1 /* -*- mode: C++; c-basic-offset: 4; tab-width: 4 -*-
3 * Copyright (c) 2006-2010 Apple Inc. All rights reserved.
5 * @APPLE_LICENSE_HEADER_START@
7 * This file contains Original Code and/or Modifications of Original Code
8 * as defined in and that are subject to the Apple Public Source License
9 * Version 2.0 (the 'License'). You may not use this file except in
10 * compliance with the License. Please obtain a copy of the License at
11 * http://www.opensource.apple.com/apsl/ and read it before using this
14 * The Original Code and all software distributed under the License are
15 * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
16 * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
17 * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
19 * Please see the License for the specific language governing rights and
20 * limitations under the License.
22 * @APPLE_LICENSE_HEADER_END@
25 #ifndef __LTO_READER_H__
26 #define __LTO_READER_H__
29 #include <sys/param.h>
30 #include <sys/fcntl.h>
34 #include <mach-o/dyld.h>
37 #include <unordered_set>
38 #include <unordered_map>
40 #include "MachOFileAbstraction.hpp"
41 #include "Architectures.hpp"
43 #include "macho_relocatable_file.h"
46 // #defines are a work around for <rdar://problem/8760268>
47 #define __STDC_LIMIT_MACROS 1
48 #define __STDC_CONSTANT_MACROS 1
49 #include "llvm-c/lto.h"
55 // ld64 only tracks non-internal symbols from an llvm bitcode file.
56 // We model this by having an InternalAtom which represent all internal functions and data.
57 // All non-interal symbols from a bitcode file are represented by an Atom
58 // and each Atom has a reference to the InternalAtom. The InternalAtom
59 // also has references to each symbol external to the bitcode file.
61 class InternalAtom
: public ld::Atom
64 InternalAtom(class File
& f
);
65 // overrides of ld::Atom
66 virtual ld::File
* file() const { return &_file
; }
67 virtual const char* name() const { return "import-atom"; }
68 virtual uint64_t size() const { return 0; }
69 virtual uint64_t objectAddress() const { return 0; }
70 virtual void copyRawContent(uint8_t buffer
[]) const { }
71 virtual void setScope(Scope
) { }
72 virtual ld::Fixup::iterator
fixupsBegin() const { return &_undefs
[0]; }
73 virtual ld::Fixup::iterator
fixupsEnd() const { return &_undefs
[_undefs
.size()]; }
75 // for adding references to symbols outside bitcode file
76 void addReference(const char* nm
)
77 { _undefs
.push_back(ld::Fixup(0, ld::Fixup::k1of1
,
78 ld::Fixup::kindNone
, false, strdup(nm
))); }
82 mutable std::vector
<ld::Fixup
> _undefs
;
89 class File
: public ld::relocatable::File
92 File(const char* path
, time_t mTime
, ld::File::Ordinal ordinal
,
93 const uint8_t* content
, uint32_t contentLength
, cpu_type_t arch
);
96 // overrides of ld::File
97 virtual bool forEachAtom(ld::File::AtomHandler
&) const;
98 virtual bool justInTimeforEachAtom(const char* name
, ld::File::AtomHandler
&) const
100 virtual uint32_t cpuSubType() const { return _cpuSubType
; }
102 // overrides of ld::relocatable::File
103 virtual DebugInfoKind
debugInfo() const { return _debugInfo
; }
104 virtual const char* debugInfoPath() const { return _debugInfoPath
; }
105 virtual time_t debugInfoModificationTime() const
106 { return _debugInfoModTime
; }
107 virtual const std::vector
<ld::relocatable::File::Stab
>* stabs() const { return NULL
; }
108 virtual bool canScatterAtoms() const { return true; }
109 virtual LinkerOptionsList
* linkerOptions() const { return NULL
; }
113 lto_module_t
module() { return _module
; }
114 class InternalAtom
& internalAtom() { return _internalAtom
; }
115 void setDebugInfo(ld::relocatable::File::DebugInfoKind k
,
116 const char* pth
, time_t modTime
, uint32_t subtype
)
118 _debugInfoPath
= pth
;
119 _debugInfoModTime
= modTime
;
120 _cpuSubType
= subtype
;}
122 static bool sSupportsLocalContext
;
123 static bool sHasTriedLocalContext
;
124 bool mergeIntoGenerator(lto_code_gen_t generator
, bool useSetModule
);
127 friend class InternalAtom
;
130 cpu_type_t _architecture
;
131 class InternalAtom _internalAtom
;
132 class Atom
* _atomArray
;
133 uint32_t _atomArrayCount
;
134 lto_module_t _module
;
136 const uint8_t* _content
;
137 uint32_t _contentLength
;
138 const char* _debugInfoPath
;
139 time_t _debugInfoModTime
;
140 ld::Section _section
;
141 ld::Fixup _fixupToInternal
;
142 ld::relocatable::File::DebugInfoKind _debugInfo
;
143 uint32_t _cpuSubType
;
147 // Atom acts as a proxy Atom for the symbols that are exported by LLVM bitcode file. Initially,
148 // Reader creates Atoms to allow linker proceed with usual symbol resolution phase. After
149 // optimization is performed, real Atoms are created for these symobls. However these real Atoms
150 // are not inserted into global symbol table. Atom holds real Atom and forwards appropriate
151 // methods to real atom.
