253.3
[darwin-xtools.git] / ld64 / src / ld / parsers / lto_file.cpp
blob26d70be8492c5b14d1842fd05abcb491f588a35a
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
12 * file.
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__
28 #include <stdlib.h>
29 #include <sys/param.h>
30 #include <sys/fcntl.h>
31 #include <sys/stat.h>
32 #include <errno.h>
33 #include <pthread.h>
34 #include <mach-o/dyld.h>
35 #include <vector>
36 #include <map>
37 #include <unordered_set>
38 #include <unordered_map>
40 #include "MachOFileAbstraction.hpp"
41 #include "Architectures.hpp"
42 #include "ld.hpp"
43 #include "macho_relocatable_file.h"
44 #include "lto_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"
51 namespace lto {
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
63 public:
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))); }
79 private:
81 ld::File& _file;
82 mutable std::vector<ld::Fixup> _undefs;
87 // LLVM bitcode file
89 class File : public ld::relocatable::File
91 public:
92 File(const char* path, time_t mTime, ld::File::Ordinal ordinal,
93 const uint8_t* content, uint32_t contentLength, cpu_type_t arch);
94 virtual ~File();
96 // overrides of ld::File
97 virtual bool forEachAtom(ld::File::AtomHandler&) const;
98 virtual bool justInTimeforEachAtom(const char* name, ld::File::AtomHandler&) const
99 { return false; }
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; }
112 void release();
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)
117 { _debugInfo = k;
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);
125 private:
126 friend class Atom;
127 friend class InternalAtom;
128 friend class Parser;
130 cpu_type_t _architecture;
131 class InternalAtom _internalAtom;
132 class Atom* _atomArray;
133 uint32_t _atomArrayCount;
134 lto_module_t _module;
135 const char* _path;
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
155 public:
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);
190 private:
192 File& _file;
193 const char* _name;
194 const ld::Atom* _compiledAtom;
203 class Parser
205 public:
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,
213 ld::Internal& state,
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(); }
221 private:
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*);
226 #endif
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 {
232 public:
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);
258 if ( !result ) {
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);
264 return result;
267 return false;
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) )
278 return t->archName;
280 else {
281 return t->archName;
285 return "unknown bitcode architecture";
287 return NULL;
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);
295 if ( logAllFiles )
296 printf("%s\n", path);
297 return f;
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;
316 objOpts.subType = 0;
317 objOpts.srcKind = ld::relocatable::File::kSourceLTO;
319 // mach-o parsing is done in-memory, but need path for debug notes
320 const char* path = "/tmp/lto.o";
321 time_t modTime = 0;
322 if ( options.tmpObjectFilePath != NULL ) {
323 path = options.tmpObjectFilePath;
324 struct stat statBuffer;
325 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
326 modTime = statBuffer.st_mtime;
329 ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
330 if ( result != NULL )
331 return result;
332 throw "LLVM LTO, file is not of required architecture";
337 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
338 : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
339 _atomArray(NULL), _atomArrayCount(0), _module(NULL), _path(pth),
340 _content(content), _contentLength(contentLength), _debugInfoPath(pth),
341 _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
342 _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
343 _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
345 const bool log = false;
347 // create llvm module
348 #if LTO_API_VERSION >= 11
349 if ( sSupportsLocalContext || !sHasTriedLocalContext ) {
350 _module = ::lto_module_create_in_local_context(content, contentLength, pth);
