src/ld/parsers/lto_file.cpp

   1 /* -*- mode: C++; c-basic-offset: 4; tab-width: 4 -*-
   2  *
   3  * Copyright (c) 2006-2010 Apple Inc. All rights reserved.
   4  *
   5  * @APPLE_LICENSE_HEADER_START@
   6  *
   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.
  13  *
  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.
  21  *
  22  * @APPLE_LICENSE_HEADER_END@
  23  */
  24
  25 #ifndef __LTO_READER_H__
  26 #define __LTO_READER_H__
  27
  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 <unordered_set>
  37 #include <unordered_map>
  38
  39 #include "MachOFileAbstraction.hpp"
  40 #include "Architectures.hpp"
  41 #include "ld.hpp"
  42 #include "macho_relocatable_file.h"
  43 #include "lto_file.h"
  44
  45 // #defines are a work around for <rdar://problem/8760268>
  46 #define __STDC_LIMIT_MACROS 1
  47 #define __STDC_CONSTANT_MACROS 1
  48 #include "llvm-c/lto.h"
  49
  50
  51 namespace lto {
  52
  53
  54 //
  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.
  60 //
  61 class InternalAtom : public ld::Atom
  62 {
  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()]; }
  74
  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, nm)); }
  79 private:
  80
  81         ld::File&                                                                       _file;
  82         mutable std::vector<ld::Fixup>                          _undefs;
  83 };
  84
  85
  86 //
  87 // LLVM bitcode file
  88 //
  89 class File : public ld::relocatable::File
  90 {
  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();
  95
  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; }
 101
 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
 110         lto_module_t                                                                            module()                                        { return _module; }
 111         class InternalAtom&                                                                     internalAtom()                          { return _internalAtom; }
 112         void                                                                                            setDebugInfo(ld::relocatable::File::DebugInfoKind k,
 113                                                                                                                                         const char* pth, time_t modTime, uint32_t subtype)
 114                                                                                                                                                                         {       _debugInfo = k;
 115                                                                                                                                                                                 _debugInfoPath = pth;
 116                                                                                                                                                                                 _debugInfoModTime = modTime;
 117                                                                                                                                                                                 _cpuSubType = subtype;}
 118
 119 private:
 120         friend class Atom;
 121         friend class InternalAtom;
 122         friend class Parser;
 123
 124         cpu_type_t                                                              _architecture;
 125         class InternalAtom                                              _internalAtom;
 126         class Atom*                                                             _atomArray;
 127         uint32_t                                                                _atomArrayCount;
 128         lto_module_t                                                    _module;
 129         const char*                                                             _debugInfoPath;
 130         time_t                                                                  _debugInfoModTime;
 131         ld::Section                                                             _section;
 132         ld::Fixup                                                               _fixupToInternal;
 133         ld::relocatable::File::DebugInfoKind    _debugInfo;
 134         uint32_t                                                                _cpuSubType;
 135 };
 136
 137 //
 138 // Atom acts as a proxy Atom for the symbols that are exported by LLVM bitcode file. Initially,
 139 // Reader creates Atoms to allow linker proceed with usual symbol resolution phase. After
 140 // optimization is performed, real Atoms are created for these symobls. However these real Atoms
 141 // are not inserted into global symbol table. Atom holds real Atom and forwards appropriate
 142 // methods to real atom.