153 class Atom
: public ld::Atom
156 Atom(File
& f
, const char* name
, ld::Atom::Scope s
,
157 ld::Atom::Definition d
, ld::Atom::Combine c
, ld::Atom::Alignment a
, bool ah
);
159 // overrides of ld::Atom
160 virtual ld::File
* file() const { return &_file
; }
161 virtual const char* translationUnitSource() const
162 { return (_compiledAtom
? _compiledAtom
->translationUnitSource() : NULL
); }
163 virtual const char* name() const { return _name
; }
164 virtual uint64_t size() const { return (_compiledAtom
? _compiledAtom
->size() : 0); }
165 virtual uint64_t objectAddress() const { return (_compiledAtom
? _compiledAtom
->objectAddress() : 0); }
166 virtual void copyRawContent(uint8_t buffer
[]) const
167 { if (_compiledAtom
) _compiledAtom
->copyRawContent(buffer
); }
168 virtual const uint8_t* rawContentPointer() const
169 { return (_compiledAtom
? _compiledAtom
->rawContentPointer() : NULL
); }
170 virtual unsigned long contentHash(const class ld::IndirectBindingTable
& ibt
) const
171 { return (_compiledAtom
? _compiledAtom
->contentHash(ibt
) : 0); }
172 virtual bool canCoalesceWith(const ld::Atom
& rhs
, const class ld::IndirectBindingTable
& ibt
) const
173 { return (_compiledAtom
? _compiledAtom
->canCoalesceWith(rhs
,ibt
) : false); }
174 virtual ld::Fixup::iterator
fixupsBegin() const
175 { return (_compiledAtom
? _compiledAtom
->fixupsBegin() : (ld::Fixup
*)&_file
._fixupToInternal
); }
176 virtual ld::Fixup::iterator
fixupsEnd() const
177 { return (_compiledAtom
? _compiledAtom
->fixupsEnd() : &((ld::Fixup
*)&_file
._fixupToInternal
)[1]); }
178 virtual ld::Atom::UnwindInfo::iterator
beginUnwind() const
179 { return (_compiledAtom
? _compiledAtom
->beginUnwind() : NULL
); }
180 virtual ld::Atom::UnwindInfo::iterator
endUnwind() const
181 { return (_compiledAtom
? _compiledAtom
->endUnwind() : NULL
); }
182 virtual ld::Atom::LineInfo::iterator
beginLineInfo() const
183 { return (_compiledAtom
? _compiledAtom
->beginLineInfo() : NULL
); }
184 virtual ld::Atom::LineInfo::iterator
endLineInfo() const
185 { return (_compiledAtom
? _compiledAtom
->endLineInfo() : NULL
); }
187 const ld::Atom
* compiledAtom() { return _compiledAtom
; }
188 void setCompiledAtom(const ld::Atom
& atom
);
194 const ld::Atom
* _compiledAtom
;
206 static bool validFile(const uint8_t* fileContent
, uint64_t fileLength
, cpu_type_t architecture
, cpu_subtype_t subarch
);
207 static const char* fileKind(const uint8_t* fileContent
, uint64_t fileLength
);
208 static File
* parse(const uint8_t* fileContent
, uint64_t fileLength
, const char* path
,
209 time_t modTime
, ld::File::Ordinal ordinal
, cpu_type_t architecture
, cpu_subtype_t subarch
,
210 bool logAllFiles
, bool verboseOptimizationHints
);
211 static bool libLTOisLoaded() { return (::lto_get_version() != NULL
); }
212 static bool optimize( const std::vector
<const ld::Atom
*>& allAtoms
,
214 const OptimizeOptions
& options
,
215 ld::File::AtomHandler
& handler
,
216 std::vector
<const ld::Atom
*>& newAtoms
,
217 std::vector
<const char*>& additionalUndefines
);
219 static const char* ltoVersion() { return ::lto_get_version(); }
222 static const char* tripletPrefixForArch(cpu_type_t arch
);
223 static ld::relocatable::File
* parseMachOFile(const uint8_t* p
, size_t len
, const OptimizeOptions
& options
);
224 #if LTO_API_VERSION >= 7
225 static void ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t
, const char*, void*);
228 typedef std::unordered_set
<const char*, ld::CStringHash
, ld::CStringEquals
> CStringSet
;
229 typedef std::unordered_map
<const char*, Atom
*, ld::CStringHash
, ld::CStringEquals
> CStringToAtom
;
231 class AtomSyncer
: public ld::File::AtomHandler
{
233 AtomSyncer(std::vector
<const char*>& a
, std::vector
<const ld::Atom
*>&na
,
234 CStringToAtom la
, CStringToAtom dla
, const OptimizeOptions
& options
) :
235 _options(options
), _additionalUndefines(a
), _newAtoms(na
), _llvmAtoms(la
), _deadllvmAtoms(dla
) { }
236 virtual void doAtom(const class ld::Atom
&);
237 virtual void doFile(const class ld::File
&) { }
240 const OptimizeOptions
& _options
;
241 std::vector
<const char*>& _additionalUndefines
;
242 std::vector
<const ld::Atom
*>& _newAtoms
;
243 CStringToAtom _llvmAtoms
;
244 CStringToAtom _deadllvmAtoms
;
247 static std::vector
<File
*> _s_files
;
250 std::vector
<File
*> Parser::_s_files
;
253 bool Parser::validFile(const uint8_t* fileContent
, uint64_t fileLength
, cpu_type_t architecture
, cpu_subtype_t subarch
)
255 for (const ArchInfo
* t
=archInfoArray
; t
->archName
!= NULL
; ++t
) {
256 if ( (architecture
== t
->cpuType
) && (!(t
->isSubType
) || (subarch
== t
->cpuSubType
)) ) {
257 bool result
= ::lto_module_is_object_file_in_memory_for_target(fileContent
, fileLength
, t
->llvmTriplePrefix
);
259 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
260 if ( t
->llvmTriplePrefixAlt
[0] != '\0' ) {
261 result
= ::lto_module_is_object_file_in_memory_for_target(fileContent
, fileLength
, t
->llvmTriplePrefixAlt
);
270 const char* Parser::fileKind(const uint8_t* p
, uint64_t fileLength
)
272 if ( (p
[0] == 0xDE) && (p
[1] == 0xC0) && (p
[2] == 0x17) && (p
[3] == 0x0B) ) {
273 cpu_type_t arch
