LLVM  13.0.0git
AArch64ExternalSymbolizer.cpp
Go to the documentation of this file.
1 //===- AArch64ExternalSymbolizer.cpp - Symbolizer for AArch64 ---*- C++ -*-===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 
11 #include "Utils/AArch64BaseInfo.h"
12 #include "llvm/MC/MCContext.h"
13 #include "llvm/MC/MCExpr.h"
14 #include "llvm/MC/MCInst.h"
15 #include "llvm/MC/MCRegisterInfo.h"
16 #include "llvm/Support/Format.h"
18 
19 using namespace llvm;
20 
21 #define DEBUG_TYPE "aarch64-disassembler"
22 
24 getVariant(uint64_t LLVMDisassembler_VariantKind) {
25  switch (LLVMDisassembler_VariantKind) {
38  default:
39  llvm_unreachable("bad LLVMDisassembler_VariantKind");
40  }
41 }
42 
43 /// tryAddingSymbolicOperand - tryAddingSymbolicOperand trys to add a symbolic
44 /// operand in place of the immediate Value in the MCInst. The immediate
45 /// Value has not had any PC adjustment made by the caller. If the instruction
46 /// is a branch that adds the PC to the immediate Value then isBranch is
47 /// Success, else Fail. If GetOpInfo is non-null, then it is called to get any
48 /// symbolic information at the Address for this instrution. If that returns
49 /// non-zero then the symbolic information it returns is used to create an
50 /// MCExpr and that is added as an operand to the MCInst. If GetOpInfo()
51 /// returns zero and isBranch is Success then a symbol look up for
52 /// Address + Value is done and if a symbol is found an MCExpr is created with
53 /// that, else an MCExpr with Address + Value is created. If GetOpInfo()
54 /// returns zero and isBranch is Fail then the Opcode of the MCInst is
55 /// tested and for ADRP an other instructions that help to load of pointers
56 /// a symbol look up is done to see it is returns a specific reference type
57 /// to add to the comment stream. This function returns Success if it adds
58 /// an operand to the MCInst and Fail otherwise.
60  MCInst &MI, raw_ostream &CommentStream, int64_t Value, uint64_t Address,
61  bool IsBranch, uint64_t Offset, uint64_t InstSize) {
62  if (!SymbolLookUp)
63  return false;
64  // FIXME: This method shares a lot of code with
65  // MCExternalSymbolizer::tryAddingSymbolicOperand. It may be possible
66  // refactor the MCExternalSymbolizer interface to allow more of this
67  // implementation to be shared.
68  //
69  struct LLVMOpInfo1 SymbolicOp;
70  memset(&SymbolicOp, '\0', sizeof(struct LLVMOpInfo1));
71  SymbolicOp.Value = Value;
72  uint64_t ReferenceType;
73  const char *ReferenceName;
74  if (!GetOpInfo ||
75  !GetOpInfo(DisInfo, Address, 0 /* Offset */, InstSize, 1, &SymbolicOp)) {
76  if (IsBranch) {
79  Address, &ReferenceName);
80  if (Name) {
81  SymbolicOp.AddSymbol.Name = Name;
82  SymbolicOp.AddSymbol.Present = true;
83  SymbolicOp.Value = 0;
84  } else {
85  SymbolicOp.Value = Address + Value;
86  }
88  CommentStream << "symbol stub for: " << ReferenceName;
89  else if (ReferenceType ==
91  CommentStream << "Objc message: " << ReferenceName;
92  } else if (MI.getOpcode() == AArch64::ADRP) {
94  // otool expects the fully encoded ADRP instruction to be passed in as
95  // the value here, so reconstruct it:
96  const MCRegisterInfo &MCRI = *Ctx.getRegisterInfo();
97  uint32_t EncodedInst = 0x90000000;
98  EncodedInst |= (Value & 0x3) << 29; // immlo
99  EncodedInst |= ((Value >> 2) & 0x7FFFF) << 5; // immhi
100  EncodedInst |= MCRI.getEncodingValue(MI.getOperand(0).getReg()); // reg
101  SymbolLookUp(DisInfo, EncodedInst, &ReferenceType, Address,
102  &ReferenceName);
103  CommentStream << format("0x%llx", (0xfffffffffffff000LL & Address) +
104  Value * 0x1000);
105  } else if (MI.getOpcode() == AArch64::ADDXri ||
