1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-24 19:52:54 +01:00

[MC] Adjust StringTableBuilder for linked Mach-O binaries

LD64 emits string tables which start with a space and a zero byte.
This diff adjusts StringTableBuilder for linked Mach-O binaries to match LD64's behavior.

Test plan: make check-all

Differential revision: https://reviews.llvm.org/D89561
This commit is contained in:
Alexander Shaposhnikov 2020-10-22 18:03:40 -07:00
parent b3e197724b
commit dceff17612
12 changed files with 241 additions and 58 deletions

View File

@ -114,7 +114,7 @@ class MachObjectWriter : public MCObjectWriter {
/// \name Symbol Table Data
/// @{
StringTableBuilder StringTable{StringTableBuilder::MachO};
StringTableBuilder StringTable;
std::vector<MachSymbolData> LocalSymbolData;
std::vector<MachSymbolData> ExternalSymbolData;
std::vector<MachSymbolData> UndefinedSymbolData;
@ -129,6 +129,8 @@ public:
MachObjectWriter(std::unique_ptr<MCMachObjectTargetWriter> MOTW,
raw_pwrite_stream &OS, bool IsLittleEndian)
: TargetObjectWriter(std::move(MOTW)),
StringTable(TargetObjectWriter->is64Bit() ? StringTableBuilder::MachO64
: StringTableBuilder::MachO),
W(OS, IsLittleEndian ? support::little : support::big) {}
support::endian::Writer W;

View File

@ -22,7 +22,17 @@ class raw_ostream;
/// Utility for building string tables with deduplicated suffixes.
class StringTableBuilder {
public:
enum Kind { ELF, WinCOFF, MachO, RAW, DWARF, XCOFF };
enum Kind {
ELF,
WinCOFF,
MachO,
MachO64,
MachOLinked,
MachO64Linked,
RAW,
DWARF,
XCOFF
};
private:
DenseMap<CachedHashStringRef, size_t> StringIndexMap;

View File

@ -33,7 +33,12 @@ void StringTableBuilder::initSize() {
case DWARF:
Size = 0;
break;
case MachOLinked:
case MachO64Linked:
Size = 2;
break;
case MachO:
case MachO64:
case ELF:
// Start the table with a NUL byte.
Size = 1;
@ -161,8 +166,16 @@ void StringTableBuilder::finalizeStringTable(bool Optimize) {
}
}
if (K == MachO)
if (K == MachO || K == MachOLinked)
Size = alignTo(Size, 4); // Pad to multiple of 4.
if (K == MachO64 || K == MachO64Linked)
Size = alignTo(Size, 8); // Pad to multiple of 8.
// According to ld64 the string table of a final linked Mach-O binary starts
// with " ", i.e. the first byte is ' ' and the second byte is zero. In
// 'initSize()' we reserved the first two bytes for holding this string.
if (K == MachOLinked || K == MachO64Linked)
StringIndexMap[CachedHashStringRef(" ")] = 0;
// The first byte in an ELF string table must be null, according to the ELF
// specification. In 'initSize()' we reserved the first byte to hold null for

View File

@ -83,7 +83,7 @@ LoadCommands:
symoff: 784
nsyms: 2
stroff: 816
strsize: 36
strsize: 40
- cmd: LC_DYSYMTAB
cmdsize: 80
ilocalsym: 0
@ -121,4 +121,8 @@ LinkEditData:
- _compilerrt_abort_impl
- ___absvdi2
- ''
- ''
- ''
- ''
- ''
...

View File

@ -83,7 +83,7 @@ LoadCommands:
symoff: 784
nsyms: 2
stroff: 816
strsize: 36
strsize: 40
- cmd: LC_DYSYMTAB
cmdsize: 80
ilocalsym: 0
@ -121,4 +121,8 @@ LinkEditData:
- _compilerrt_abort_impl
- ___absvdi2
- ''
- ''
- ''
- ''
- ''
...

View File

@ -5,7 +5,7 @@
# CHECK: cmd LC_CODE_SIGNATURE
# CHECK-NEXT: cmdsize 16
# CHECK-NEXT: dataoff 124
# CHECK-NEXT: dataoff 128
# CHECK-NEXT: datasize 16
# RUN: llvm-objcopy %t %t.copy
@ -28,13 +28,13 @@ LoadCommands:
vmaddr: 4294979584
vmsize: 4096
fileoff: 120
filesize: 20
filesize: 24
maxprot: 7
initprot: 1
nsects: 0
flags: 0
- cmd: LC_CODE_SIGNATURE
cmdsize: 16
dataoff: 124
dataoff: 128
datasize: 16
...

