), CPUString(CPU),
TargetTriple(TT), IsLittleEndian(LittleEndian) {
// Determine default and user-specified characteristics
if (CPUString.empty())
CPUString = "generic";
ParseSubtargetFeatures(CPUString, FS);
}
/// ClassifyGlobalReference - Find the target operand flags that describe
/// how a global value should be referenced for the current subtarget.
unsigned char
AArch64Subtarget::ClassifyGlobalReference(const GlobalValue *GV,
const TargetMachine &TM) const {
// Determine whether this is a reference to a definition or a declaration.
// Materializable GVs (in JIT lazy compilation mode) do not require an extra
// load from stub.
bool isDecl = GV->hasAvailableExternallyLinkage();
if (GV->isDeclaration() && !GV->isMaterializable())
isDecl = true;
// MachO large model always goes via a GOT, simply to get a single 8-byte
// absolute relocation on all global addresses.
if (TM.getCodeModel() == CodeModel::Large && isTargetMachO())
return AArch64II::MO_GOT;
// The small code mode's direct accesses use ADRP, which cannot necessarily
// produce the value 0 (if the code is above 4GB). Therefore they must use the
// GOT.
if (TM.getCodeModel() == CodeModel::Small && GV->isWeakForLinker() && isDecl)
return AArch64II::MO_GOT;
// If symbol visibility is hidden, the extra load is not needed if
// the symbol is definitely defined in the current translation unit.
// The handling of non-hidden symbols in PIC mode is rather target-dependent:
// + On MachO, if the symbol is defined in this module the GOT can be
// skipped.
// + On ELF, the R_AARCH64_COPY relocation means that even symbols actually
// defined could end up in unexpected places. Use a GOT.
if (TM.getRelocationModel() != Reloc::Static && GV->hasDefaultVisibility()) {
if (isTargetMachO())
return (isDecl || GV->isWeakForLinker()) ? AArch64II::MO_GOT
: AArch64II::MO_NO_FLAG;
else
// No need to go through the GOT for local symbols on ELF.
return GV->hasLocalLinkage() ? AArch64II::MO_NO_FLAG : AArch64II::MO_GOT;
}
return AArch64II::MO_NO_FLAG;
}
/// This function returns the name of a function which has an interface
/// like the non-standard bzero function, if such a function exists on
/// the current subtarget and it is considered prefereable over
/// memset with zero passed as the second argument. Otherwise it
/// returns null.
const char *AArch64Subtarget::getBZeroEntry() const {
// Prefer bzero on Darwin only.
if(isTargetDarwin())
return "bzero";
return nullptr;
}
void AArch64Subtarget::overrideSchedPolicy(MachineSchedPolicy &Policy,
MachineInstr *begin, MachineInstr *end,
unsigned NumRegionInstrs) const {
// LNT run (at least on Cyclone) showed reasonably significant gains for
// bi-directional scheduling. 253.perlbmk.
Policy.OnlyTopDown = false;
Policy.OnlyBottomUp = false;
}
bool AArch64Subtarget::enableEarlyIfConversion() const {
return EnableEarlyIfConvert;
}
|