352 if ( !sHasTriedLocalContext ) {
353 sHasTriedLocalContext = true;
354 sSupportsLocalContext = (_module != NULL);
356 if ( (_module == NULL) && !sSupportsLocalContext )
357 #endif
358 #if LTO_API_VERSION >= 9
359 _module = ::lto_module_create_from_memory_with_path(content, contentLength, pth);
360 if ( _module == NULL && !sSupportsLocalContext )
361 #endif
362 _module = ::lto_module_create_from_memory(content, contentLength);
363 if ( _module == NULL )
364 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth, ::lto_get_error_message(), ::lto_get_version());
366 if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
368 // create atom for each global symbol in module
369 uint32_t count = ::lto_module_get_num_symbols(_module);
370 _atomArray = (Atom*)malloc(sizeof(Atom)*count);
371 for (uint32_t i=0; i < count; ++i) {
372 const char* name = ::lto_module_get_symbol_name(_module, i);
373 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
375 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
376 // ignore dtrace static probes for now
377 // later when codegen is done and a mach-o file is produces the probes will be processed
378 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
379 continue;
381 ld::Atom::Definition def;
382 ld::Atom::Combine combine = ld::Atom::combineNever;
383 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
384 case LTO_SYMBOL_DEFINITION_REGULAR:
385 def = ld::Atom::definitionRegular;
386 break;
387 case LTO_SYMBOL_DEFINITION_TENTATIVE:
388 def = ld::Atom::definitionTentative;
389 break;
390 case LTO_SYMBOL_DEFINITION_WEAK:
391 def = ld::Atom::definitionRegular;
392 combine = ld::Atom::combineByName;
393 break;
394 case LTO_SYMBOL_DEFINITION_UNDEFINED:
395 case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
396 def = ld::Atom::definitionProxy;
397 break;
398 default:
399 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
402 // make LLVM atoms for definitions and a reference for undefines
403 if ( def != ld::Atom::definitionProxy ) {
404 ld::Atom::Scope scope;
405 bool autohide = false;
406 switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
407 case LTO_SYMBOL_SCOPE_INTERNAL:
408 scope = ld::Atom::scopeTranslationUnit;
409 break;
410 case LTO_SYMBOL_SCOPE_HIDDEN:
411 scope = ld::Atom::scopeLinkageUnit;
412 break;
413 case LTO_SYMBOL_SCOPE_DEFAULT:
414 scope = ld::Atom::scopeGlobal;
415 break;
416 #if LTO_API_VERSION >= 4
417 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
418 scope = ld::Atom::scopeGlobal;
419 autohide = true;
420 break;
421 #endif
422 default:
423 throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
425 // only make atoms for non-internal symbols
426 if ( scope == ld::Atom::scopeTranslationUnit )
427 continue;
428 uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
429 // make Atom using placement new operator
430 new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
431 if ( scope != ld::Atom::scopeTranslationUnit )
432 _internalAtom.addReference(name);
433 if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
435 else {
436 // add to list of external references
437 _internalAtom.addReference(name);
438 if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
442 #if LTO_API_VERSION >= 11
443 if ( sSupportsLocalContext )
444 this->release();
445 #endif
448 File::~File()
450 this->release();
453 bool File::mergeIntoGenerator(lto_code_gen_t generator, bool useSetModule) {
454 #if LTO_API_VERSION >= 11
455 if ( sSupportsLocalContext ) {
456 assert(!_module && "Expected module to be disposed");
457 _module = ::lto_module_create_in_codegen_context(_content, _contentLength,
458 _path, generator);
459 if ( _module == NULL )
460 throwf("could not reparse object file %s: '%s', using libLTO version '%s'",
461 _path, ::lto_get_error_message(), ::lto_get_version());
463 #endif
464 assert(_module && "Expected module to stick around");
465 #if LTO_API_VERSION >= 13
466 if (useSetModule) {
467 // lto_codegen_set_module will transfer ownership of the module to LTO code generator,
468 // so we don't need to release the module here.