= LittleEndian::get32(*((uint32_t*)(&p
[16])));
274 for (const ArchInfo
* t
=archInfoArray
; t
->archName
!= NULL
; ++t
) {
275 if ( arch
== t
->cpuType
) {
276 if ( t
->isSubType
) {
277 if ( ::lto_module_is_object_file_in_memory_for_target(p
, fileLength
, t
->llvmTriplePrefix
) )
285 return "unknown bitcode architecture";
290 File
* Parser::parse(const uint8_t* fileContent
, uint64_t fileLength
, const char* path
, time_t modTime
, ld::File::Ordinal ordinal
,
291 cpu_type_t architecture
, cpu_subtype_t subarch
, bool logAllFiles
, bool verboseOptimizationHints
)
293 File
* f
= new File(path
, modTime
, ordinal
, fileContent
, fileLength
, architecture
);
294 _s_files
.push_back(f
);
296 printf("%s\n", path
);
301 ld::relocatable::File
* Parser::parseMachOFile(const uint8_t* p
, size_t len
, const OptimizeOptions
& options
)
303 mach_o::relocatable::ParserOptions objOpts
;
304 objOpts
.architecture
= options
.arch
;
305 objOpts
.objSubtypeMustMatch
= false;
306 objOpts
.logAllFiles
= false;
307 objOpts
.warnUnwindConversionProblems
= options
.needsUnwindInfoSection
;
308 objOpts
.keepDwarfUnwind
= options
.keepDwarfUnwind
;
309 objOpts
.forceDwarfConversion
= false;
310 objOpts
.neverConvertDwarf
= false;
311 objOpts
.verboseOptimizationHints
= options
.verboseOptimizationHints
;
312 objOpts
.armUsesZeroCostExceptions
= options
.armUsesZeroCostExceptions
;
313 objOpts
.simulator
= options
.simulator
;
314 objOpts
.ignoreMismatchPlatform
= options
.ignoreMismatchPlatform
;
315 objOpts
.platform
= options
.platform
;
317 objOpts
.srcKind
= ld::relocatable::File::kSourceLTO
;
318 objOpts
.treateBitcodeAsData
= false;
319 objOpts
.usingBitcode
= options
.bitcodeBundle
;
321 // mach-o parsing is done in-memory, but need path for debug notes
322 const char* path
= "/tmp/lto.o";
324 if ( options
.tmpObjectFilePath
!= NULL
) {
325 path
= options
.tmpObjectFilePath
;
326 struct stat statBuffer
;
327 if ( stat(options
.tmpObjectFilePath
, &statBuffer
) == 0 )
328 modTime
= statBuffer
.st_mtime
;
331 ld::relocatable::File
* result
= mach_o::relocatable::parse(p
, len
, path
, modTime
, ld::File::Ordinal::LTOOrdinal(), objOpts
);
332 if ( result
!= NULL
)
334 throw "LLVM LTO, file is not of required architecture";
339 File::File(const char* pth
, time_t mTime
, ld::File::Ordinal ordinal
, const uint8_t* content
, uint32_t contentLength
, cpu_type_t arch
)
340 : ld::relocatable::File(pth
,mTime
,ordinal
), _architecture(arch
), _internalAtom(*this),
341 _atomArray(NULL
), _atomArrayCount(0), _module(NULL
), _path(pth
),
342 _content(content
), _contentLength(contentLength
), _debugInfoPath(pth
),
343 _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO
),
344 _fixupToInternal(0, ld::Fixup::k1of1
, ld::Fixup::kindNone
, &_internalAtom
),
345 _debugInfo(ld::relocatable::File::kDebugInfoNone
), _cpuSubType(0)
347 const bool log
= false;
349 // create llvm module
350 #if LTO_API_VERSION >= 11
351 if ( sSupportsLocalContext
|| !sHasTriedLocalContext
) {
352 _module
= ::lto_module_create_in_local_context(content
, contentLength
, pth
);
354 if ( !sHasTriedLocalContext
) {
355 sHasTriedLocalContext
= true;
356 sSupportsLocalContext
= (_module
!= NULL
);
358 if ( (_module
== NULL
) && !sSupportsLocalContext
)
360 #if LTO_API_VERSION >= 9
361 _module
= ::lto_module_create_from_memory_with_path(content
, contentLength
, pth
);
362 if ( _module
== NULL
&& !sSupportsLocalContext
)
364 _module
= ::lto_module_create_from_memory(content
, contentLength
);
365 if ( _module
== NULL
)
366 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth
, ::lto_get_error_message(), ::lto_get_version());
368 if ( log
) fprintf(stderr
, "bitcode file: %s\n", pth
);
370 // create atom for each global symbol in module
371 uint32_t count
= ::lto_module_get_num_symbols(_module
);
372 _atomArray
= (Atom
*)malloc(sizeof(Atom
)*count
);
373 for (uint32_t i
=0; i
< count
; ++i
) {
374 const char* name
= ::lto_module_get_symbol_name(_module
, i
);
375 lto_symbol_attributes attr
= lto_module_get_symbol_attribute(_module
, i
);
377 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
378 // ignore dtrace static probes for now
379 // later when codegen is done and a mach-o file is produces the probes will be processed
380 if ( (strncmp(name
, "___dtrace_probe$", 16) == 0) || (strncmp(name
, "___dtrace_isenabled$", 20) == 0) )
383 ld::Atom::Definition def
;
384 ld::Atom::Combine combine
= ld::Atom::combineNever
;
385 switch ( attr
& LTO_SYMBOL_DEFINITION_MASK
) {
386 case LTO_SYMBOL_DEFINITION_REGULAR
:
387 def
= ld::Atom::definitionRegular
;
389 case LTO_SYMBOL_DEFINITION_TENTATIVE
:
390 def
= ld::Atom::definitionTentative
;
392 case LTO_SYMBOL_DEFINITION_WEAK
:
393 def
= ld::Atom::definitionRegular
;
394 combine
= ld::Atom::combineByName
;
396 case LTO_SYMBOL_DEFINITION_UNDEFINED
:
397 case LTO_SYMBOL_DEFINITION_WEAKUNDEF
:
398 def
= ld::Atom::definitionProxy
;
401 throwf("unknown definition kind for symbol %s in bitcode file %s", name
, pth
);
404 // make LLVM atoms for definitions and a reference for undefines
405 if ( def
!= ld::Atom::definitionProxy
) {
406 ld::Atom::Scope scope
;
407 bool autohide
