1 //===- DWARFGdbIndex.cpp --------------------------------------------------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 
10 #include "llvm/DebugInfo/DWARF/DWARFGdbIndex.h"
11 #include "llvm/ADT/SmallVector.h"
12 #include "llvm/ADT/StringRef.h"
13 #include "llvm/Support/Format.h"
14 #include "llvm/Support/raw_ostream.h"
15 #include <algorithm>
16 #include <cassert>
17 #include <cinttypes>
18 #include <cstdint>
19 #include <utility>
20 
21 using namespace llvm;
22 
23 // .gdb_index section format reference:
24 // https://sourceware.org/gdb/onlinedocs/gdb/Index-Section-Format.html
25 
26 void DWARFGdbIndex::dumpCUList(raw_ostream &OS) const {
27   OS << format("\n  CU list offset = 0x%x, has %" PRId64 " entries:",
28                CuListOffset, (uint64_t)CuList.size())
29      << '\n';
30   uint32_t I = 0;
31   for (const CompUnitEntry &CU : CuList)
32     OS << format("    %d: Offset = 0x%llx, Length = 0x%llx\n", I++, CU.Offset,
33                  CU.Length);
34 }
35 
36 void DWARFGdbIndex::dumpAddressArea(raw_ostream &OS) const {
37   OS << format("\n  Address area offset = 0x%x, has %" PRId64 " entries:",
38                AddressAreaOffset, (uint64_t)AddressArea.size())
39      << '\n';
40   for (const AddressEntry &Addr : AddressArea)
41     OS << format(
42         "    Low/High address = [0x%llx, 0x%llx) (Size: 0x%llx), CU id = %d\n",
43         Addr.LowAddress, Addr.HighAddress, Addr.HighAddress - Addr.LowAddress,
44         Addr.CuIndex);
45 }
46 
47 void DWARFGdbIndex::dumpSymbolTable(raw_ostream &OS) const {
48   OS << format("\n  Symbol table offset = 0x%x, size = %" PRId64
49                ", filled slots:",
50                SymbolTableOffset, (uint64_t)SymbolTable.size())
51      << '\n';
52   uint32_t I = -1;
53   for (const SymTableEntry &E : SymbolTable) {
54     ++I;
55     if (!E.NameOffset && !E.VecOffset)
56       continue;
57 
58     OS << format("    %d: Name offset = 0x%x, CU vector offset = 0x%x\n", I,
59                  E.NameOffset, E.VecOffset);
60 
61     StringRef Name = ConstantPoolStrings.substr(
62         ConstantPoolOffset - StringPoolOffset + E.NameOffset);
63 
64     auto CuVector = std::find_if(
65         ConstantPoolVectors.begin(), ConstantPoolVectors.end(),
66         [&](const std::pair<uint32_t, SmallVector<uint32_t, 0>> &V) {
67           return V.first == E.VecOffset;
68         });
69     assert(CuVector != ConstantPoolVectors.end() && "Invalid symbol table");
70     uint32_t CuVectorId = CuVector - ConstantPoolVectors.begin();
71     OS << format("      String name: %s, CU vector index: %d\n", Name.data(),
72                  CuVectorId);
73   }
74 }
75 
76 void DWARFGdbIndex::dumpConstantPool(raw_ostream &OS) const {
77   OS << format("\n  Constant pool offset = 0x%x, has %" PRId64 " CU vectors:",
78                ConstantPoolOffset, (uint64_t)ConstantPoolVectors.size());
79   uint32_t I = 0;
80   for (const auto &V : ConstantPoolVectors) {
81     OS << format("\n    %d(0x%x): ", I++, V.first);
82     for (uint32_t Val : V.second)
83       OS << format("0x%x ", Val);
84   }
85   OS << '\n';
86 }
87 
88 void DWARFGdbIndex::dump(raw_ostream &OS) {
89   if (HasError) {
90     OS << "\n<error parsing>\n";
91     return;
92   }
93 
94   if (HasContent) {
95     OS << "  Version = " << Version << '\n';
96     dumpCUList(OS);
97     dumpAddressArea(OS);
98     dumpSymbolTable(OS);
99     dumpConstantPool(OS);
100   }
101 }
102 
103 bool DWARFGdbIndex::parseImpl(DataExtractor Data) {
104   uint32_t Offset = 0;
105 
106   // Only version 7 is supported at this moment.