469 ::lto_codegen_set_module(generator, _module);
470 return false;
472 #endif
473 if ( ::lto_codegen_add_module(generator, _module) )
474 return true;
476 // <rdar://problem/15471128> linker should release module as soon as possible
477 this->release();
478 return false;
481 void File::release()
483 if ( _module != NULL )
484 ::lto_module_dispose(_module);
485 _module = NULL;
488 bool File::forEachAtom(ld::File::AtomHandler& handler) const
490 handler.doAtom(_internalAtom);
491 for(uint32_t i=0; i < _atomArrayCount; ++i) {
492 handler.doAtom(_atomArray[i]);
494 return true;
497 InternalAtom::InternalAtom(File& f)
498 : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
499 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
500 _file(f)
504 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
505 ld::Atom::Alignment a, bool ah)
506 : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
507 ld::Atom::symbolTableIn, false, false, false, a),
508 _file(f), _name(strdup(nm)), _compiledAtom(NULL)
510 if ( ah )
511 this->setAutoHide();
514 void Atom::setCompiledAtom(const ld::Atom& atom)
516 // set delegate so virtual methods go to it
517 _compiledAtom = &atom;
519 //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
521 // update fields in ld::Atom to match newly constructed mach-o atom
522 (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
527 // <rdar://problem/12379604> The order that files are merged must match command line order
528 struct CommandLineOrderFileSorter
530 bool operator()(File* left, File* right)
532 return ( left->ordinal() < right->ordinal() );
537 #if LTO_API_VERSION >= 7
538 void Parser::ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t severity, const char* message, void*)
540 switch ( severity ) {
541 #if LTO_API_VERSION >= 10
542 case LTO_DS_REMARK:
543 fprintf(stderr, "ld: LTO remark: %s\n", message);
544 break;
545 #endif
546 case LTO_DS_NOTE:
547 case LTO_DS_WARNING:
548 warning("%s", message);
549 break;
550 case LTO_DS_ERROR:
551 throwf("%s", message);
554 #endif
556 bool Parser::optimize( const std::vector<const ld::Atom*>& allAtoms,
557 ld::Internal& state,
558 const OptimizeOptions& options,
559 ld::File::AtomHandler& handler,
560 std::vector<const ld::Atom*>& newAtoms,
561 std::vector<const char*>& additionalUndefines)
563 const bool logMustPreserve = false;
564 const bool logExtraOptions = false;
565 const bool logBitcodeFiles = false;
566 const bool logAtomsBeforeSync = false;
568 // exit quickly if nothing to do
569 if ( _s_files.size() == 0 )
570 return false;
572 // print out LTO version string if -v was used
573 if ( options.verbose )
574 fprintf(stderr, "%s\n", ::lto_get_version());
576 // create optimizer and add each Reader
577 lto_code_gen_t generator = NULL;
578 #if LTO_API_VERSION >= 11
579 if ( File::sSupportsLocalContext )
580 generator = ::lto_codegen_create_in_local_context();
581 else
582 #endif
583 generator = ::lto_codegen_create();
584 #if LTO_API_VERSION >= 7
585 lto_codegen_set_diagnostic_handler(generator, ltoDiagnosticHandler, NULL);
586 #endif
588 // <rdar://problem/12379604> The order that files are merged must match command line order
589 std::sort(_s_files.begin(), _s_files.end(), CommandLineOrderFileSorter());
590 ld::File::Ordinal lastOrdinal;
592 // When flto_codegen_only is on and we have a single .bc file, use lto_codegen_set_module instead of
593 // lto_codegen_add_module, to make sure the the destination module will be the same as the input .bc file.
594 bool useSetModule = false;
595 #if LTO_API_VERSION >= 13
596 useSetModule = (_s_files.size() == 1) && options.ltoCodegenOnly && (::lto_api_version() >= 13);
597 #endif
598 for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
599 File* f = *it;
600 assert(f->ordinal() > lastOrdinal);
601 if ( logBitcodeFiles && !useSetModule) fprintf(stderr, "lto_codegen_add_module(%s)\n", f->path());
602 if ( logBitcodeFiles && useSetModule) fprintf(stderr, "lto_codegen_set_module(%s)\n", f->path());
603 if ( f->mergeIntoGenerator(generator, useSetModule) )
604 throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", f->path(), ::lto_get_error_message(), ::lto_get_version());
605 lastOrdinal = f->ordinal();
608 // add any -mllvm command line options
609 for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
610 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
611 ::lto_codegen_debug_options(generator, *it);
614 // <rdar://problem/13687397> Need a way for LTO to get cpu variants (until that info is in bitcode)
615 if ( options.mcpu != NULL )
616 ::lto_codegen_set_cpu(generator, options.mcpu);
618 // The atom graph uses directed edges (references). Collect all references where
619 // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
620 // external (i.e. not originated from same .o file) reference if all originating atoms are also
621 // defined in llvm bitcode file.