106  MI.getOpcode() == AArch64::LDRXui ||
107  MI.getOpcode() == AArch64::LDRXl ||
108  MI.getOpcode() == AArch64::ADR) {
109  if (MI.getOpcode() == AArch64::ADDXri)
111  else if (MI.getOpcode() == AArch64::LDRXui)
113  if (MI.getOpcode() == AArch64::LDRXl) {
116  &ReferenceName);
117  } else if (MI.getOpcode() == AArch64::ADR) {
120  &ReferenceName);
121  } else {
122  const MCRegisterInfo &MCRI = *Ctx.getRegisterInfo();
123  // otool expects the fully encoded ADD/LDR instruction to be passed in
124  // as the value here, so reconstruct it:
125  unsigned EncodedInst =
126  MI.getOpcode() == AArch64::ADDXri ? 0x91000000: 0xF9400000;
127  EncodedInst |= Value << 10; // imm12 [+ shift:2 for ADD]
128  EncodedInst |=
129  MCRI.getEncodingValue(MI.getOperand(1).getReg()) << 5; // Rn
130  EncodedInst |= MCRI.getEncodingValue(MI.getOperand(0).getReg()); // Rd
131 
132  SymbolLookUp(DisInfo, EncodedInst, &ReferenceType, Address,
133  &ReferenceName);
134  }
136  CommentStream << "literal pool symbol address: " << ReferenceName;
137  else if (ReferenceType ==
139  CommentStream << "literal pool for: \"";
140  CommentStream.write_escaped(ReferenceName);
141  CommentStream << "\"";
142  } else if (ReferenceType ==
144  CommentStream << "Objc cfstring ref: @\"" << ReferenceName << "\"";
145  else if (ReferenceType ==
147  CommentStream << "Objc message: " << ReferenceName;
148  else if (ReferenceType ==
150  CommentStream << "Objc message ref: " << ReferenceName;
151  else if (ReferenceType ==
153  CommentStream << "Objc selector ref: " << ReferenceName;
154  else if (ReferenceType ==
156  CommentStream << "Objc class ref: " << ReferenceName;
157  // For these instructions, the SymbolLookUp() above is just to get the
158  // ReferenceType and ReferenceName. We want to make sure not to
159  // fall through so we don't build an MCExpr to leave the disassembly
160  // of the immediate values of these instructions to the InstPrinter.
161  return false;
162  } else {
163  return false;
164  }
165  }
166 
167  const MCExpr *Add = nullptr;
168  if (SymbolicOp.AddSymbol.Present) {
169  if (SymbolicOp.AddSymbol.Name) {
170  StringRef Name(SymbolicOp.AddSymbol.Name);
173  if (Variant != MCSymbolRefExpr::VK_None)
174  Add = MCSymbolRefExpr::create(Sym, Variant, Ctx);
175  else
177  } else {
179  }
180  }
181 
182  const MCExpr *Sub = nullptr;
183  if (SymbolicOp.SubtractSymbol.Present) {
184  if (SymbolicOp.SubtractSymbol.Name) {
185  StringRef Name(SymbolicOp.SubtractSymbol.Name);
187  Sub = MCSymbolRefExpr::create(Sym, Ctx);
188  } else {
189  Sub = MCConstantExpr::create(SymbolicOp.SubtractSymbol.Value, Ctx);
190  }
191  }
192 
193  const MCExpr *Off = nullptr;
194  if (SymbolicOp.Value != 0)
195  Off = MCConstantExpr::create(SymbolicOp.Value, Ctx);
196 
197  const MCExpr *Expr;
198  if (Sub) {
199  const MCExpr *LHS;
200  if (Add)
201  LHS = MCBinaryExpr::createSub(Add, Sub, Ctx);
202  else
203  LHS = MCUnaryExpr::createMinus(Sub, Ctx);
204  if (Off)
205  Expr = MCBinaryExpr::createAdd(LHS, Off, Ctx);
206  else
207  Expr = LHS;
208  } else if (Add) {
209  if (Off)
210  Expr = MCBinaryExpr::createAdd(Add, Off, Ctx);
211  else
212  Expr = Add;
213  } else {
214  if (Off)
215  Expr = Off;
216  else
217  Expr = MCConstantExpr::create(0, Ctx);
218  }
219 
220  MI.addOperand(MCOperand::createExpr(Expr));
221 
222  return true;
223 }
MI
IRTranslator LLVM IR MI
Definition: IRTranslator.cpp:102
llvm
Definition: AllocatorList.h:23
llvm::MCSymbolizer::Ctx
MCContext & Ctx
Definition: MCSymbolizer.h:41
llvm::MCSymbol
MCSymbol - Instances of this class represent a symbol name in the MC file, and MCSymbols are created ...