View File

@ -19,7 +19,7 @@
# NO-SWIFT-SYMBOLS: Symbols [
# NO-SWIFT-SYMBOLS-NEXT: Symbol {
# NO-SWIFT-SYMBOLS-NEXT: Name: _main (1)
# NO-SWIFT-SYMBOLS-NEXT: Name: _main (2)
# NO-SWIFT-SYMBOLS-NEXT: Extern
# NO-SWIFT-SYMBOLS-NEXT: Type: Section (0xE)
# NO-SWIFT-SYMBOLS-NEXT: Section: __text (0x1)
@ -73,7 +73,7 @@
# SWIFT-SYMBOLS: Symbols [
# SWIFT-SYMBOLS-NEXT: Symbol {
# SWIFT-SYMBOLS-NEXT: Name: _$S1a13PublicSymbol1Sivp (26)
# SWIFT-SYMBOLS-NEXT: Name: _$S1a13PublicSymbol1Sivp (27)
# SWIFT-SYMBOLS-NEXT: Extern
# SWIFT-SYMBOLS-NEXT: Type: Section (0xE)
# SWIFT-SYMBOLS-NEXT: Section: __text (0x1)
@ -83,7 +83,7 @@
# SWIFT-SYMBOLS-NEXT: Value: 0x100001160
# SWIFT-SYMBOLS-NEXT: }
# SWIFT-SYMBOLS-NEXT: Symbol {
# SWIFT-SYMBOLS-NEXT: Name: _$s1a13PublicSymbol2Sivp (1)
# SWIFT-SYMBOLS-NEXT: Name: _$s1a13PublicSymbol2Sivp (2)
# SWIFT-SYMBOLS-NEXT: Extern
# SWIFT-SYMBOLS-NEXT: Type: Section (0xE)
# SWIFT-SYMBOLS-NEXT: Section: __text (0x1)
@ -93,7 +93,7 @@
# SWIFT-SYMBOLS-NEXT: Value: 0x100001168
# SWIFT-SYMBOLS-NEXT: }
# SWIFT-SYMBOLS-NEXT: Symbol {
# SWIFT-SYMBOLS-NEXT: Name: _main (51)
# SWIFT-SYMBOLS-NEXT: Name: _main (52)
# SWIFT-SYMBOLS-NEXT: Extern
# SWIFT-SYMBOLS-NEXT: Type: Section (0xE)
# SWIFT-SYMBOLS-NEXT: Section: __text (0x1)

View File

@ -240,20 +240,19 @@ LinkEditData:
Other: 0x0000000000000000
ImportName: ''
NameList:
- n_strx: 18
- n_strx: 19
n_type: 0x0F
n_sect: 1
n_desc: 16
n_value: 4294967296
- n_strx: 1
- n_strx: 2
n_type: 0x01
n_sect: 0
n_desc: 256
n_value: 0
StringTable:
- ''
- ' '
- dyld_stub_binder
- __mh_execute_header
- ''
- ''
...