= false;
408 switch ( attr
& LTO_SYMBOL_SCOPE_MASK
) {
409 case LTO_SYMBOL_SCOPE_INTERNAL
:
410 scope
= ld::Atom::scopeTranslationUnit
;
412 case LTO_SYMBOL_SCOPE_HIDDEN
:
413 scope
= ld::Atom::scopeLinkageUnit
;
415 case LTO_SYMBOL_SCOPE_DEFAULT
:
416 scope
= ld::Atom::scopeGlobal
;
418 #if LTO_API_VERSION >= 4
419 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN
:
420 scope
= ld::Atom::scopeGlobal
;
425 throwf("unknown scope for symbol %s in bitcode file %s", name
, pth
);
427 // only make atoms for non-internal symbols
428 if ( scope
== ld::Atom::scopeTranslationUnit
)
430 uint8_t alignment
= (attr
& LTO_SYMBOL_ALIGNMENT_MASK
);
431 // make Atom using placement new operator
432 new (&_atomArray
[_atomArrayCount
++]) Atom(*this, name
, scope
, def
, combine
, alignment
, autohide
);
433 if ( scope
!= ld::Atom::scopeTranslationUnit
)
434 _internalAtom
.addReference(name
);
435 if ( log
) fprintf(stderr
, "\t0x%08X %s\n", attr
, name
);
438 // add to list of external references
439 _internalAtom
.addReference(name
);
440 if ( log
) fprintf(stderr
, "\t%s (undefined)\n", name
);
444 #if LTO_API_VERSION >= 11
445 if ( sSupportsLocalContext
)
455 bool File::mergeIntoGenerator(lto_code_gen_t generator
, bool useSetModule
) {
456 #if LTO_API_VERSION >= 11
457 if ( sSupportsLocalContext
) {
458 assert(!_module
&& "Expected module to be disposed");
459 _module
= ::lto_module_create_in_codegen_context(_content
, _contentLength
,
461 if ( _module
== NULL
)
462 throwf("could not reparse object file %s: '%s', using libLTO version '%s'",
463 _path
, ::lto_get_error_message(), ::lto_get_version());
466 assert(_module
&& "Expected module to stick around");
467 #if LTO_API_VERSION >= 13
469 // lto_codegen_set_module will transfer ownership of the module to LTO code generator,
470 // so we don't need to release the module here.
471 ::lto_codegen_set_module(generator
, _module
);
475 if ( ::lto_codegen_add_module(generator
, _module
) )
478 // <rdar://problem/15471128> linker should release module as soon as possible
485 if ( _module
!= NULL
)
486 ::lto_module_dispose(_module
);
490 bool File::forEachAtom(ld::File::AtomHandler
& handler
) const
492 handler
.doAtom(_internalAtom
);
493 for(uint32_t i
=0; i
< _atomArrayCount
; ++i
) {
494 handler
.doAtom(_atomArray
[i
]);
499 InternalAtom::InternalAtom(File
& f
)
500 : ld::Atom(f
._section
, ld::Atom::definitionRegular
, ld::Atom::combineNever
, ld::Atom::scopeTranslationUnit
,
501 ld::Atom::typeLTOtemporary
, ld::Atom::symbolTableNotIn
, true, false, false, ld::Atom::Alignment(0)),
506 Atom::Atom(File
& f
, const char* nm
, ld::Atom::Scope s
, ld::Atom::Definition d
, ld::Atom::Combine c
,
507 ld::Atom::Alignment a
, bool ah
)
508 : ld::Atom(f
._section
, d
, c
, s
, ld::Atom::typeLTOtemporary
,
509 ld::Atom::symbolTableIn
, false, false, false, a
),
510 _file(f
), _name(strdup(nm
)), _compiledAtom(NULL
)
516 void Atom::setCompiledAtom(const ld::Atom
& atom
)
518 // set delegate so virtual methods go to it
519 _compiledAtom
= &atom
;
521 //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
523 // update fields in ld::Atom to match newly constructed mach-o atom
524 (const_cast<Atom
*>(this))->setAttributesFromAtom(atom
);
529 // <rdar://problem/12379604> The order that files are merged must match command line order
530 struct CommandLineOrderFileSorter
532 bool operator()(File
* left
, File
* right
)
534 return ( left
->ordinal() < right
->ordinal() );
539 #if LTO_API_VERSION >= 7
540 void Parser::ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t severity
, const char* message
, void*)
542 switch ( severity
) {
543 #if LTO_API_VERSION >= 10
545 fprintf(stderr
, "ld: LTO remark: %s\n", message
);
550 warning("%s", message
);
553 throwf("%s", message
);
558 bool Parser::optimize( const std::vector
<const ld::Atom
*>& allAtoms
,
560 const OptimizeOptions
& options
,
561 ld::File::AtomHandler
& handler
,
562 std::vector
<const ld::Atom
*>& newAtoms
,
563 std::vector
<const char*>& additionalUndefines
)
565 const bool logMustPreserve
= false;
566 const bool logExtraOptions
= false;
567 const bool logBitcodeFiles
= false;
568 const bool logAtomsBeforeSync
= false;
570 // exit quickly if nothing to do
571 if ( _s_files
.size() == 0 )
574 // print out LTO version string if -v was used
575 if ( options
.verbose
)
576 fprintf(stderr
, "%s\n", ::lto_get_version());
578 // create optimizer and add each Reader
579 lto_code_gen_t generator
= NULL
;
580 #if LTO_API_VERSION >= 11
581 if ( File::sSupportsLocalContext
)
582 generator
= ::lto_codegen_create_in_local_context();
585 generator
= ::lto_codegen_create();
586 #if LTO_API_VERSION >= 7
587 lto_codegen_set_diagnostic_handler(generator
, ltoDiagnosticHandler
, NULL
);
590 // <rdar://problem/12379604> The order that files are merged must match command line order
591 std::sort(_s_files
.begin(), _s_files
.end(), CommandLineOrderFileSorter());
592 ld::File::Ordinal lastOrdinal
;
594 // When flto_codegen_only is on and we have a single .bc file, use lto_codegen_set_module instead of
595 // lto_codegen_add_module, to make sure the the destination module will be the same as the input .bc file.
