blob: 8f3b76ccc89449ca4b1eaabd0757269ff9a494a8 [file] [log] [blame]
Armando Montanez31f0f652019-01-03 18:32:36 +00001//===- ELFObjHandler.cpp --------------------------------------------------===//
2//
Chandler Carruth2946cd72019-01-19 08:50:56 +00003// 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
Armando Montanez31f0f652019-01-03 18:32:36 +00006//
7//===-----------------------------------------------------------------------===/
8
9#include "ELFObjHandler.h"
10#include "llvm/Object/Binary.h"
11#include "llvm/Object/ELFObjectFile.h"
12#include "llvm/Object/ELFTypes.h"
13#include "llvm/Support/Errc.h"
14#include "llvm/Support/Error.h"
15#include "llvm/Support/MemoryBuffer.h"
16#include "llvm/TextAPI/ELF/ELFStub.h"
17
18using llvm::MemoryBufferRef;
19using llvm::object::ELFObjectFile;
20
21using namespace llvm;
22using namespace llvm::object;
Armando Montanez31f0f652019-01-03 18:32:36 +000023using namespace llvm::ELF;
24
25namespace llvm {
26namespace elfabi {
27
Armando Montanezfe7ab3c2019-01-16 17:47:16 +000028// Simple struct to hold relevant .dynamic entries.
29struct DynamicEntries {
30 uint64_t StrTabAddr = 0;
31 uint64_t StrSize = 0;
32 Optional<uint64_t> SONameOffset;
Armando Montanez56d18122019-01-18 20:56:03 +000033 std::vector<uint64_t> NeededLibNames;
Armando Montanez8367b072019-01-24 22:39:21 +000034 // Symbol table:
35 uint64_t DynSymAddr = 0;
36 // Hash tables:
37 Optional<uint64_t> ElfHash;
38 Optional<uint64_t> GnuHash;
Armando Montanezfe7ab3c2019-01-16 17:47:16 +000039};
40
41/// This function behaves similarly to StringRef::substr(), but attempts to
42/// terminate the returned StringRef at the first null terminator. If no null
43/// terminator is found, an error is returned.
44///
45/// @param Str Source string to create a substring from.
46/// @param Offset The start index of the desired substring.
47static Expected<StringRef> terminatedSubstr(StringRef Str, size_t Offset) {
48 size_t StrEnd = Str.find('\0', Offset);
49 if (StrEnd == StringLiteral::npos) {
50 return createError(
51 "String overran bounds of string table (no null terminator)");
52 }
53
54 size_t StrLen = StrEnd - Offset;
55 return Str.substr(Offset, StrLen);
56}
57
58/// This function takes an error, and appends a string of text to the end of
59/// that error. Since "appending" to an Error isn't supported behavior of an
60/// Error, this function technically creates a new error with the combined
61/// message and consumes the old error.
62///
63/// @param Err Source error.
64/// @param After Text to append at the end of Err's error message.
65Error appendToError(Error Err, StringRef After) {
66 std::string Message;
67 raw_string_ostream Stream(Message);
68 Stream << Err;
69 Stream << " " << After;
70 consumeError(std::move(Err));
71 return createError(Stream.str().c_str());
72}
73
74/// This function populates a DynamicEntries struct using an ELFT::DynRange.
75/// After populating the struct, the members are validated with
76/// some basic sanity checks.
77///
78/// @param Dyn Target DynamicEntries struct to populate.
79/// @param DynTable Source dynamic table.
80template <class ELFT>
81static Error populateDynamic(DynamicEntries &Dyn,
82 typename ELFT::DynRange DynTable) {
83 if (DynTable.empty())
84 return createError("No .dynamic section found");
85
86 // Search .dynamic for relevant entries.
