LLVM 20.0.0git
MachOObject.cpp
Go to the documentation of this file.
1//===- MachOObject.cpp - Mach-O object file model ---------------*- 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
9#include "MachOObject.h"
12#include <unordered_set>
13
14using namespace llvm;
15using namespace llvm::objcopy::macho;
16
18 : Segname(SegName), Sectname(SectName),
19 CanonicalName((Twine(SegName) + Twine(',') + SectName).str()) {}
20
22 : Segname(SegName), Sectname(SectName),
23 CanonicalName((Twine(SegName) + Twine(',') + SectName).str()),
25
27 assert(Index < Symbols.size() && "invalid symbol index");
28 return Symbols[Index].get();
29}
30
32 return const_cast<SymbolEntry *>(
33 static_cast<const SymbolTable *>(this)->getSymbolByIndex(Index));
34}
35
37 function_ref<bool(const std::unique_ptr<SymbolEntry> &)> ToRemove) {
39}
40
42 static constexpr char TextSegmentName[] = "__TEXT";
43 // Update indices of special load commands
44 for (size_t Index = 0, Size = LoadCommands.size(); Index < Size; ++Index) {
46 switch (LC.MachOLoadCommand.load_command_data.cmd) {
47 case MachO::LC_CODE_SIGNATURE:
49 break;
50 case MachO::LC_SEGMENT:
51 if (StringRef(LC.MachOLoadCommand.segment_command_data.segname) ==
52 TextSegmentName)
54 break;
55 case MachO::LC_SEGMENT_64:
56 if (StringRef(LC.MachOLoadCommand.segment_command_64_data.segname) ==
57 TextSegmentName)
59 break;
60 case MachO::LC_SYMTAB:
62 break;
63 case MachO::LC_DYSYMTAB:
65 break;
66 case MachO::LC_DYLD_INFO:
67 case MachO::LC_DYLD_INFO_ONLY:
69 break;
70 case MachO::LC_DATA_IN_CODE:
72 break;
73 case MachO::LC_LINKER_OPTIMIZATION_HINT:
75 break;
76 case MachO::LC_FUNCTION_STARTS:
78 break;
79 case MachO::LC_DYLIB_CODE_SIGN_DRS:
81 break;
82 case MachO::LC_DYLD_CHAINED_FIXUPS:
84 break;
85 case MachO::LC_DYLD_EXPORTS_TRIE:
87 break;
88 }
89 }
90}
91
93 function_ref<bool(const LoadCommand &)> ToRemove) {
94 auto It = std::stable_partition(
95 LoadCommands.begin(), LoadCommands.end(),
96 [&](const LoadCommand &LC) { return !ToRemove(LC); });
97 LoadCommands.erase(It, LoadCommands.end());
98
100 return Error::success();
101}
102
104 function_ref<bool(const std::unique_ptr<Section> &)> ToRemove) {
105 DenseMap<uint32_t, const Section *> OldIndexToSection;
106 uint32_t NextSectionIndex = 1;
107 for (LoadCommand &LC : LoadCommands) {
108 auto It = std::stable_partition(
109 std::begin(LC.Sections), std::end(LC.Sections),
110 [&](const std::unique_ptr<Section> &Sec) { return !ToRemove(Sec); });
111 for (auto I = LC.Sections.begin(), End = It; I != End; ++I) {
112 OldIndexToSection[(*I)->Index] = I->get();
113 (*I)->Index = NextSectionIndex++;
114 }
115 LC.Sections.erase(It, LC.Sections.end());
116 }
117
118 auto IsDead = [&](const std::unique_ptr<SymbolEntry> &S) -> bool {
119 std::optional<uint32_t> Section = S->section();
120 return (Section && !OldIndexToSection.count(*Section));
121 };
122
124 for (const std::unique_ptr<SymbolEntry> &Sym : SymTable.Symbols)
125 if (IsDead(Sym))
126 DeadSymbols.insert(Sym.get());
127
128 for (const LoadCommand &LC : LoadCommands)
129 for (const std::unique_ptr<Section> &Sec : LC.Sections)
130 for (const RelocationInfo &R : Sec->Relocations)
131 if (R.Symbol && *R.Symbol && DeadSymbols.count(*R.Symbol))
132 return createStringError(std::errc::invalid_argument,
133 "symbol '%s' defined in section with index "
134 "'%u' cannot be removed because it is "
135 "referenced by a relocation in section '%s'",
136 (*R.Symbol)->Name.c_str(),
137 *((*R.Symbol)->section()),
138 Sec->CanonicalName.c_str());