View File

@ -24,7 +24,7 @@
# CHECK: Symbols [
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _PrivateSymbol (169)
# CHECK-NEXT: Name: _PrivateSymbol (170)
# CHECK-NEXT: Type: Section (0xE)
# CHECK-NEXT: Section: __bss (0x4)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -33,7 +33,7 @@
# CHECK-NEXT: Value: 0x100001008
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _PrivateExternalSymbol (121)
# CHECK-NEXT: Name: _PrivateExternalSymbol (122)
# CHECK-NEXT: PrivateExtern
# CHECK-NEXT: Type: Section (0xE)
# CHECK-NEXT: Section: __common (0x5)
@ -43,7 +43,7 @@
# CHECK-NEXT: Value: 0x100001010
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: /Users/aaaaaaaa/ (191)
# CHECK-NEXT: Name: /Users/aaaaaaaa/ (192)
# CHECK-NEXT: Type: SymDebugTable (0x64)
# CHECK-NEXT: Section: (0x0)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -52,7 +52,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: main.c (184)
# CHECK-NEXT: Name: main.c (185)
# CHECK-NEXT: Type: SymDebugTable (0x64)
# CHECK-NEXT: Section: (0x0)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -61,7 +61,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: /var/folders/1d/zyfdpp7j2995h5hqspjy28bc0000gn/T/main-c5ac21.o (38)
# CHECK-NEXT: Name: /var/folders/1d/zyfdpp7j2995h5hqspjy28bc0000gn/T/main-c5ac21.o (39)
# CHECK-NEXT: Type: SymDebugTable (0x66)
# CHECK-NEXT: Section: (0x3)
# CHECK-NEXT: RefType: ReferenceFlagUndefinedLazy (0x1)
@ -70,7 +70,7 @@
# CHECK-NEXT: Value: 0x5EA74C81
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: (207)
# CHECK-NEXT: Name: (208)
# CHECK-NEXT: Type: SymDebugTable (0x2E)
# CHECK-NEXT: Section: (0x1)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -79,7 +79,7 @@
# CHECK-NEXT: Value: 0x100000F80
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _main (101)
# CHECK-NEXT: Name: _main (102)
# CHECK-NEXT: Type: SymDebugTable (0x24)
# CHECK-NEXT: Section: (0x1)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -88,7 +88,7 @@
# CHECK-NEXT: Value: 0x100000F80
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: (207)
# CHECK-NEXT: Name: (208)
# CHECK-NEXT: Type: SymDebugTable (0x24)
# CHECK-NEXT: Section: (0x0)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -97,7 +97,7 @@
# CHECK-NEXT: Value: 0x2D
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: (207)
# CHECK-NEXT: Name: (208)
# CHECK-NEXT: Type: SymDebugTable (0x4E)
# CHECK-NEXT: Section: (0x1)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -106,7 +106,7 @@
# CHECK-NEXT: Value: 0x2D
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _PrivateSymbol (169)
# CHECK-NEXT: Name: _PrivateSymbol (170)
# CHECK-NEXT: Type: SymDebugTable (0x26)
# CHECK-NEXT: Section: (0x4)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -115,7 +115,7 @@
# CHECK-NEXT: Value: 0x100001008
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _CommonSymbol (107)
# CHECK-NEXT: Name: _CommonSymbol (108)
# CHECK-NEXT: Type: SymDebugTable (0x20)
# CHECK-NEXT: Section: (0x0)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -124,7 +124,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _PrivateExternalSymbol (121)
# CHECK-NEXT: Name: _PrivateExternalSymbol (122)
# CHECK-NEXT: Type: SymDebugTable (0x20)
# CHECK-NEXT: Section: (0x0)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -133,7 +133,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: (207)
# CHECK-NEXT: Name: (208)
# CHECK-NEXT: Type: SymDebugTable (0x64)
# CHECK-NEXT: Section: (0x1)
# CHECK-NEXT: RefType: UndefinedNonLazy (0x0)
@ -142,7 +142,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _CommonSymbol (107)