622 CStringSet nonLLVMRefs;
623 CStringToAtom llvmAtoms;
624 bool hasNonllvmAtoms = false;
625 for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
626 const ld::Atom* atom = *it;
627 // only look at references that come from an atom that is not an llvm atom
628 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
629 if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
630 hasNonllvmAtoms = true;
632 const ld::Atom* target;
633 for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
634 switch ( fit->binding ) {
635 case ld::Fixup::bindingDirectlyBound:
636 // that reference an llvm atom
637 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
638 nonLLVMRefs.insert(fit->u.target->name());
639 break;
640 case ld::Fixup::bindingsIndirectlyBound:
641 target = state.indirectBindingTable[fit->u.bindingIndex];
642 if ( (target != NULL) && (target->contentType() == ld::Atom::typeLTOtemporary) )
643 nonLLVMRefs.insert(target->name());
644 default:
645 break;
649 else if ( atom->scope() >= ld::Atom::scopeLinkageUnit ) {
650 llvmAtoms[atom->name()] = (Atom*)atom;
653 // if entry point is in a llvm bitcode file, it must be preserved by LTO
654 if ( state.entryPoint!= NULL ) {
655 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
656 nonLLVMRefs.insert(state.entryPoint->name());
659 // deadAtoms are the atoms that the linker coalesced. For instance weak or tentative definitions
660 // overriden by another atom. If any of these deadAtoms are llvm atoms and they were replaced
661 // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
662 // atom so that the linker can replace it with the mach-o one later.
663 CStringToAtom deadllvmAtoms;
664 for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
665 const ld::Atom* atom = *it;
666 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
667 const char* name = atom->name();
668 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
669 ::lto_codegen_add_must_preserve_symbol(generator, name);
670 deadllvmAtoms[name] = (Atom*)atom;
673 for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
674 File* file = *it;
675 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
676 Atom* llvmAtom = &file->_atomArray[i];
677 if ( llvmAtom->coalescedAway() ) {
678 const char* name = llvmAtom->name();
679 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
680 if ( logMustPreserve )
681 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
682 ::lto_codegen_add_must_preserve_symbol(generator, name);
683 deadllvmAtoms[name] = (Atom*)llvmAtom;
686 else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
687 const char* name = llvmAtom->name();
688 deadllvmAtoms[name] = (Atom*)llvmAtom;
693 // tell code generator about symbols that must be preserved
694 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
695 const char* name = it->first;
696 Atom* atom = it->second;
697 // Include llvm Symbol in export list if it meets one of following two conditions
698 // 1 - atom scope is global (and not linkage unit).
699 // 2 - included in nonLLVMRefs set.
700 // If a symbol is not listed in exportList then LTO is free to optimize it away.
701 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
702 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
703 ::lto_codegen_add_must_preserve_symbol(generator, name);
705 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
706 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
707 ::lto_codegen_add_must_preserve_symbol(generator, name);
709 else if ( options.relocatable && hasNonllvmAtoms ) {
710 // <rdar://problem/14334895> ld -r mode but merging in some mach-o files, so need to keep libLTO from optimizing away anything
711 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because -r mode disable LTO dead stripping\n", name);
712 ::lto_codegen_add_must_preserve_symbol(generator, name);
716 // <rdar://problem/16165191> tell code generator to preserve initial undefines
717 for( std::vector<const char*>::const_iterator it=options.initialUndefines->begin(); it != options.initialUndefines->end(); ++it) {
718 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because it is an initial undefine\n", *it);
719 ::lto_codegen_add_must_preserve_symbol(generator, *it);
722 // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
723 if ( options.relocatable && !hasNonllvmAtoms ) {
724 #if LTO_API_VERSION >= 15
725 ::lto_codegen_set_should_embed_uselists(generator, false);
726 #endif
727 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