107   Version = Data.getU32(&Offset);
108   if (Version != 7)
109     return false;
110 
111   CuListOffset = Data.getU32(&Offset);
112   uint32_t CuTypesOffset = Data.getU32(&Offset);
113   AddressAreaOffset = Data.getU32(&Offset);
114   SymbolTableOffset = Data.getU32(&Offset);
115   ConstantPoolOffset = Data.getU32(&Offset);
116 
117   if (Offset != CuListOffset)
118     return false;
119 
120   uint32_t CuListSize = (CuTypesOffset - CuListOffset) / 16;
121   CuList.reserve(CuListSize);
122   for (uint32_t i = 0; i < CuListSize; ++i) {
123     uint64_t CuOffset = Data.getU64(&Offset);
124     uint64_t CuLength = Data.getU64(&Offset);
125     CuList.push_back({CuOffset, CuLength});
126   }
127 
128   // CU Types are no longer needed as DWARF skeleton type units never made it
129   // into the standard.
130   uint32_t CuTypesListSize = (AddressAreaOffset - CuTypesOffset) / 24;
131   if (CuTypesListSize != 0)
132     return false;
133 
134   uint32_t AddressAreaSize = (SymbolTableOffset - AddressAreaOffset) / 20;
135   AddressArea.reserve(AddressAreaSize);
136   for (uint32_t i = 0; i < AddressAreaSize; ++i) {
137     uint64_t LowAddress = Data.getU64(&Offset);
138     uint64_t HighAddress = Data.getU64(&Offset);
139     uint32_t CuIndex = Data.getU32(&Offset);
140     AddressArea.push_back({LowAddress, HighAddress, CuIndex});
141   }
142 
143   // The symbol table. This is an open addressed hash table. The size of the
144   // hash table is always a power of 2.
145   // Each slot in the hash table consists of a pair of offset_type values. The
146   // first value is the offset of the symbol's name in the constant pool. The
147   // second value is the offset of the CU vector in the constant pool.
148   // If both values are 0, then this slot in the hash table is empty. This is ok
149   // because while 0 is a valid constant pool index, it cannot be a valid index
150   // for both a string and a CU vector.
151   uint32_t SymTableSize = (ConstantPoolOffset - SymbolTableOffset) / 8;
152   SymbolTable.reserve(SymTableSize);
153   uint32_t CuVectorsTotal = 0;
154   for (uint32_t i = 0; i < SymTableSize; ++i) {
155     uint32_t NameOffset = Data.getU32(&Offset);
156     uint32_t CuVecOffset = Data.getU32(&Offset);
157     SymbolTable.push_back({NameOffset, CuVecOffset});
158     if (NameOffset || CuVecOffset)
159       ++CuVectorsTotal;
160   }
161 
162   // The constant pool. CU vectors are stored first, followed by strings.
163   // The first value is the number of CU indices in the vector. Each subsequent
164   // value is the index and symbol attributes of a CU in the CU list.
165   for (uint32_t i = 0; i < CuVectorsTotal; ++i) {
166     ConstantPoolVectors.emplace_back(0, SmallVector<uint32_t, 0>());
167     auto &Vec = ConstantPoolVectors.back();
168     Vec.first = Offset - ConstantPoolOffset;
169 
170     uint32_t Num = Data.getU32(&Offset);
171     for (uint32_t j = 0; j < Num; ++j)
172       Vec.second.push_back(Data.getU32(&Offset));
173   }
174 
175   ConstantPoolStrings = Data.getData().drop_front(Offset);
176   StringPoolOffset = Offset;
177   return true;
178 }
179 
180 void DWARFGdbIndex::parse(DataExtractor Data) {
181   HasContent = !Data.getData().empty();
182   HasError = HasContent && !parseImpl(Data);
183 }
184