# CHECK-NEXT: Name: _CommonSymbol (108)
# CHECK-NEXT: Extern
# CHECK-NEXT: Type: Section (0xE)
# CHECK-NEXT: Section: __common (0x5)
@ -152,7 +152,7 @@
# CHECK-NEXT: Value: 0x10000100C
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: __mh_execute_header (18)
# CHECK-NEXT: Name: __mh_execute_header (19)
# CHECK-NEXT: Extern
# CHECK-NEXT: Type: Section (0xE)
# CHECK-NEXT: Section: __text (0x1)
@ -163,7 +163,7 @@
# CHECK-NEXT: Value: 0x100000000
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _main (101)
# CHECK-NEXT: Name: _main (102)
# CHECK-NEXT: Extern
# CHECK-NEXT: Type: Section (0xE)
# CHECK-NEXT: Section: __text (0x1)
@ -173,7 +173,7 @@
# CHECK-NEXT: Value: 0x100000F80
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: _UndefinedExternalSymbol (144)
# CHECK-NEXT: Name: _UndefinedExternalSymbol (145)
# CHECK-NEXT: Extern
# CHECK-NEXT: Type: Undef (0x0)
# CHECK-NEXT: Section: (0x0)
@ -184,7 +184,7 @@
# CHECK-NEXT: Value: 0x0
# CHECK-NEXT: }
# CHECK-NEXT: Symbol {
# CHECK-NEXT: Name: dyld_stub_binder (1)
# CHECK-NEXT: Name: dyld_stub_binder (2)
# CHECK-NEXT: Extern
# CHECK-NEXT: Type: Undef (0x0)
# CHECK-NEXT: Section: (0x0)
@ -299,7 +299,7 @@ LoadCommands:
vmaddr: 4294975488
vmsize: 4096
fileoff: 8192
filesize: 500
filesize: 508
maxprot: 1
initprot: 1
nsects: 0
@ -309,7 +309,7 @@ LoadCommands:
symoff: 8192
nsyms: 18
stroff: 8484
strsize: 208
strsize: 216
- cmd: LC_DYSYMTAB
cmdsize: 80
ilocalsym: 0
@ -332,98 +332,98 @@ LoadCommands:
nlocrel: 0
LinkEditData:
NameList:
- n_strx: 169
- n_strx: 170
n_type: 0x0E
n_sect: 4
n_desc: 0
n_value: 4294971400
- n_strx: 121
- n_strx: 122
n_type: 0x1E
n_sect: 5
n_desc: 0
n_value: 4294971408
- n_strx: 191
- n_strx: 192
n_type: 0x64
n_sect: 0
n_desc: 0
n_value: 0
- n_strx: 184
- n_strx: 185
n_type: 0x64
n_sect: 0
n_desc: 0
n_value: 0
- n_strx: 38
- n_strx: 39
n_type: 0x66
n_sect: 3
n_desc: 1
n_value: 1588022401
- n_strx: 207
- n_strx: 208
n_type: 0x2E
n_sect: 1
n_desc: 0
n_value: 4294971264
- n_strx: 101
- n_strx: 102
n_type: 0x24
n_sect: 1
n_desc: 0
n_value: 4294971264
- n_strx: 207
- n_strx: 208
n_type: 0x24
n_sect: 0
n_desc: 0
n_value: 45
- n_strx: 207
- n_strx: 208
n_type: 0x4E
n_sect: 1
n_desc: 0
n_value: 45
- n_strx: 169
- n_strx: 170
n_type: 0x26
n_sect: 4
n_desc: 0
n_value: 4294971400
- n_strx: 107
- n_strx: 108
n_type: 0x20
n_sect: 0
n_desc: 0
n_value: 0
- n_strx: 121
- n_strx: 122
n_type: 0x20
n_sect: 0
n_desc: 0
n_value: 0
- n_strx: 207
- n_strx: 208
n_type: 0x64
n_sect: 1
n_desc: 0
n_value: 0
- n_strx: 107
- n_strx: 108
n_type: 0x0F
n_sect: 5
n_desc: 0
n_value: 4294971404
- n_strx: 18
- n_strx: 19
n_type: 0x0F
n_sect: 1
n_desc: 16
n_value: 4294967296
- n_strx: 101
- n_strx: 102
n_type: 0x0F
n_sect: 1
n_desc: 0
n_value: 4294971264
- n_strx: 144
- n_strx: 145
n_type: 0x01
n_sect: 0
n_desc: 65024
n_value: 0
- n_strx: 1
- n_strx: 2
n_type: 0x01
n_sect: 0
n_desc: 256
n_value: 0
StringTable:
- ''
- ' '
- dyld_stub_binder
- __mh_execute_header
- '/var/folders/1d/zyfdpp7j2995h5hqspjy28bc0000gn/T/main-c5ac21.o'
@ -434,4 +434,11 @@ LinkEditData:
- _PrivateSymbol
- main.c
- '/Users/aaaaaaaa/'
- ''
- ''
- ''
- ''
- ''
- ''
- ''
...