728 // HACK, no good way to tell linker we are all done, so just quit
729 exit(0);
731 warning("could not produce merged bitcode file");
734 // set code-gen model
735 lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
736 if ( options.mainExecutable ) {
737 if ( options.staticExecutable ) {
738 // x86_64 "static" or any "-static -pie" is really dynamic code model
739 if ( (options.arch == CPU_TYPE_X86_64) || options.pie )
740 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
741 else
742 model = LTO_CODEGEN_PIC_MODEL_STATIC;
744 else {
745 if ( options.pie )
746 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
747 else
748 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
751 else {
752 if ( options.allowTextRelocs )
753 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
754 else
755 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
757 if ( ::lto_codegen_set_pic_model(generator, model) )
758 throwf("could not create set codegen model: %s", lto_get_error_message());
760 // if requested, save off merged bitcode file
761 if ( options.saveTemps ) {
762 char tempBitcodePath[MAXPATHLEN];
763 strcpy(tempBitcodePath, options.outputFilePath);
764 strcat(tempBitcodePath, ".lto.bc");
765 #if LTO_API_VERSION >= 15
766 ::lto_codegen_set_should_embed_uselists(generator, true);
767 #endif
768 ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
771 #if LTO_API_VERSION >= 3
772 // find assembler next to linker
773 char path[PATH_MAX];
774 uint32_t bufSize = PATH_MAX;
775 if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
776 char* lastSlash = strrchr(path, '/');
777 if ( lastSlash != NULL ) {
778 strcpy(lastSlash+1, "as");
779 struct stat statInfo;
780 if ( stat(path, &statInfo) == 0 )
781 ::lto_codegen_set_assembler_path(generator, path);
784 #endif
786 // When lto API version is greater than or equal to 12, we use lto_codegen_optimize and lto_codegen_compile_optimized
787 // instead of lto_codegen_compile, and we save the merged bitcode file in between.
788 bool useSplitAPI = false;
789 #if LTO_API_VERSION >= 12
790 if ( ::lto_api_version() >= 12)
791 useSplitAPI = true;
792 #endif
794 size_t machOFileLen = 0;
795 const uint8_t* machOFile = NULL;
796 if ( useSplitAPI) {
797 #if LTO_API_VERSION >= 12
798 #if LTO_API_VERSION >= 14
799 if ( ::lto_api_version() >= 14 && options.ltoCodegenOnly)
800 lto_codegen_set_should_internalize(generator, false);
801 #endif
802 // run optimizer
803 if ( !options.ltoCodegenOnly && ::lto_codegen_optimize(generator) )
804 throwf("could not do LTO optimization: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
806 if ( options.saveTemps || options.bitcodeBundle ) {
807 // save off merged bitcode file
808 char tempOptBitcodePath[MAXPATHLEN];
809 strcpy(tempOptBitcodePath, options.outputFilePath);
810 strcat(tempOptBitcodePath, ".lto.opt.bc");
811 #if LTO_API_VERSION >= 15
812 ::lto_codegen_set_should_embed_uselists(generator, true);
813 #endif
814 ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
815 if ( options.bitcodeBundle )
816 state.ltoBitcodePath = tempOptBitcodePath;
819 // run code generator
820 machOFile = (uint8_t*)::lto_codegen_compile_optimized(generator, &machOFileLen);
821 #endif
822 if ( machOFile == NULL )
823 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
825 else {
826 // run optimizer and code generator
827 machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
828 if ( machOFile == NULL )
829 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
830 if ( options.saveTemps ) {
831 // save off merged bitcode file
832 char tempOptBitcodePath[MAXPATHLEN];
833 strcpy(tempOptBitcodePath, options.outputFilePath);
834 strcat(tempOptBitcodePath, ".lto.opt.bc");
835 #if LTO_API_VERSION >= 15
836 ::lto_codegen_set_should_embed_uselists(generator, true);
837 #endif
838 ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
842 // if requested, save off temp mach-o file
843 if ( options.saveTemps ) {
844 char tempMachoPath[MAXPATHLEN];
845 strcpy(tempMachoPath, options.outputFilePath);
846 strcat(tempMachoPath, ".lto.o");
847 int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
848 if ( fd != -1) {
849 ::write(fd, machOFile, machOFileLen);
850 ::close(fd);
854 // if needed, save temp mach-o file to specific location
855 if ( options.tmpObjectFilePath != NULL ) {
856 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
857 if ( fd != -1) {
858 ::write(fd, machOFile, machOFileLen);
859 ::close(fd);
861 else {