596 bool useSetModule
= false;
597 #if LTO_API_VERSION >= 13
598 useSetModule
= (_s_files
.size() == 1) && options
.ltoCodegenOnly
&& (::lto_api_version() >= 13);
600 for (std::vector
<File
*>::iterator it
=_s_files
.begin(); it
!= _s_files
.end(); ++it
) {
602 assert(f
->ordinal() > lastOrdinal
);
603 if ( logBitcodeFiles
&& !useSetModule
) fprintf(stderr
, "lto_codegen_add_module(%s)\n", f
->path());
604 if ( logBitcodeFiles
&& useSetModule
) fprintf(stderr
, "lto_codegen_set_module(%s)\n", f
->path());
605 if ( f
->mergeIntoGenerator(generator
, useSetModule
) )
606 throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", f
->path(), ::lto_get_error_message(), ::lto_get_version());
607 lastOrdinal
= f
->ordinal();
610 // add any -mllvm command line options
611 for (std::vector
<const char*>::const_iterator it
=options
.llvmOptions
->begin(); it
!= options
.llvmOptions
->end(); ++it
) {
612 if ( logExtraOptions
) fprintf(stderr
, "passing option to llvm: %s\n", *it
);
613 ::lto_codegen_debug_options(generator
, *it
);
616 // <rdar://problem/13687397> Need a way for LTO to get cpu variants (until that info is in bitcode)
617 if ( options
.mcpu
!= NULL
)
618 ::lto_codegen_set_cpu(generator
, options
.mcpu
);
620 // The atom graph uses directed edges (references). Collect all references where
621 // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
622 // external (i.e. not originated from same .o file) reference if all originating atoms are also
623 // defined in llvm bitcode file.
624 CStringSet nonLLVMRefs
;
625 CStringToAtom llvmAtoms
;
626 bool hasNonllvmAtoms
= false;
627 for (std::vector
<const ld::Atom
*>::const_iterator it
= allAtoms
.begin(); it
!= allAtoms
.end(); ++it
) {
628 const ld::Atom
* atom
= *it
;
629 // only look at references that come from an atom that is not an llvm atom
630 if ( atom
->contentType() != ld::Atom::typeLTOtemporary
) {
631 if ( (atom
->section().type() != ld::Section::typeMachHeader
) && (atom
->definition() != ld::Atom::definitionProxy
) ) {
632 hasNonllvmAtoms
= true;
634 const ld::Atom
* target
;
635 for (ld::Fixup::iterator fit
=atom
->fixupsBegin(); fit
!= atom
->fixupsEnd(); ++fit
) {
636 switch ( fit
->binding
) {
637 case ld::Fixup::bindingDirectlyBound
:
638 // that reference an llvm atom
639 if ( fit
->u
.target
->contentType() == ld::Atom::typeLTOtemporary
)
640 nonLLVMRefs
.insert(fit
->u
.target
->name());
642 case ld::Fixup::bindingsIndirectlyBound
:
643 target
= state
.indirectBindingTable
[fit
->u
.bindingIndex
];
644 if ( (target
!= NULL
) && (target
->contentType() == ld::Atom::typeLTOtemporary
) )
645 nonLLVMRefs
.insert(target
->name());
651 else if ( atom
->scope() >= ld::Atom::scopeLinkageUnit
) {
652 llvmAtoms
[atom
->name()] = (Atom
*)atom
;
655 // if entry point is in a llvm bitcode file, it must be preserved by LTO
656 if ( state
.entryPoint
!= NULL
) {
657 if ( state
.entryPoint
->contentType() == ld::Atom::typeLTOtemporary
)
658 nonLLVMRefs
.insert(state
.entryPoint
->name());
661 // deadAtoms are the atoms that the linker coalesced. For instance weak or tentative definitions
662 // overriden by another atom. If any of these deadAtoms are llvm atoms and they were replaced
663 // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
664 // atom so that the linker can replace it with the mach-o one later.
665 CStringToAtom deadllvmAtoms
;
666 for (std::vector
<const ld::Atom
*>::const_iterator it
= allAtoms
.begin(); it
!= allAtoms
.end(); ++it
) {
667 const ld::Atom
* atom
= *it
;
668 if ( atom
->coalescedAway() && (atom
->contentType() == ld::Atom::typeLTOtemporary
) ) {
669 const char* name
= atom
->name();
670 if ( logMustPreserve
) fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name
);
671 ::lto_codegen_add_must_preserve_symbol(generator
, name
);
672 deadllvmAtoms
[name
] = (Atom
*)atom
;
675 for (std::vector
<File
*>::iterator it
=_s_files
.begin(); it
!= _s_files
.end(); ++it
) {
677 for(uint32_t i
=0; i
< file
->_atomArrayCount
; ++i
) {
678 Atom
* llvmAtom
= &file
->_atomArray
[i
];
679 if ( llvmAtom
->coalescedAway() ) {
680 const char* name
= llvmAtom
->name();
681 if ( deadllvmAtoms
.find(name
) == deadllvmAtoms
.end() ) {
682 if ( logMustPreserve
)
683 fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name
);
684 ::lto_codegen_add_must_preserve_symbol(generator
, name
);
685 deadllvmAtoms
[name
] = (Atom
*)llvmAtom
;
688 else if ( options
.linkerDeadStripping
&& !llvmAtom
->live() ) {
689 const char* name
= llvmAtom
->name();
690 deadllvmAtoms
[name
] = (Atom
*)llvmAtom
;
695 // tell code generator about symbols that must be preserved
696 for (CStringToAtom::iterator it
= llvmAtoms
.begin(); it
!= llvmAtoms
.end(); ++it
) {
697 const char* name
= it
->first
;
698 Atom
* atom
= it
->second
;
699 // Include llvm Symbol in export list if it meets one of following two conditions
700 // 1 - atom scope is global (and not linkage unit).