View File

@ -252,6 +252,8 @@ Error MachOLayoutBuilder::layoutTail(uint64_t Offset) {
sizeof(uint32_t) * O.IndirectSymTable.Symbols.size();
uint64_t StartOfCodeSignature =
StartOfSymbolStrings + StrTableBuilder.getSize();
if (O.CodeSignatureCommandIndex)
StartOfCodeSignature = alignTo(StartOfCodeSignature, 16);
uint64_t LinkEditSize =
(StartOfCodeSignature + O.CodeSignature.Data.size()) - StartOfLinkEdit;

View File

@ -23,7 +23,7 @@ class MachOLayoutBuilder {
// Points to the __LINKEDIT segment if it exists.
MachO::macho_load_command *LinkEditLoadCommand = nullptr;
StringTableBuilder StrTableBuilder{StringTableBuilder::MachO};
StringTableBuilder StrTableBuilder;
uint32_t computeSizeOfCmds() const;
void constructStringTable();
@ -33,9 +33,18 @@ class MachOLayoutBuilder {
uint64_t layoutRelocations(uint64_t Offset);
Error layoutTail(uint64_t Offset);
static StringTableBuilder::Kind getStringTableBuilderKind(const Object &O,
bool Is64Bit) {
if (O.Header.FileType == MachO::HeaderFileType::MH_OBJECT)
return Is64Bit ? StringTableBuilder::MachO64 : StringTableBuilder::MachO;
return Is64Bit ? StringTableBuilder::MachO64Linked
: StringTableBuilder::MachOLinked;
}
public:
MachOLayoutBuilder(Object &O, bool Is64Bit, uint64_t PageSize)
: O(O), Is64Bit(Is64Bit), PageSize(PageSize) {}
: O(O), Is64Bit(Is64Bit), PageSize(PageSize),
StrTableBuilder(getStringTableBuilderKind(O, Is64Bit)) {}
// Recomputes and updates fields in the given object such as file offsets.
Error layout();

View File

@ -103,4 +103,137 @@ TEST(StringTableBuilderTest, ELFInOrder) {
EXPECT_EQ(9U, B.getOffset("foobar"));
}
TEST(StringTableBuilderTest, MachOInOrder) {
StringTableBuilder B(StringTableBuilder::MachO);
B.add("foo");
B.add("bar");
B.add("fooba");
B.finalizeInOrder();
std::string Expected;
Expected += '\x00';
Expected += "foo";
Expected += '\x00';
Expected += "bar";
Expected += '\x00';
Expected += "fooba";
Expected += '\x00';
// Mach-O pads to 4 bytes
Expected += '\x00';
SmallString<64> Data;
raw_svector_ostream OS(Data);
B.write(OS);
EXPECT_EQ(Expected, Data);
EXPECT_EQ(1U, B.getOffset("foo"));
EXPECT_EQ(5U, B.getOffset("bar"));
EXPECT_EQ(9U, B.getOffset("fooba"));
}
TEST(StringTableBuilderTest, MachO64InOrder) {
StringTableBuilder B(StringTableBuilder::MachO64);
B.add("foo");
B.add("bar");
B.add("f");
B.finalizeInOrder();
std::string Expected;
Expected += '\x00';
Expected += "foo";
Expected += '\x00';
Expected += "bar";
Expected += '\x00';
Expected += "f";
Expected += '\x00';
// 64 bit Mach-O pads to 8 bytes
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
SmallString<64> Data;
raw_svector_ostream OS(Data);
B.write(OS);
EXPECT_EQ(Expected, Data);
EXPECT_EQ(1U, B.getOffset("foo"));
EXPECT_EQ(5U, B.getOffset("bar"));
EXPECT_EQ(9U, B.getOffset("f"));
}
TEST(StringTableBuilderTest, MachOLinkedInOrder) {
StringTableBuilder B(StringTableBuilder::MachOLinked);
B.add("foo");
B.add("bar");
B.add("foob");
B.finalizeInOrder();
std::string Expected;
Expected += ' ';
Expected += '\x00';
Expected += "foo";
Expected += '\x00';
Expected += "bar";
Expected += '\x00';
Expected += "foob";
Expected += '\x00';
// Mach-O pads to 4 bytes
Expected += '\x00';
SmallString<64> Data;
raw_svector_ostream OS(Data);
B.write(OS);
EXPECT_EQ(Expected, Data);
EXPECT_EQ(2U, B.getOffset("foo"));
EXPECT_EQ(6U, B.getOffset("bar"));
EXPECT_EQ(10U, B.getOffset("foob"));
}
TEST(StringTableBuilderTest, MachO64LinkedInOrder) {
StringTableBuilder B(StringTableBuilder::MachO64Linked);
B.add("foo");
B.add("ba");
B.add("f");
B.finalizeInOrder();
std::string Expected;
Expected += ' ';
Expected += '\x00';
Expected += "foo";
Expected += '\x00';
Expected += "ba";
Expected += '\x00';
Expected += "f";
Expected += '\x00';
// 64 bit Mach-O pads to 8 bytes
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
Expected += '\x00';
SmallString<64> Data;
raw_svector_ostream OS(Data);
B.write(OS);
EXPECT_EQ(Expected, Data);
EXPECT_EQ(2U, B.getOffset("foo"));
EXPECT_EQ(6U, B.getOffset("ba"));
EXPECT_EQ(9U, B.getOffset("f"));
}
}