Definition: MCSymbol.h:41
llvm::MCSymbolRefExpr::VK_PAGEOFF
@ VK_PAGEOFF
Definition: MCExpr.h:219
llvm::MCOperand::createExpr
static MCOperand createExpr(const MCExpr *Val)
Definition: MCInst.h:162
LLVMDisassembler_VariantKind_ARM64_TLVP
#define LLVMDisassembler_VariantKind_ARM64_TLVP
Definition: DisassemblerTypes.h:97
ReferenceType
Definition: ItaniumDemangle.h:525
llvm::MCContext::getRegisterInfo
const MCRegisterInfo * getRegisterInfo() const
Definition: MCContext.h:422
llvm::MCConstantExpr::create
static const MCConstantExpr * create(int64_t Value, MCContext &Ctx, bool PrintInHex=false, unsigned SizeInBytes=0)
Definition: MCExpr.cpp:194
LLVMOpInfo1::AddSymbol
struct LLVMOpInfoSymbol1 AddSymbol
Definition: DisassemblerTypes.h:73
llvm::MCUnaryExpr::createMinus
static const MCUnaryExpr * createMinus(const MCExpr *Expr, MCContext &Ctx, SMLoc Loc=SMLoc())
Definition: MCExpr.h:449
llvm::raw_ostream::write_escaped
raw_ostream & write_escaped(StringRef Str, bool UseHexEscapes=false)
Output Str, turning '\', '\t', ' ', '"', and anything that doesn't satisfy llvm::isPrint into an esca...
Definition: raw_ostream.cpp:163
LLVMDisassembler_VariantKind_ARM64_GOTPAGE
#define LLVMDisassembler_VariantKind_ARM64_GOTPAGE
Definition: DisassemblerTypes.h:95
AArch64BaseInfo.h
LLVMDisassembler_ReferenceType_Out_Objc_Selector_Ref
#define LLVMDisassembler_ReferenceType_Out_Objc_Selector_Ref
Definition: DisassemblerTypes.h:153
llvm::AArch64ExternalSymbolizer::tryAddingSymbolicOperand
bool tryAddingSymbolicOperand(MCInst &MI, raw_ostream &CommentStream, int64_t Value, uint64_t Address, bool IsBranch, uint64_t Offset, uint64_t InstSize) override
tryAddingSymbolicOperand - tryAddingSymbolicOperand trys to add a symbolic operand in place of the im...
Definition: AArch64ExternalSymbolizer.cpp:59
LLVMDisassembler_ReferenceType_In_Branch
#define LLVMDisassembler_ReferenceType_In_Branch
Definition: DisassemblerTypes.h:124
LLVMDisassembler_VariantKind_ARM64_PAGEOFF
#define LLVMDisassembler_VariantKind_ARM64_PAGEOFF
Definition: DisassemblerTypes.h:94
llvm::MCRegisterInfo::getEncodingValue
uint16_t getEncodingValue(MCRegister RegNo) const
Returns the encoding for RegNo.