87 bool FoundDynStr = false;
88 bool FoundDynStrSz = false;
Armando Montanez8367b072019-01-24 22:39:21 +000089 bool FoundDynSym = false;
Armando Montanezfe7ab3c2019-01-16 17:47:16 +000090 for (auto &Entry : DynTable) {
91 switch (Entry.d_tag) {
92 case DT_SONAME:
93 Dyn.SONameOffset = Entry.d_un.d_val;
94 break;
95 case DT_STRTAB:
96 Dyn.StrTabAddr = Entry.d_un.d_ptr;
97 FoundDynStr = true;
98 break;
99 case DT_STRSZ:
100 Dyn.StrSize = Entry.d_un.d_val;
101 FoundDynStrSz = true;
102 break;
Armando Montanez56d18122019-01-18 20:56:03 +0000103 case DT_NEEDED:
104 Dyn.NeededLibNames.push_back(Entry.d_un.d_val);
105 break;
Armando Montanez8367b072019-01-24 22:39:21 +0000106 case DT_SYMTAB:
107 Dyn.DynSymAddr = Entry.d_un.d_ptr;
108 FoundDynSym = true;
109 break;
110 case DT_HASH:
111 Dyn.ElfHash = Entry.d_un.d_ptr;
112 break;
113 case DT_GNU_HASH:
114 Dyn.GnuHash = Entry.d_un.d_ptr;
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000115 }
116 }
117
118 if (!FoundDynStr) {
119 return createError(
120 "Couldn't locate dynamic string table (no DT_STRTAB entry)");
121 }
122 if (!FoundDynStrSz) {
123 return createError(
124 "Couldn't determine dynamic string table size (no DT_STRSZ entry)");
125 }
Armando Montanez8367b072019-01-24 22:39:21 +0000126 if (!FoundDynSym) {
127 return createError(
128 "Couldn't locate dynamic symbol table (no DT_SYMTAB entry)");
129 }
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000130 if (Dyn.SONameOffset.hasValue() && *Dyn.SONameOffset >= Dyn.StrSize) {
131 return createStringError(
132 object_error::parse_failed,
133 "DT_SONAME string offset (0x%016x) outside of dynamic string table",
134 *Dyn.SONameOffset);
135 }
Armando Montanez56d18122019-01-18 20:56:03 +0000136 for (uint64_t Offset : Dyn.NeededLibNames) {
137 if (Offset >= Dyn.StrSize) {
138 return createStringError(
139 object_error::parse_failed,
140 "DT_NEEDED string offset (0x%016x) outside of dynamic string table",
141 Offset);
142 }
143 }
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000144
145 return Error::success();
146}
147
Armando Montanez8367b072019-01-24 22:39:21 +0000148/// This function finds the number of dynamic symbols using a GNU hash table.
149///
150/// @param Table The GNU hash table for .dynsym.
151template <class ELFT>
152static uint64_t getDynSymtabSize(const typename ELFT::GnuHash &Table) {
153 using Elf_Word = typename ELFT::Word;
154 if (Table.nbuckets == 0)
155 return Table.symndx + 1;
156 uint64_t LastSymIdx = 0;
157 uint64_t BucketVal = 0;
158 // Find the index of the first symbol in the last chain.
159 for (Elf_Word Val : Table.buckets()) {
160 BucketVal = std::max(BucketVal, (uint64_t)Val);
161 }
162 LastSymIdx += BucketVal;
163 const Elf_Word *It =
164 reinterpret_cast<const Elf_Word *>(Table.values(BucketVal).end());
165 // Locate the end of the chain to find the last symbol index.
166 while ((*It & 1) == 0) {
167 LastSymIdx++;
168 It++;
169 }
170 return LastSymIdx + 1;
171}
172
173/// This function determines the number of dynamic symbols.
174/// Without access to section headers, the number of symbols must be determined
175/// by parsing dynamic hash tables.
176///
177/// @param Dyn Entries with the locations of hash tables.
178/// @param ElfFile The ElfFile that the section contents reside in.
179template <class ELFT>
180static Expected<uint64_t> getNumSyms(DynamicEntries &Dyn,
181 const ELFFile<ELFT> &ElfFile) {
182 using Elf_Hash = typename ELFT::Hash;
183 using Elf_GnuHash = typename ELFT::GnuHash;
184 // Search GNU hash table to try to find the upper bound of dynsym.
185 if (Dyn.GnuHash.hasValue()) {
186 Expected<const uint8_t *> TablePtr = ElfFile.toMappedAddr(*Dyn.GnuHash);
187 if (!TablePtr)
188 return TablePtr.takeError();
189 const Elf_GnuHash *Table =
190 reinterpret_cast<const Elf_GnuHash *>(TablePtr.get());
191 return getDynSymtabSize<ELFT>(*Table);
192 }
193 // Search SYSV hash table to try to find the upper bound of dynsym.
194 if (Dyn.ElfHash.hasValue()) {
195 Expected<const uint8_t *> TablePtr = ElfFile.toMappedAddr(*Dyn.ElfHash);
196 if (!TablePtr)
197 return TablePtr.takeError();
198 const Elf_Hash *Table = reinterpret_cast<const Elf_Hash *>(TablePtr.get());
199 return Table->nchain;
200 }
201 return 0;
202}
203
204/// This function extracts symbol type from a symbol's st_info member and
205/// maps it to an ELFSymbolType enum.