862 warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
866 // parse generated mach-o file into a MachOReader
867 ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
869 // sync generated mach-o atoms with existing atoms ld knows about
870 if ( logAtomsBeforeSync ) {
871 fprintf(stderr, "llvmAtoms:\n");
872 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
873 const char* name = it->first;
874 Atom* atom = it->second;
875 fprintf(stderr, "\t%p\t%s\n", atom, name);
877 fprintf(stderr, "deadllvmAtoms:\n");
878 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
879 const char* name = it->first;
880 Atom* atom = it->second;
881 fprintf(stderr, "\t%p\t%s\n", atom, name);
884 AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
885 machoFile->forEachAtom(syncer);
887 // Remove InternalAtoms from ld
888 for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
889 (*it)->internalAtom().setCoalescedAway();
891 // Remove Atoms from ld if code generator optimized them away
892 for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
893 // check if setRealAtom() called on this Atom
894 if ( li->second->compiledAtom() == NULL ) {
895 //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
896 li->second->setCoalescedAway();
900 // notify about file level attributes
901 handler.doFile(*machoFile);
903 // if final mach-o file has debug info, update original bitcode files to match
904 for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
905 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
906 machoFile->modificationTime(), machoFile->cpuSubType());
909 return true;
913 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
915 static const bool log = false;
916 static const ld::Atom* lastProxiedAtom = NULL;
917 static const ld::File* lastProxiedFile = NULL;
918 // update proxy atoms to point to real atoms and find new atoms
919 const char* name = machoAtom.name();
920 CStringToAtom::iterator pos = _llvmAtoms.find(name);
921 if ( pos != _llvmAtoms.end() ) {
922 // turn Atom into a proxy for this mach-o atom
923 pos->second->setCompiledAtom(machoAtom);
924 lastProxiedAtom = &machoAtom;
925 lastProxiedFile = pos->second->file();
926 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p synced to lto atom %p (name=%s)\n", &machoAtom, pos->second, machoAtom.name());
928 else {
929 // an atom of this name was not in the allAtoms list the linker gave us
930 if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
931 // this corresponding to an atom that the linker coalesced away or marked not-live
932 if ( _options.linkerDeadStripping ) {
933 // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
934 Atom* llvmAtom = _deadllvmAtoms[name];
935 llvmAtom->setCompiledAtom(machoAtom);
936 _newAtoms.push_back(&machoAtom);
937 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p matches dead lto atom %p but adding back (name=%s)\n", &machoAtom, llvmAtom, machoAtom.name());
939 else {
940 // Don't pass it back as a new atom
941 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p matches dead lto atom %p (name=%s)\n", &machoAtom, _deadllvmAtoms[name], machoAtom.name());
944 else
946 // this is something new that lto conjured up, tell ld its new
947 _newAtoms.push_back(&machoAtom);
948 // <rdar://problem/15469363> if new static atom in same section as previous non-static atom, assign to same file as previous
949 if ( (lastProxiedAtom != NULL) && (lastProxiedAtom->section() == machoAtom.section()) ) {
950 ld::Atom* ma = const_cast<ld::Atom*>(&machoAtom);
951 ma->setFile(lastProxiedFile);
953 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p is totally new (name=%s)\n", &machoAtom, machoAtom.name());
957 // adjust fixups to go through proxy atoms
958 if (log) fprintf(stderr, " adjusting fixups in atom: %s\n", machoAtom.name());
959 for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
960 switch ( fit->binding ) {
961 case ld::Fixup::bindingNone:
962 break;
963 case ld::Fixup::bindingByNameUnbound:
964 // don't know if this target has been seen by linker before or if it is new
965 // be conservative and tell linker it is new
966 _additionalUndefines.push_back(fit->u.name);
967 if (log) fprintf(stderr, " adding by-name symbol %s\n", fit->u.name);
968 break;
969 case ld::Fixup::bindingDirectlyBound:
970 // If mach-o atom is referencing another mach-o atom then
971 // reference is not going through Atom proxy. Fix it here to ensure that all
972 // llvm symbol references always go through Atom proxy.