 143 //
 144 class Atom : public ld::Atom
 145 {
 146 public:
 147                                                                                 Atom(File& f, const char* name, ld::Atom::Scope s,
 148                                                                                                 ld::Atom::Definition d, ld::Atom::Combine c, ld::Atom::Alignment a, bool ah);
 149
 150         // overrides of ld::Atom
 151         virtual ld::File*                                       file() const            { return &_file; }
 152         virtual const char*                                     translationUnitSource() const
 153                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource() : NULL); }
 154         virtual const char*                                     name() const            { return _name; }
 155         virtual uint64_t                                        size() const            { return (_compiledAtom ? _compiledAtom->size() : 0); }
 156         virtual uint64_t                                        objectAddress() const { return (_compiledAtom ? _compiledAtom->objectAddress() : 0); }
 157         virtual void                                            copyRawContent(uint8_t buffer[]) const
 158                                                                                                                         { if (_compiledAtom) _compiledAtom->copyRawContent(buffer); }
 159         virtual const uint8_t*                          rawContentPointer() const
 160                                                                                                                         { return (_compiledAtom ? _compiledAtom->rawContentPointer() : NULL);  }
 161         virtual unsigned long                           contentHash(const class ld::IndirectBindingTable& ibt) const
 162                                                                                                                         { return (_compiledAtom ? _compiledAtom->contentHash(ibt) : 0);  }
 163         virtual bool                                            canCoalesceWith(const ld::Atom& rhs, const class ld::IndirectBindingTable& ibt) const
 164                                                                                                                         { return (_compiledAtom ? _compiledAtom->canCoalesceWith(rhs,ibt) : false); }
 165         virtual ld::Fixup::iterator                             fixupsBegin() const
 166                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsBegin() : (ld::Fixup*)&_file._fixupToInternal); }
 167         virtual ld::Fixup::iterator                             fixupsEnd() const
 168                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsEnd() : &((ld::Fixup*)&_file._fixupToInternal)[1]); }
 169         virtual ld::Atom::UnwindInfo::iterator  beginUnwind() const
 170                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginUnwind() : NULL); }
 171         virtual ld::Atom::UnwindInfo::iterator  endUnwind() const
 172                                                                                                                         { return (_compiledAtom ? _compiledAtom->endUnwind() : NULL); }
 173         virtual ld::Atom::LineInfo::iterator    beginLineInfo() const
 174                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginLineInfo() : NULL); }
 175         virtual ld::Atom::LineInfo::iterator    endLineInfo() const
 176                                                                                                                         { return (_compiledAtom ? _compiledAtom->endLineInfo() : NULL); }
 177
 178         const ld::Atom*                                         compiledAtom()          { return _compiledAtom; }
 179         void                                                            setCompiledAtom(const ld::Atom& atom);
 180
 181 private:
 182
 183         File&                                                           _file;
 184         const char*                                                     _name;
 185         const ld::Atom*                                         _compiledAtom;
 186 };
 187
 188
 189
 190
 191
 192
 193
 194 class Parser
 195 {
 196 public:
 197         static bool                                             validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch);
 198         static const char*                              fileKind(const uint8_t* fileContent, uint64_t fileLength);
 199         static File*                                    parse(const uint8_t* fileContent, uint64_t fileLength, const char* path,
 200                                                                                         time_t modTime, ld::File::Ordinal ordinal, cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles);
 201         static bool                                             libLTOisLoaded() { return (::lto_get_version() != NULL); }
 202         static bool                                             optimize(   const std::vector<const ld::Atom*>& allAtoms,
 203                                                                                                 ld::Internal&                                           state,
 204                                                                                                 const OptimizeOptions&                          options,
 205                                                                                                 ld::File::AtomHandler&                          handler,
 206                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 207                                                                                                 std::vector<const char*>&                       additionalUndefines);
 208
 209         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 210
 211 private:
 212         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 213         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options);
 214
 215         typedef std::unordered_set<const char*, ld::CStringHash, ld::CStringEquals>  CStringSet;
 216         typedef std::unordered_map<const char*, Atom*, ld::CStringHash, ld::CStringEquals> CStringToAtom;
 217
 218         class AtomSyncer : public ld::File::AtomHandler {
 219         public:
 220                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 221                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 222                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 223                 virtual void            doAtom(const class ld::Atom&);
 224                 virtual void            doFile(const class ld::File&) { }
 225
 226                 const OptimizeOptions&                  _options;
 227                 std::vector<const char*>&               _additionalUndefines;
 228                 std::vector<const ld::Atom*>&   _newAtoms;
 229                 CStringToAtom                                   _llvmAtoms;
 230                 CStringToAtom                                   _deadllvmAtoms;