206/// Currently, STT_NOTYPE, STT_OBJECT, STT_FUNC, and STT_TLS are supported.
207/// Other symbol types are mapped to ELFSymbolType::Unknown.
208///
209/// @param Info Binary symbol st_info to extract symbol type from.
210static ELFSymbolType convertInfoToType(uint8_t Info) {
211 Info = Info & 0xf;
212 switch (Info) {
213 case ELF::STT_NOTYPE:
214 return ELFSymbolType::NoType;
215 case ELF::STT_OBJECT:
216 return ELFSymbolType::Object;
217 case ELF::STT_FUNC:
218 return ELFSymbolType::Func;
219 case ELF::STT_TLS:
220 return ELFSymbolType::TLS;
221 default:
222 return ELFSymbolType::Unknown;
223 }
224}
225
226/// This function creates an ELFSymbol and populates all members using
227/// information from a binary ELFT::Sym.
228///
229/// @param SymName The desired name of the ELFSymbol.
230/// @param RawSym ELFT::Sym to extract symbol information from.
231template <class ELFT>
232static ELFSymbol createELFSym(StringRef SymName,
233 const typename ELFT::Sym &RawSym) {
234 ELFSymbol TargetSym(SymName);
235 uint8_t Binding = RawSym.getBinding();
236 if (Binding == STB_WEAK)
237 TargetSym.Weak = true;
238 else
239 TargetSym.Weak = false;
240
241 TargetSym.Undefined = RawSym.isUndefined();
242 TargetSym.Type = convertInfoToType(RawSym.st_info);
243
244 if (TargetSym.Type == ELFSymbolType::Func) {
245 TargetSym.Size = 0;
246 } else {
247 TargetSym.Size = RawSym.st_size;
248 }
249 return TargetSym;
250}
251
252/// This function populates an ELFStub with symbols using information read
253/// from an ELF binary.
254///
255/// @param TargetStub ELFStub to add symbols to.
256/// @param DynSym Range of dynamic symbols to add to TargetStub.
257/// @param DynStr StringRef to the dynamic string table.
258template <class ELFT>
259static Error populateSymbols(ELFStub &TargetStub,
260 const typename ELFT::SymRange DynSym,
261 StringRef DynStr) {
262 // Skips the first symbol since it's the NULL symbol.
263 for (auto RawSym : DynSym.drop_front(1)) {
264 // If a symbol does not have global or weak binding, ignore it.
265 uint8_t Binding = RawSym.getBinding();
266 if (!(Binding == STB_GLOBAL || Binding == STB_WEAK))
267 continue;
268 // If a symbol doesn't have default or protected visibility, ignore it.
269 uint8_t Visibility = RawSym.getVisibility();
270 if (!(Visibility == STV_DEFAULT || Visibility == STV_PROTECTED))
271 continue;
272 // Create an ELFSymbol and populate it with information from the symbol
273 // table entry.
274 Expected<StringRef> SymName = terminatedSubstr(DynStr, RawSym.st_name);
275 if (!SymName)
276 return SymName.takeError();
277 ELFSymbol Sym = createELFSym<ELFT>(*SymName, RawSym);
278 TargetStub.Symbols.insert(std::move(Sym));
279 // TODO: Populate symbol warning.
280 }
281 return Error::success();
282}
283
Armando Montanez31f0f652019-01-03 18:32:36 +0000284/// Returns a new ELFStub with all members populated from an ELFObjectFile.
285/// @param ElfObj Source ELFObjectFile.
286template <class ELFT>
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000287static Expected<std::unique_ptr<ELFStub>>
Armando Montanez31f0f652019-01-03 18:32:36 +0000288buildStub(const ELFObjectFile<ELFT> &ElfObj) {
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000289 using Elf_Dyn_Range = typename ELFT::DynRange;
290 using Elf_Phdr_Range = typename ELFT::PhdrRange;
Armando Montanez8367b072019-01-24 22:39:21 +0000291 using Elf_Sym_Range = typename ELFT::SymRange;
292 using Elf_Sym = typename ELFT::Sym;
Armando Montanez31f0f652019-01-03 18:32:36 +0000293 std::unique_ptr<ELFStub> DestStub = make_unique<ELFStub>();
294 const ELFFile<ELFT> *ElfFile = ElfObj.getELFFile();
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000295 // Fetch .dynamic table.
296 Expected<Elf_Dyn_Range> DynTable = ElfFile->dynamicEntries();
297 if (!DynTable) {
298 return DynTable.takeError();
299 }
Armando Montanez31f0f652019-01-03 18:32:36 +0000300
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000301 // Fetch program headers.