701 // 2 - included in nonLLVMRefs set.
702 // If a symbol is not listed in exportList then LTO is free to optimize it away.
703 if ( (atom
->scope() == ld::Atom::scopeGlobal
) && options
.preserveAllGlobals
) {
704 if ( logMustPreserve
) fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name
);
705 ::lto_codegen_add_must_preserve_symbol(generator
, name
);
707 else if ( nonLLVMRefs
.find(name
) != nonLLVMRefs
.end() ) {
708 if ( logMustPreserve
) fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name
);
709 ::lto_codegen_add_must_preserve_symbol(generator
, name
);
711 else if ( options
.relocatable
&& hasNonllvmAtoms
) {
712 // <rdar://problem/14334895> ld -r mode but merging in some mach-o files, so need to keep libLTO from optimizing away anything
713 if ( logMustPreserve
) fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because -r mode disable LTO dead stripping\n", name
);
714 ::lto_codegen_add_must_preserve_symbol(generator
, name
);
718 // <rdar://problem/16165191> tell code generator to preserve initial undefines
719 for( std::vector
<const char*>::const_iterator it
=options
.initialUndefines
->begin(); it
!= options
.initialUndefines
->end(); ++it
) {
720 if ( logMustPreserve
) fprintf(stderr
, "lto_codegen_add_must_preserve_symbol(%s) because it is an initial undefine\n", *it
);
721 ::lto_codegen_add_must_preserve_symbol(generator
, *it
);
724 // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
725 if ( options
.relocatable
&& !hasNonllvmAtoms
) {
726 #if LTO_API_VERSION >= 15
727 ::lto_codegen_set_should_embed_uselists(generator
, false);
729 if ( ! ::lto_codegen_write_merged_modules(generator
, options
.outputFilePath
) ) {
730 // HACK, no good way to tell linker we are all done, so just quit
733 warning("could not produce merged bitcode file");
736 // set code-gen model
737 lto_codegen_model model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC
;
738 if ( options
.mainExecutable
) {
739 if ( options
.staticExecutable
) {
740 // x86_64 "static" or any "-static -pie" is really dynamic code model
741 if ( (options
.arch
== CPU_TYPE_X86_64
) || options
.pie
)
742 model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC
;
744 model
= LTO_CODEGEN_PIC_MODEL_STATIC
;
748 model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC
;
750 model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC
;
754 if ( options
.allowTextRelocs
)
755 model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC
;
757 model
= LTO_CODEGEN_PIC_MODEL_DYNAMIC
;
759 if ( ::lto_codegen_set_pic_model(generator
, model
) )
760 throwf("could not create set codegen model: %s", lto_get_error_message());
762 // if requested, save off merged bitcode file
763 if ( options
.saveTemps
) {
764 char tempBitcodePath
[MAXPATHLEN
];
765 strcpy(tempBitcodePath
, options
.outputFilePath
);
766 strcat(tempBitcodePath
, ".lto.bc");
767 #if LTO_API_VERSION >= 15
768 ::lto_codegen_set_should_embed_uselists(generator
, true);
770 ::lto_codegen_write_merged_modules(generator
, tempBitcodePath
);
773 #if LTO_API_VERSION >= 3
774 // find assembler next to linker
776 uint32_t bufSize
= PATH_MAX
;
777 if ( _NSGetExecutablePath(path
, &bufSize
) != -1 ) {
778 char* lastSlash
= strrchr(path
, '/');
779 if ( lastSlash
!= NULL
) {
780 strcpy(lastSlash
+1, "as");
781 struct stat statInfo
;
782 if ( stat(path
, &statInfo
) == 0 )
783 ::lto_codegen_set_assembler_path(generator
, path
);
788 // When lto API version is greater than or equal to 12, we use lto_codegen_optimize and lto_codegen_compile_optimized
789 // instead of lto_codegen_compile, and we save the merged bitcode file in between.