Definition: MCRegisterInfo.h:553
llvm::MCExternalSymbolizer::GetOpInfo
LLVMOpInfoCallback GetOpInfo
Definition: MCExternalSymbolizer.h:32
AArch64ExternalSymbolizer.h
LLVMDisassembler_ReferenceType_In_ARM64_LDRXl
#define LLVMDisassembler_ReferenceType_In_ARM64_LDRXl
Definition: DisassemblerTypes.h:135
Offset
uint64_t Offset
Definition: ELFObjHandler.cpp:81
llvm::MCInst
Instances of this class represent a single low-level machine instruction.
Definition: MCInst.h:184
Format.h
LLVMDisassembler_ReferenceType_In_ARM64_ADDXri
#define LLVMDisassembler_ReferenceType_In_ARM64_ADDXri
Definition: DisassemblerTypes.h:131
LLVMDisassembler_ReferenceType_In_ARM64_ADR
#define LLVMDisassembler_ReferenceType_In_ARM64_ADR
Definition: DisassemblerTypes.h:137
llvm::MCContext::getOrCreateSymbol
MCSymbol * getOrCreateSymbol(const Twine &Name)
Lookup the symbol inside with the specified Name.
Definition: MCContext.cpp:187
llvm::MCSymbolRefExpr::VK_GOTPAGEOFF
@ VK_GOTPAGEOFF
Definition: MCExpr.h:221
LLVMDisassembler_VariantKind_ARM64_PAGE
#define LLVMDisassembler_VariantKind_ARM64_PAGE
The ARM64 target VariantKinds.
Definition: DisassemblerTypes.h:93
MCContext.h
LLVMOpInfoSymbol1::Name
const char * Name
Definition: DisassemblerTypes.h:68
LLVMOpInfoSymbol1::Value
uint64_t Value
Definition: DisassemblerTypes.h:69
LLVMOpInfo1
Definition: DisassemblerTypes.h:72
MCInst.h
LLVMDisassembler_VariantKind_ARM64_GOTPAGEOFF
#define LLVMDisassembler_VariantKind_ARM64_GOTPAGEOFF
Definition: DisassemblerTypes.h:96
LLVMDisassembler_ReferenceType_Out_SymbolStub
#define LLVMDisassembler_ReferenceType_Out_SymbolStub
Definition: DisassemblerTypes.h:140
llvm::raw_ostream
This class implements an extremely fast bulk output stream that can only output to a stream.
Definition: raw_ostream.h:50
getVariant
static MCSymbolRefExpr::VariantKind getVariant(uint64_t LLVMDisassembler_VariantKind)
Definition: AArch64ExternalSymbolizer.cpp:24
LLVMDisassembler_ReferenceType_In_ARM64_ADRP
#define LLVMDisassembler_ReferenceType_In_ARM64_ADRP
Definition: DisassemblerTypes.h:129
AArch64AddressingModes.h
LLVMDisassembler_ReferenceType_Out_Objc_CFString_Ref
#define LLVMDisassembler_ReferenceType_Out_Objc_CFString_Ref
Definition: DisassemblerTypes.h:147
llvm::MCSymbolRefExpr::VariantKind
VariantKind
Definition: MCExpr.h:194
llvm::MCExternalSymbolizer::DisInfo
void * DisInfo
The pointer to the block of symbolic information for above call back.
Definition: MCExternalSymbolizer.h:36
LLVMOpInfo1::Value
uint64_t Value
Definition: DisassemblerTypes.h:75
MCRegisterInfo.h
llvm::HighlightColor::Address
@ Address
llvm::MCExternalSymbolizer::SymbolLookUp
LLVMSymbolLookupCallback SymbolLookUp
The function to lookup a symbol name.