302 Expected<Elf_Phdr_Range> PHdrs = ElfFile->program_headers();
303 if (!PHdrs) {
304 return PHdrs.takeError();
305 }
306
307 // Collect relevant .dynamic entries.
308 DynamicEntries DynEnt;
309 if (Error Err = populateDynamic<ELFT>(DynEnt, *DynTable))
310 return std::move(Err);
311
Armando Montanez8367b072019-01-24 22:39:21 +0000312 // Get pointer to in-memory location of .dynstr section.
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000313 Expected<const uint8_t *> DynStrPtr =
314 ElfFile->toMappedAddr(DynEnt.StrTabAddr);
315 if (!DynStrPtr)
316 return appendToError(DynStrPtr.takeError(),
317 "when locating .dynstr section contents");
318
319 StringRef DynStr(reinterpret_cast<const char *>(DynStrPtr.get()),
320 DynEnt.StrSize);
321
322 // Populate Arch from ELF header.
Armando Montanez31f0f652019-01-03 18:32:36 +0000323 DestStub->Arch = ElfFile->getHeader()->e_machine;
324
Armando Montanezfe7ab3c2019-01-16 17:47:16 +0000325 // Populate SoName from .dynamic entries and dynamic string table.
326 if (DynEnt.SONameOffset.hasValue()) {
327 Expected<StringRef> NameOrErr =
328 terminatedSubstr(DynStr, *DynEnt.SONameOffset);
329 if (!NameOrErr) {
330 return appendToError(NameOrErr.takeError(), "when reading DT_SONAME");
331 }
332 DestStub->SoName = *NameOrErr;
333 }
334
Armando Montanez56d18122019-01-18 20:56:03 +0000335 // Populate NeededLibs from .dynamic entries and dynamic string table.
336 for (uint64_t NeededStrOffset : DynEnt.NeededLibNames) {
337 Expected<StringRef> LibNameOrErr =
338 terminatedSubstr(DynStr, NeededStrOffset);
339 if (!LibNameOrErr) {
340 return appendToError(LibNameOrErr.takeError(), "when reading DT_NEEDED");
341 }
342 DestStub->NeededLibs.push_back(*LibNameOrErr);
343 }
344
Armando Montanez8367b072019-01-24 22:39:21 +0000345 // Populate Symbols from .dynsym table and dynamic string table.
346 Expected<uint64_t> SymCount = getNumSyms(DynEnt, *ElfFile);
347 if (!SymCount)
348 return SymCount.takeError();
349 if (*SymCount > 0) {
350 // Get pointer to in-memory location of .dynsym section.
351 Expected<const uint8_t *> DynSymPtr =
352 ElfFile->toMappedAddr(DynEnt.DynSymAddr);
353 if (!DynSymPtr)
354 return appendToError(DynSymPtr.takeError(),
355 "when locating .dynsym section contents");
356 Elf_Sym_Range DynSyms =
357 ArrayRef<Elf_Sym>(reinterpret_cast<const Elf_Sym *>(*DynSymPtr),
358 *SymCount);
359 Error SymReadError = populateSymbols<ELFT>(*DestStub, DynSyms, DynStr);
360 if (SymReadError)
361 return appendToError(std::move(SymReadError),
362 "when reading dynamic symbols");
363 }
Armando Montanez31f0f652019-01-03 18:32:36 +0000364
365 return std::move(DestStub);
366}
367
368Expected<std::unique_ptr<ELFStub>> readELFFile(MemoryBufferRef Buf) {
369 Expected<std::unique_ptr<Binary>> BinOrErr = createBinary(Buf);
370 if (!BinOrErr) {
371 return BinOrErr.takeError();
372 }
373
374 Binary *Bin = BinOrErr->get();
375 if (auto Obj = dyn_cast<ELFObjectFile<ELF32LE>>(Bin)) {
376 return buildStub(*Obj);
377 } else if (auto Obj = dyn_cast<ELFObjectFile<ELF64LE>>(Bin)) {
378 return buildStub(*Obj);
379 } else if (auto Obj = dyn_cast<ELFObjectFile<ELF32BE>>(Bin)) {
380 return buildStub(*Obj);
381 } else if (auto Obj = dyn_cast<ELFObjectFile<ELF64BE>>(Bin)) {
382 return buildStub(*Obj);
383 }
384
385 return createStringError(errc::not_supported, "Unsupported binary format");
386}
387
388} // end namespace elfabi
389} // end namespace llvm