140 for (std::unique_ptr<SymbolEntry> &S : SymTable.Symbols)
141 if (S->section())
142 S->n_sect = OldIndexToSection[S->n_sect]->Index;
143 return Error::success();
144}
145
147 uint64_t HeaderSize =
149 uint64_t Addr = HeaderSize + Header.SizeOfCmds;
150 for (const LoadCommand &LC : LoadCommands) {
151 const MachO::macho_load_command &MLC = LC.MachOLoadCommand;
152 switch (MLC.load_command_data.cmd) {
153 case MachO::LC_SEGMENT:
154 Addr = std::max(Addr,
155 static_cast<uint64_t>(MLC.segment_command_data.vmaddr) +
156 MLC.segment_command_data.vmsize);
157 break;
158 case MachO::LC_SEGMENT_64:
159 Addr = std::max(Addr, MLC.segment_command_64_data.vmaddr +
160 MLC.segment_command_64_data.vmsize);
161 break;
162 default:
163 continue;
164 }
165 }
166 return Addr;
167}
168
169template <typename SegmentType>
170static void
172 StringRef SegName, uint64_t SegVMAddr, uint64_t SegVMSize) {
173 assert(SegName.size() <= sizeof(Seg.segname) && "too long segment name");
174 memset(&Seg, 0, sizeof(SegmentType));
175 Seg.cmd = CmdType;
176 strncpy(Seg.segname, SegName.data(), SegName.size());
177 Seg.maxprot |=
179 Seg.initprot |=
181 Seg.vmaddr = SegVMAddr;
182 Seg.vmsize = SegVMSize;
183}
184
186 LoadCommand LC;
187 const uint64_t SegVMAddr = nextAvailableSegmentAddress();
188 if (is64Bit())
189 constructSegment(LC.MachOLoadCommand.segment_command_64_data,
190 MachO::LC_SEGMENT_64, SegName, SegVMAddr, SegVMSize);
191 else
192 constructSegment(LC.MachOLoadCommand.segment_command_data,
193 MachO::LC_SEGMENT, SegName, SegVMAddr, SegVMSize);
194
195 LoadCommands.push_back(std::move(LC));
196 return LoadCommands.back();
197}
198
199/// Extracts a segment name from a string which is possibly non-null-terminated.
200static StringRef extractSegmentName(const char *SegName) {
201 return StringRef(SegName,
202 strnlen(SegName, sizeof(MachO::segment_command::segname)));
203}
204
205std::optional<StringRef> LoadCommand::getSegmentName() const {
207 switch (MLC.load_command_data.cmd) {
208 case MachO::LC_SEGMENT:
209 return extractSegmentName(MLC.segment_command_data.segname);
210 case MachO::LC_SEGMENT_64:
211 return extractSegmentName(MLC.segment_command_64_data.segname);
212 default:
213 return std::nullopt;
214 }
215}
216
217std::optional<uint64_t> LoadCommand::getSegmentVMAddr() const {
219 switch (MLC.load_command_data.cmd) {
220 case MachO::LC_SEGMENT:
221 return MLC.segment_command_data.vmaddr;
222 case MachO::LC_SEGMENT_64:
223 return MLC.segment_command_64_data.vmaddr;
224 default:
225 return std::nullopt;
226 }
227}
ReachingDefAnalysis InstSet & ToRemove
T Content
uint64_t Addr
uint64_t Size
bool End
Definition: ELF_riscv.cpp:480
Symbol * Sym
Definition: ELF_riscv.cpp:479
#define I(x, y, z)
Definition: MD5.cpp:58
static void constructSegment(SegmentType &Seg, llvm::MachO::LoadCommandType CmdType, StringRef SegName, uint64_t SegVMAddr, uint64_t SegVMSize)
static StringRef extractSegmentName(const char *SegName)
Extracts a segment name from a string which is possibly non-null-terminated.
assert(ImpDefSCC.getReg()==AMDGPU::SCC &&ImpDefSCC.isDef())
bool IsDead
This file defines the SmallPtrSet class.
size_type count(const_arg_type_t< KeyT > Val) const
Return 1 if the specified key is in the map, 0 otherwise.
Definition: DenseMap.h:151
Lightweight error class with error context and mandatory checking.
Definition: Error.h:160
static ErrorSuccess success()
Create a success value.
Definition: Error.h:337
size_type count(ConstPtrType Ptr) const
count - Return 1 if the specified pointer is in the set, 0 otherwise.