 231         };
 232
 233         static std::vector<File*>               _s_files;
 234 };
 235
 236 std::vector<File*> Parser::_s_files;
 237
 238
 239 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 240 {
 241         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 242                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 243                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 244                         if ( !result ) {
 245                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 246                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 247                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 248                                 }
 249                         }
 250                         return result;
 251                 }
 252         }
 253         return false;
 254 }
 255
 256 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 257 {
 258         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 259                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 260                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 261                         if ( arch == t->cpuType ) {
 262                                  if ( t->isSubType ) {
 263                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 264                                                 return t->archName;
 265                                 }
 266                                 else {
 267                                         return t->archName;
 268                                 }
 269                         }
 270                 }
 271                 return "unknown bitcode architecture";
 272         }
 273         return NULL;
 274 }
 275
 276 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime, ld::File::Ordinal ordinal,
 277                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 278 {
 279         File* f = new File(path, modTime, ordinal, fileContent, fileLength, architecture);
 280         _s_files.push_back(f);
 281         if ( logAllFiles )
 282                 printf("%s\n", path);
 283         return f;
 284 }
 285
 286
 287 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 288 {
 289         mach_o::relocatable::ParserOptions objOpts;
 290         objOpts.architecture            = options.arch;
 291         objOpts.objSubtypeMustMatch = false;
 292         objOpts.logAllFiles                     = false;
 293         objOpts.convertUnwindInfo       = true;
 294         objOpts.subType                         = 0;
 295
 296         // mach-o parsing is done in-memory, but need path for debug notes
 297         const char* path = "/tmp/lto.o";
 298         time_t modTime = 0;
 299         if ( options.tmpObjectFilePath != NULL ) {
 300                 path = options.tmpObjectFilePath;
 301                 struct stat statBuffer;
 302                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 303                         modTime = statBuffer.st_mtime;
 304         }
 305
 306         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 307         if ( result != NULL )
 308                 return result;
 309         throw "LLVM LTO, file is not of required architecture";
 310 }
 311
 312
 313
 314 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 315         : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
 316         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 317         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 318         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 319         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 320 {
 321         const bool log = false;
 322
 323         // create llvm module
 324         _module = ::lto_module_create_from_memory(content, contentLength);
 325     if ( _module == NULL )
 326                 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth, ::lto_get_error_message(), ::lto_get_version());
 327
 328         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 329
 330         // create atom for each global symbol in module
 331         uint32_t count = ::lto_module_get_num_symbols(_module);
 332         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 333         for (uint32_t i=0; i < count; ++i) {
 334                 const char* name = ::lto_module_get_symbol_name(_module, i);
 335                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 336
 337                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 338                 // ignore dtrace static probes for now
 339                 // later when codegen is done and a mach-o file is produces the probes will be processed
 340                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 341                         continue;
 342
 343                 ld::Atom::Definition def;
 344                 ld::Atom::Combine combine = ld::Atom::combineNever;
 345                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 346                         case LTO_SYMBOL_DEFINITION_REGULAR:
 347                                 def = ld::Atom::definitionRegular;
 348                                 break;
 349                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 350                                 def = ld::Atom::definitionTentative;
 351                                 break;
 352                         case LTO_SYMBOL_DEFINITION_WEAK:
 353                                 def = ld::Atom::definitionRegular;
 354                                 combine = ld::Atom::combineByName;
 355                                 break;
 356                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 357                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 358                                 def = ld::Atom::definitionProxy;
 359                                 break;
 360                         default:
 361                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 362                 }
 363
 364                 // make LLVM atoms for definitions and a reference for undefines
 365                 if ( def != ld::Atom::definitionProxy ) {
 366                         ld::Atom::Scope scope;
 367                         bool autohide = false;
 368                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 369                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 370                                         scope = ld::Atom::scopeTranslationUnit;
 371                                         break;
 372                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 373                                         scope = ld::Atom::scopeLinkageUnit;
 374                                         break;
 375                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 376                                         scope = ld::Atom::scopeGlobal;
 377                                         break;
 378 #if LTO_API_VERSION >= 4