790 bool useSplitAPI
= false;
791 #if LTO_API_VERSION >= 12
792 if ( ::lto_api_version() >= 12)
796 size_t machOFileLen
= 0;
797 const uint8_t* machOFile
= NULL
;
799 #if LTO_API_VERSION >= 12
800 #if LTO_API_VERSION >= 14
801 if ( ::lto_api_version() >= 14 && options
.ltoCodegenOnly
)
802 lto_codegen_set_should_internalize(generator
, false);
805 if ( !options
.ltoCodegenOnly
&& ::lto_codegen_optimize(generator
) )
806 throwf("could not do LTO optimization: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
808 if ( options
.saveTemps
|| options
.bitcodeBundle
) {
809 // save off merged bitcode file
810 char tempOptBitcodePath
[MAXPATHLEN
];
811 strcpy(tempOptBitcodePath
, options
.outputFilePath
);
812 strcat(tempOptBitcodePath
, ".lto.opt.bc");
813 #if LTO_API_VERSION >= 15
814 ::lto_codegen_set_should_embed_uselists(generator
, true);
816 ::lto_codegen_write_merged_modules(generator
, tempOptBitcodePath
);
817 if ( options
.bitcodeBundle
)
818 state
.ltoBitcodePath
= tempOptBitcodePath
;
821 // run code generator
822 machOFile
= (uint8_t*)::lto_codegen_compile_optimized(generator
, &machOFileLen
);
824 if ( machOFile
== NULL
)
825 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
828 // run optimizer and code generator
829 machOFile
= (uint8_t*)::lto_codegen_compile(generator
, &machOFileLen
);
830 if ( machOFile
== NULL
)
831 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
832 if ( options
.saveTemps
) {
833 // save off merged bitcode file
834 char tempOptBitcodePath
[MAXPATHLEN
];
835 strcpy(tempOptBitcodePath
, options
.outputFilePath
);
836 strcat(tempOptBitcodePath
, ".lto.opt.bc");
837 #if LTO_API_VERSION >= 15
838 ::lto_codegen_set_should_embed_uselists(generator
, true);
840 ::lto_codegen_write_merged_modules(generator
, tempOptBitcodePath
);
844 // if requested, save off temp mach-o file
845 if ( options
.saveTemps
) {
846 char tempMachoPath
[MAXPATHLEN
];
847 strcpy(tempMachoPath
, options
.outputFilePath
);
848 strcat(tempMachoPath
, ".lto.o");
849 int fd
= ::open(tempMachoPath
, O_CREAT
| O_WRONLY
| O_TRUNC
, 0666);
851 ::write(fd
, machOFile
, machOFileLen
);
856 // if needed, save temp mach-o file to specific location
857 if ( options
.tmpObjectFilePath
!= NULL
) {
858 int fd
= ::open(options
.tmpObjectFilePath
, O_CREAT
| O_WRONLY
| O_TRUNC
, 0666);
860 ::write(fd
, machOFile
, machOFileLen
);
864 warning("could not write LTO temp file '%s', errno=%d", options
.tmpObjectFilePath
, errno
);
868 // parse generated mach-o file into a MachOReader
869 ld::relocatable::File
* machoFile
= parseMachOFile(machOFile
, machOFileLen
, options
);
871 // sync generated mach-o atoms with existing atoms ld knows about
872 if ( logAtomsBeforeSync
) {
873 fprintf(stderr
, "llvmAtoms:\n");
874 for (CStringToAtom::iterator it
= llvmAtoms
.begin(); it
!= llvmAtoms
.end(); ++it
) {
875 const char* name
= it
->first
;
876 Atom
* atom
= it
->second
;
877 fprintf(stderr
, "\t%p\t%s\n", atom
, name
);
879 fprintf(stderr
, "deadllvmAtoms:\n");
880 for (CStringToAtom::iterator it
= deadllvmAtoms
.begin(); it
!= deadllvmAtoms
.end(); ++it
) {
881 const char* name
= it
->first
;
882 Atom
* atom
= it
->second
;
883 fprintf(stderr
, "\t%p\t%s\n", atom
, name
);
886 AtomSyncer
syncer(additionalUndefines
, newAtoms
, llvmAtoms
, deadllvmAtoms
, options
);
887 machoFile
->forEachAtom(syncer
);
889 // Remove InternalAtoms from ld
890 for (std::vector
<File
*>::iterator it
=_s_files
.begin(); it
!= _s_files
.end(); ++it
) {
891 (*it
)->internalAtom().setCoalescedAway();
893 // Remove Atoms from ld if code generator optimized them away
894 for (CStringToAtom::iterator li
= llvmAtoms
.begin(), le
= llvmAtoms
.end(); li
!= le
; ++li
) {
895 // check if setRealAtom() called on this Atom
896 if ( li
->second
->compiledAtom() == NULL
) {
897 //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
898 li
->second
->setCoalescedAway();
902 // notify about file level attributes
903 handler
.doFile(*machoFile
);
905 // if final mach-o file has debug info, update original bitcode files to match
906 for (std::vector
<File
*>::iterator it
=_s_files
.begin(); it
!= _s_files
.end(); ++it
) {
907 (*it
)->setDebugInfo(machoFile
->debugInfo(), machoFile
->path(),
908 machoFile
->modificationTime(), machoFile
->cpuSubType());
915 void Parser::AtomSyncer::doAtom(const ld::Atom
& machoAtom
)
917 static const bool log
= false;
918 static const ld::Atom
* lastProxiedAtom
= NULL
;
919 static const ld::File
* lastProxiedFile
= NULL
;
920 // update proxy atoms to point to real atoms and find new atoms
921 const char* name
= machoAtom
.name();
922 CStringToAtom::iterator pos
= _llvmAtoms
.find(name
);
923 if ( pos
!= _llvmAtoms
.end() ) {
924 // turn Atom into a proxy for this mach-o atom
925 pos
->second
->setCompiledAtom(machoAtom
);
926 lastProxiedAtom
= &machoAtom
;
927 lastProxiedFile
= pos
->second
->file();
928 if (log
) fprintf(stderr
, "AtomSyncer, mach-o atom %p synced to lto atom %p (name=%s)\n", &machoAtom
, pos
->second
, machoAtom
.name());
931 // an atom of this name was not in the allAtoms list the linker gave us
932 if ( _deadllvmAtoms
.find(name
) != _deadllvmAtoms
.end() ) {
933 // this corresponding to an atom that the linker coalesced away or marked not-live
934 if ( _options
.linkerDeadStripping
) {
935 // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
936 Atom
* llvmAtom
= _deadllvmAtoms
[name
];
937 llvmAtom
->setCompiledAtom(machoAtom
);
938 _newAtoms
.push_back(&machoAtom
);
939 if (log
) fprintf(stderr
, "AtomSyncer, mach-o atom %p matches dead lto atom %p but adding back (name=%s)\n", &machoAtom
, llvmAtom
, machoAtom
.name());
942 // Don't pass it back as a new atom
943 if (log
) fprintf(stderr
, "AtomSyncer, mach-o atom %p matches dead lto atom %p (name=%s)\n", &machoAtom
, _deadllvmAtoms
[name
], machoAtom
.name());
948 // this is something new that lto conjured up, tell ld its new
949 _newAtoms
.push_back(&machoAtom
);
950 // <rdar://problem/15469363> if new static atom in same section as previous non-static atom, assign to same file as previous
951 if ( (lastProxiedAtom
!= NULL
) && (lastProxiedAtom
->section() == machoAtom
.section()) ) {
952 ld::Atom
* ma
= const_cast<ld::Atom
*>(&machoAtom
);
953 ma
->setFile(lastProxiedFile
);
955 if (log
) fprintf(stderr
, "AtomSyncer, mach-o atom %p is totally new (name=%s)\n", &machoAtom
, machoAtom
.name());
959 // adjust fixups to go through proxy atoms
960 if (log
) fprintf(stderr
, " adjusting fixups in atom: %s\n", machoAtom
.name());
961 for (ld::Fixup::iterator fit
=machoAtom
.fixupsBegin(); fit
!= machoAtom
.fixupsEnd(); ++fit
) {
962 switch ( fit
->binding
) {
963 case ld::Fixup::bindingNone
:
965 case ld::Fixup::bindingByNameUnbound
:
966 // don't know if this target has been seen by linker before or if it is new
967 // be conservative and tell linker it is new
968 _additionalUndefines
.push_back(fit
->u
.name
);
969 if (log
) fprintf(stderr
, " adding by-name symbol %s\n", fit
->u
.name
);
971 case ld::Fixup::bindingDirectlyBound
:
972 // If mach-o atom is referencing another mach-o atom then
973 // reference is not going through Atom proxy. Fix it here to ensure that all
974 // llvm symbol references always go through Atom proxy.