Definition: SmallPtrSet.h:436
std::pair< iterator, bool > insert(PtrType Ptr)
Inserts Ptr if and only if there is no element in the container equal to Ptr.
Definition: SmallPtrSet.h:368
SmallPtrSet - This class implements a set which is optimized for holding SmallSize or less elements.
Definition: SmallPtrSet.h:503
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:50
constexpr size_t size() const
size - Get the string size.
Definition: StringRef.h:137
constexpr const char * data() const
data - Get a pointer to the start of the string (which may not be null terminated).
Definition: StringRef.h:131
Twine - A lightweight data structure for efficiently representing the concatenation of temporary valu...
Definition: Twine.h:81
An efficient, type-erasing, non-owning reference to a callable.
@ VM_PROT_EXECUTE
Definition: MachO.h:497
@ VM_PROT_READ
Definition: MachO.h:497
@ VM_PROT_WRITE
Definition: MachO.h:497
LoadCommandType
Definition: MachO.h:98
This is an optimization pass for GlobalISel generic memory operations.
Definition: AddressRanges.h:18
Error createStringError(std::error_code EC, char const *Fmt, const Ts &... Vals)
Create formatted StringError object.
Definition: Error.h:1286
void erase_if(Container &C, UnaryPredicate P)
Provide a container algorithm similar to C++ Library Fundamentals v2's erase_if which is equivalent t...
Definition: STLExtras.h:2057
MachO::macho_load_command MachOLoadCommand
Definition: MachOObject.h:82
std::optional< StringRef > getSegmentName() const
std::optional< uint64_t > getSegmentVMAddr() const
std::optional< size_t > FunctionStartsCommandIndex
The index LC_FUNCTION_STARTS load command if present.
Definition: MachOObject.h:335
Error removeLoadCommands(function_ref< bool(const LoadCommand &)> ToRemove)
Definition: MachOObject.cpp:92
std::optional< size_t > ChainedFixupsCommandIndex
The index LC_DYLD_CHAINED_FIXUPS load command if present.
Definition: MachOObject.h:337
std::optional< size_t > ExportsTrieCommandIndex
The index LC_DYLD_EXPORTS_TRIE load command if present.
Definition: MachOObject.h:339
Error removeSections(function_ref< bool(const std::unique_ptr< Section > &)> ToRemove)
std::optional< size_t > DylibCodeSignDRsIndex
The index of LC_DYLIB_CODE_SIGN_DRS load command if present.
Definition: MachOObject.h:323
std::optional< size_t > SymTabCommandIndex
The index of LC_SYMTAB load command if present.
Definition: MachOObject.h:325
std::optional< size_t > DyLdInfoCommandIndex
The index of LC_DYLD_INFO or LC_DYLD_INFO_ONLY load command if present.
Definition: MachOObject.h:327
std::vector< LoadCommand > LoadCommands
Definition: MachOObject.h:300
std::optional< size_t > DataInCodeCommandIndex
The index LC_DATA_IN_CODE load command if present.
Definition: MachOObject.h:331
std::optional< size_t > DySymTabCommandIndex
The index LC_DYSYMTAB load command if present.
Definition: MachOObject.h:329
std::optional< size_t > TextSegmentCommandIndex
The index of the LC_SEGMENT or LC_SEGMENT_64 load command corresponding to the __TEXT segment.
Definition: MachOObject.h:342
uint64_t nextAvailableSegmentAddress() const
std::optional< size_t > CodeSignatureCommandIndex
The index of LC_CODE_SIGNATURE load command if present.
Definition: MachOObject.h:321
LoadCommand & addSegment(StringRef SegName, uint64_t SegVMSize)
Creates a new segment load command in the object and returns a reference to the newly created load co...
std::optional< size_t > LinkerOptimizationHintCommandIndex
The index of LC_LINKER_OPTIMIZATIN_HINT load command if present.
Definition: MachOObject.h:333
Section(StringRef SegName, StringRef SectName)
Definition: MachOObject.cpp:17
The location of the symbol table inside the binary is described by LC_SYMTAB load command.
Definition: MachOObject.h:134
const SymbolEntry * getSymbolByIndex(uint32_t Index) const
Definition: MachOObject.cpp:26
std::vector< std::unique_ptr< SymbolEntry > > Symbols
Definition: MachOObject.h:135
void removeSymbols(function_ref< bool(const std::unique_ptr< SymbolEntry > &)> ToRemove)
Definition: MachOObject.cpp:36