Definition: MCExternalSymbolizer.h:34
llvm::MCBinaryExpr::createSub
static const MCBinaryExpr * createSub(const MCExpr *LHS, const MCExpr *RHS, MCContext &Ctx)
Definition: MCExpr.h:608
LLVMDisassembler_ReferenceType_Out_LitPool_CstrAddr
#define LLVMDisassembler_ReferenceType_Out_LitPool_CstrAddr
Definition: DisassemblerTypes.h:144
llvm::MCBinaryExpr::createAdd
static const MCBinaryExpr * createAdd(const MCExpr *LHS, const MCExpr *RHS, MCContext &Ctx)
Definition: MCExpr.h:523
LLVMDisassembler_ReferenceType_Out_Objc_Message_Ref
#define LLVMDisassembler_ReferenceType_Out_Objc_Message_Ref
Definition: DisassemblerTypes.h:151
LLVMDisassembler_VariantKind_None
#define LLVMDisassembler_VariantKind_None
The operand VariantKinds for symbolic disassembly.
Definition: DisassemblerTypes.h:82
llvm::StringRef
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:58
llvm_unreachable
#define llvm_unreachable(msg)
Marks that the current location is not supposed to be reachable.
Definition: ErrorHandling.h:136
uint32_t
llvm::AArch64ISD::ADR
@ ADR
Definition: AArch64ISelLowering.h:62
llvm::format
format_object< Ts... > format(const char *Fmt, const Ts &... Vals)
These are helper functions used to produce formatted output.
Definition: Format.h:124
llvm::MCRegisterInfo
MCRegisterInfo base class - We assume that the target defines a static array of MCRegisterDesc object...
Definition: MCRegisterInfo.h:135
LLVMDisassembler_ReferenceType_Out_Objc_Class_Ref
#define LLVMDisassembler_ReferenceType_Out_Objc_Class_Ref
Definition: DisassemblerTypes.h:155
llvm::GraphProgram::Name
Name
Definition: GraphWriter.h:52
LLVMDisassembler_VariantKind_ARM64_TLVOFF
#define LLVMDisassembler_VariantKind_ARM64_TLVOFF
Definition: DisassemblerTypes.h:98
LLVMOpInfo1::VariantKind
uint64_t VariantKind
Definition: DisassemblerTypes.h:76
llvm::TargetStackID::Value
Value
Definition: TargetFrameLowering.h:27
llvm::MCSymbolRefExpr::create
static const MCSymbolRefExpr * create(const MCSymbol *Symbol, MCContext &Ctx)
Definition: MCExpr.h:384
LLVMDisassembler_ReferenceType_Out_LitPool_SymAddr
#define LLVMDisassembler_ReferenceType_Out_LitPool_SymAddr
Definition: DisassemblerTypes.h:142
llvm::MCID::Add
@ Add
Definition: MCInstrDesc.h:183
llvm::AArch64ISD::ADRP
@ ADRP
Definition: AArch64ISelLowering.h:61
LLVMDisassembler_ReferenceType_In_ARM64_LDRXui
#define LLVMDisassembler_ReferenceType_In_ARM64_LDRXui
Definition: DisassemblerTypes.h:133
LLVMDisassembler_ReferenceType_Out_Objc_Message
#define LLVMDisassembler_ReferenceType_Out_Objc_Message
Definition: DisassemblerTypes.h:149
llvm::MCSymbolRefExpr::VK_None
@ VK_None
Definition: MCExpr.h:195
LLVMOpInfoSymbol1::Present
uint64_t Present
Definition: DisassemblerTypes.h:67
llvm::MCSymbolRefExpr::VK_GOTPAGE
@ VK_GOTPAGE
Definition: MCExpr.h:220
raw_ostream.h
llvm::MCSymbolRefExpr::VK_PAGE
@ VK_PAGE
Definition: MCExpr.h:218
MCExpr.h
LLVMOpInfo1::SubtractSymbol
struct LLVMOpInfoSymbol1 SubtractSymbol
Definition: DisassemblerTypes.h:74
llvm::Value
LLVM Value Representation.
Definition: Value.h:75
llvm::MCExpr
Base class for the full range of assembler expressions which are needed for parsing.
Definition: MCExpr.h:35