976 const char* targetName
= fit
->u
.target
->name();
977 CStringToAtom::iterator post
= _llvmAtoms
.find(targetName
);
978 if ( post
!= _llvmAtoms
.end() ) {
979 const ld::Atom
* t
= post
->second
;
980 if (log
) fprintf(stderr
, " updating direct reference to %p to be ref to %p: %s\n", fit
->u
.target
, t
, targetName
);
984 // <rdar://problem/12859831> Don't unbind follow-on reference into by-name reference
985 if ( (_deadllvmAtoms
.find(targetName
) != _deadllvmAtoms
.end()) && (fit
->kind
!= ld::Fixup::kindNoneFollowOn
) && (fit
->u
.target
->scope() != ld::Atom::scopeTranslationUnit
) ) {
986 // target was coalesed away and replace by mach-o atom from a non llvm .o file
987 fit
->binding
= ld::Fixup::bindingByNameUnbound
;
988 fit
->u
.name
= targetName
;
992 //fprintf(stderr, " direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
994 case ld::Fixup::bindingByContentBound
:
995 //fprintf(stderr, " direct by content to: %s\n", fit->u.target->name());
997 case ld::Fixup::bindingsIndirectlyBound
:
998 assert(0 && "indirect binding found in initial mach-o file?");
999 //fprintf(stderr, " indirect by content to: %u\n", fit->u.bindingIndex);
1007 static pthread_mutex_t lto_lock
;
1009 Mutex() { pthread_mutex_lock(<o_lock
); }
1010 ~Mutex() { pthread_mutex_unlock(<o_lock
); }
1012 pthread_mutex_t
Mutex::lto_lock
= PTHREAD_MUTEX_INITIALIZER
;
1013 bool File::sSupportsLocalContext
= false;
1014 bool File::sHasTriedLocalContext
= false;
1017 // Used by archive reader to see if member is an llvm bitcode file
1019 bool isObjectFile(const uint8_t* fileContent
, uint64_t fileLength
, cpu_type_t architecture
, cpu_subtype_t subarch
)
1022 return Parser::validFile(fileContent
, fileLength
, architecture
, subarch
);
1026 static ld::relocatable::File
*parseImpl(
1027 const uint8_t *fileContent
, uint64_t fileLength
, const char *path
,
1028 time_t modTime
, ld::File::Ordinal ordinal
, cpu_type_t architecture
,
1029 cpu_subtype_t subarch
, bool logAllFiles
,
1030 bool verboseOptimizationHints
)
1032 if ( Parser::validFile(fileContent
, fileLength
, architecture
, subarch
) )
1033 return Parser::parse(fileContent
, fileLength
, path
, modTime
, ordinal
, architecture
, subarch
, logAllFiles
, verboseOptimizationHints
);
1039 // main function used by linker to instantiate ld::Files
1041 ld::relocatable::File
* parse(const uint8_t* fileContent
, uint64_t fileLength
,
1042 const char* path
, time_t modTime
, ld::File::Ordinal ordinal
,
1043 cpu_type_t architecture
, cpu_subtype_t subarch
, bool logAllFiles
,
1044 bool verboseOptimizationHints
)
1046 // Note: Once lto_module_create_in_local_context() and friends are thread safe
1047 // this lock can be removed.
1049 return parseImpl(fileContent
, fileLength
, path
, modTime
, ordinal
,
1050 architecture
, subarch
, logAllFiles
,
1051 verboseOptimizationHints
);
1055 // used by "ld -v" to report version of libLTO.dylib being used
1057 const char* version()
1060 return ::lto_get_version();
1065 // used by ld for error reporting
1067 bool libLTOisLoaded()
1070 return (::lto_get_version() != NULL
);
1074 // used by ld for error reporting
1076 const char* archName(const uint8_t* fileContent
, uint64_t fileLength
)
1079 return Parser::fileKind(fileContent
, fileLength
);
1083 // used by ld for doing link time optimization
1085 bool optimize( const std::vector
<const ld::Atom
*>& allAtoms
,
1086 ld::Internal
& state
,
1087 const OptimizeOptions
& options
,
1088 ld::File::AtomHandler
& handler
,
1089 std::vector
<const ld::Atom
*>& newAtoms
,
1090 std::vector
<const char*>& additionalUndefines
)
1093 return Parser::optimize(allAtoms
, state
, options
, handler
, newAtoms
, additionalUndefines
);