974 const char* targetName = fit->u.target->name();
975 CStringToAtom::iterator post = _llvmAtoms.find(targetName);
976 if ( post != _llvmAtoms.end() ) {
977 const ld::Atom* t = post->second;
978 if (log) fprintf(stderr, " updating direct reference to %p to be ref to %p: %s\n", fit->u.target, t, targetName);
979 fit->u.target = t;
981 else {
982 // <rdar://problem/12859831> Don't unbind follow-on reference into by-name reference
983 if ( (_deadllvmAtoms.find(targetName) != _deadllvmAtoms.end()) && (fit->kind != ld::Fixup::kindNoneFollowOn) && (fit->u.target->scope() != ld::Atom::scopeTranslationUnit) ) {
984 // target was coalesed away and replace by mach-o atom from a non llvm .o file
985 fit->binding = ld::Fixup::bindingByNameUnbound;
986 fit->u.name = targetName;
990 //fprintf(stderr, " direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
991 break;
992 case ld::Fixup::bindingByContentBound:
993 //fprintf(stderr, " direct by content to: %s\n", fit->u.target->name());
994 break;
995 case ld::Fixup::bindingsIndirectlyBound:
996 assert(0 && "indirect binding found in initial mach-o file?");
997 //fprintf(stderr, " indirect by content to: %u\n", fit->u.bindingIndex);
998 break;
1004 class Mutex {
1005 static pthread_mutex_t lto_lock;
1006 public:
1007 Mutex() { pthread_mutex_lock(&lto_lock); }
1008 ~Mutex() { pthread_mutex_unlock(&lto_lock); }
1010 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
1011 bool File::sSupportsLocalContext = false;
1012 bool File::sHasTriedLocalContext = false;
1015 // Used by archive reader to see if member is an llvm bitcode file
1017 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
1019 Mutex lock;
1020 return Parser::validFile(fileContent, fileLength, architecture, subarch);
1024 static ld::relocatable::File *parseImpl(
1025 const uint8_t *fileContent, uint64_t fileLength, const char *path,
1026 time_t modTime, ld::File::Ordinal ordinal, cpu_type_t architecture,
1027 cpu_subtype_t subarch, bool logAllFiles,
1028 bool verboseOptimizationHints)
1030 if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
1031 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles, verboseOptimizationHints);
1032 else
1033 return NULL;
1037 // main function used by linker to instantiate ld::Files
1039 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
1040 const char* path, time_t modTime, ld::File::Ordinal ordinal,
1041 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles,
1042 bool verboseOptimizationHints)
1044 // Note: Once lto_module_create_in_local_context() and friends are thread safe
1045 // this lock can be removed.
1046 Mutex lock;
1047 return parseImpl(fileContent, fileLength, path, modTime, ordinal,
1048 architecture, subarch, logAllFiles,
1049 verboseOptimizationHints);
1053 // used by "ld -v" to report version of libLTO.dylib being used
1055 const char* version()
1057 Mutex lock;
1058 return ::lto_get_version();
1063 // used by ld for error reporting
1065 bool libLTOisLoaded()
1067 Mutex lock;
1068 return (::lto_get_version() != NULL);
1072 // used by ld for error reporting
1074 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
1076 Mutex lock;
1077 return Parser::fileKind(fileContent, fileLength);
1081 // used by ld for doing link time optimization
1083 bool optimize( const std::vector<const ld::Atom*>& allAtoms,
1084 ld::Internal& state,
1085 const OptimizeOptions& options,
1086 ld::File::AtomHandler& handler,
1087 std::vector<const ld::Atom*>& newAtoms,
1088 std::vector<const char*>& additionalUndefines)
1090 Mutex lock;
1091 return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
1096 }; // namespace lto
1099 #endif