1 //===- Disassembler.cpp - Disassembler for hex strings --------------------===// 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 // 9 // This class implements the disassembler of strings of bytes written in 10 // hexadecimal, from standard input or from a file. 11 // 12 //===----------------------------------------------------------------------===// 13 14 #include "Disassembler.h" 15 #include "llvm/ADT/Triple.h" 16 #include "llvm/MC/MCAsmInfo.h" 17 #include "llvm/MC/MCContext.h" 18 #include "llvm/MC/MCDisassembler/MCDisassembler.h" 19 #include "llvm/MC/MCInst.h" 20 #include "llvm/MC/MCObjectFileInfo.h" 21 #include "llvm/MC/MCRegisterInfo.h" 22 #include "llvm/MC/MCStreamer.h" 23 #include "llvm/MC/MCSubtargetInfo.h" 24 #include "llvm/Support/MemoryBuffer.h" 25 #include "llvm/Support/SourceMgr.h" 26 #include "llvm/Support/TargetRegistry.h" 27 #include "llvm/Support/raw_ostream.h" 28 29 using namespace llvm; 30 31 typedef std::pair<std::vector<unsigned char>, std::vector<const char *>> 32 ByteArrayTy; 33 34 static bool PrintInsts(const MCDisassembler &DisAsm, 35 const ByteArrayTy &Bytes, 36 SourceMgr &SM, raw_ostream &Out, 37 MCStreamer &Streamer, bool InAtomicBlock, 38 const MCSubtargetInfo &STI) { 39 ArrayRef<uint8_t> Data(Bytes.first.data(), Bytes.first.size()); 40 41 // Disassemble it to strings. 42 uint64_t Size; 43 uint64_t Index; 44 45 for (Index = 0; Index < Bytes.first.size(); Index += Size) { 46 MCInst Inst; 47 48 MCDisassembler::DecodeStatus S; 49 S = DisAsm.getInstruction(Inst, Size, Data.slice(Index), Index, 50 /*REMOVE*/ nulls(), nulls()); 51 switch (S) { 52 case MCDisassembler::Fail: 53 SM.PrintMessage(SMLoc::getFromPointer(Bytes.second[Index]), 54 SourceMgr::DK_Warning, 55 "invalid instruction encoding"); 56 // Don't try to resynchronise the stream in a block 57 if (InAtomicBlock) 58 return true; 59 60 if (Size == 0) 61 Size = 1; // skip illegible bytes 62 63 break; 64 65 case MCDisassembler::SoftFail: 66 SM.PrintMessage(SMLoc::getFromPointer(Bytes.second[Index]), 67 SourceMgr::DK_Warning, 68 "potentially undefined instruction encoding"); 69 LLVM_FALLTHROUGH; 70 71 case MCDisassembler::Success: 72 Streamer.EmitInstruction(Inst, STI); 73 break; 74 } 75 } 76 77 return false; 78 } 79 80 static bool SkipToToken(StringRef &Str) { 81 for (;;) { 82 if (Str.empty()) 83 return false; 84 85 // Strip horizontal whitespace and commas. 86 if (size_t Pos = Str.find_first_not_of(" \t\r\n,")) { 87 Str = Str.substr(Pos); 88 continue; 89 } 90 91 // If this is the start of a comment, remove the rest of the line. 92 if (Str[0] == '#') { 93 Str = Str.substr(Str.find_first_of('\n')); 94 continue; 95 } 96 return true; 97 } 98 } 99 100 101 static bool ByteArrayFromString(ByteArrayTy &ByteArray, 102 StringRef &Str, 103 SourceMgr &SM) { 104 while (SkipToToken(Str)) { 105 // Handled by higher level 106 if (Str[0] == '[' || Str[0] == ']') 107 return false; 108 109 // Get the current token. 110 size_t Next = Str.find_first_of(" \t\n\r,#[]"); 111 StringRef Value = Str.substr(0, Next); 112 113 // Convert to a byte and add to the byte vector. 114 unsigned ByteVal; 115 if (Value.getAsInteger(0, ByteVal) || ByteVal > 255) { 116 // If we have an error, print it and skip to the end of line. 117 SM.PrintMessage(SMLoc::getFromPointer(Value.data()), SourceMgr::DK_Error, 118 "invalid input token"); 119 Str = Str.substr(Str.find('\n')); 120 ByteArray.first.clear(); 121 ByteArray.second.clear(); 122 continue; 123 } 124 125 ByteArray.first.push_back(ByteVal); 126 ByteArray.second.push_back(Value.data()); 127 Str = Str.substr(Next); 128 } 129 130 return false; 131 } 132 133 int Disassembler::disassemble(const Target &T, const std::string &Triple, 134 MCSubtargetInfo &STI, MCStreamer &Streamer, 135 MemoryBuffer &Buffer, SourceMgr &SM, 136 MCContext &Ctx, raw_ostream &Out) { 137 138 std::unique_ptr<const MCRegisterInfo> MRI(T.createMCRegInfo(Triple)); 139 if (!MRI) { 140 errs() << "error: no register info for target " << Triple << "\n"; 141 return -1; 142 } 143 144 std::unique_ptr<const MCAsmInfo> MAI(T.createMCAsmInfo(*MRI, Triple)); 145 if (!MAI) { 146 errs() << "error: no assembly info for target " << Triple << "\n"; 147 return -1; 148 } 149 150 std::unique_ptr<const MCDisassembler> DisAsm( 151 T.createMCDisassembler(STI, Ctx)); 152 if (!DisAsm) { 153 errs() << "error: no disassembler for target " << Triple << "\n"; 154 return -1; 155 } 156 157 // Set up initial section manually here 158 Streamer.InitSections(false); 159 160 bool ErrorOccurred = false; 161 162 // Convert the input to a vector for disassembly. 163 ByteArrayTy ByteArray; 164 StringRef Str = Buffer.getBuffer(); 165 bool InAtomicBlock = false; 166 167 while (SkipToToken(Str)) { 168 ByteArray.first.clear(); 169 ByteArray.second.clear(); 170 171 if (Str[0] == '[') { 172 if (InAtomicBlock) { 173 SM.PrintMessage(SMLoc::getFromPointer(Str.data()), SourceMgr::DK_Error, 174 "nested atomic blocks make no sense"); 175 ErrorOccurred = true; 176 } 177 InAtomicBlock = true; 178 Str = Str.drop_front(); 179 continue; 180 } else if (Str[0] == ']') { 181 if (!InAtomicBlock) { 182 SM.PrintMessage(SMLoc::getFromPointer(Str.data()), SourceMgr::DK_Error, 183 "attempt to close atomic block without opening"); 184 ErrorOccurred = true; 185 } 186 InAtomicBlock = false; 187 Str = Str.drop_front(); 188 continue; 189 } 190 191 // It's a real token, get the bytes and emit them 192 ErrorOccurred |= ByteArrayFromString(ByteArray, Str, SM); 193 194 if (!ByteArray.first.empty()) 195 ErrorOccurred |= PrintInsts(*DisAsm, ByteArray, SM, Out, Streamer, 196 InAtomicBlock, STI); 197 } 198 199 if (InAtomicBlock) { 200 SM.PrintMessage(SMLoc::getFromPointer(Str.data()), SourceMgr::DK_Error, 201 "unclosed atomic block"); 202 ErrorOccurred = true; 203 } 204 205 return ErrorOccurred; 206 } 207