 379                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 380                                         scope = ld::Atom::scopeGlobal;
 381                                         autohide = true;
 382                                         break;
 383 #endif
 384                                 default:
 385                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 386                         }
 387                         // only make atoms for non-internal symbols
 388                         if ( scope == ld::Atom::scopeTranslationUnit )
 389                                 continue;
 390                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 391                         // make Atom using placement new operator
 392                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 393                         if ( scope != ld::Atom::scopeTranslationUnit )
 394                                 _internalAtom.addReference(name);
 395                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 396                 }
 397                 else {
 398                         // add to list of external references
 399                         _internalAtom.addReference(name);
 400                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 401                 }
 402         }
 403 }
 404
 405 File::~File()
 406 {
 407         if ( _module != NULL )
 408                 ::lto_module_dispose(_module);
 409 }
 410
 411 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 412 {
 413         handler.doAtom(_internalAtom);
 414         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 415                 handler.doAtom(_atomArray[i]);
 416         }
 417         return true;
 418 }
 419
 420 InternalAtom::InternalAtom(File& f)
 421         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 422                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 423                 _file(f)
 424 {
 425 }
 426
 427 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 428                         ld::Atom::Alignment a, bool ah)
 429         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 430                                 ld::Atom::symbolTableIn, false, false, false, a),
 431                 _file(f), _name(nm), _compiledAtom(NULL)
 432 {
 433         if ( ah )
 434                 this->setAutoHide();
 435 }
 436
 437 void Atom::setCompiledAtom(const ld::Atom& atom)
 438 {
 439         // set delegate so virtual methods go to it
 440         _compiledAtom = &atom;
 441
 442         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 443
 444         // update fields in ld::Atom to match newly constructed mach-o atom
 445         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 446 }
 447
 448
 449
 450 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 451                                                 ld::Internal&                                           state,
 452                                                 const OptimizeOptions&                          options,
 453                                                 ld::File::AtomHandler&                          handler,
 454                                                 std::vector<const ld::Atom*>&           newAtoms,
 455                                                 std::vector<const char*>&                       additionalUndefines)
 456 {
 457         const bool logMustPreserve = false;
 458         const bool logExtraOptions = false;
 459         const bool logBitcodeFiles = false;
 460         const bool logAtomsBeforeSync = false;
 461
 462         // exit quickly if nothing to do
 463         if ( _s_files.size() == 0 )
 464                 return false;
 465
 466         // print out LTO version string if -v was used
 467         if ( options.verbose )
 468                 fprintf(stderr, "%s\n", ::lto_get_version());
 469
 470         // create optimizer and add each Reader
 471         lto_code_gen_t generator = ::lto_codegen_create();
 472         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 473                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", (*it)->path());
 474                 if ( ::lto_codegen_add_module(generator, (*it)->module()) )
 475                         throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", (*it)->path(), ::lto_get_error_message(), ::lto_get_version());
 476         }
 477
 478         // add any -mllvm command line options
 479         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 480                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 481                 ::lto_codegen_debug_options(generator, *it);
 482         }
 483
 484         // The atom graph uses directed edges (references). Collect all references where
 485         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 486         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 487         // defined in llvm bitcode file.
 488         CStringSet nonLLVMRefs;
 489         CStringToAtom llvmAtoms;
 490     bool hasNonllvmAtoms = false;
 491         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 492                 const ld::Atom* atom = *it;
 493                 // only look at references that come from an atom that is not an llvm atom
 494                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 495                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 496                                 hasNonllvmAtoms = true;
 497                         }
 498                         const ld::Atom* target;
 499                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 500                                 switch ( fit->binding ) {
 501                                         case ld::Fixup::bindingDirectlyBound:
 502                                                 // that reference an llvm atom
 503                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 504                                                         nonLLVMRefs.insert(fit->u.target->name());
 505                                                 break;
 506                                         case ld::Fixup::bindingsIndirectlyBound:
 507                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 508                                                 if ( target == NULL )
 509                                                         throwf("'%s' in %s contains undefined reference", atom->name(), atom->file()->path());
 510                                                 assert(target != NULL);
 511                                                 if ( target->contentType() == ld::Atom::typeLTOtemporary )
 512                                                         nonLLVMRefs.insert(target->name());
 513                                         default:
 514                                                 break;
 515                                 }
 516                         }
 517                 }
 518                 else {
 519                         llvmAtoms[atom->name()] = (Atom*)atom;
 520                 }
 521         }
 522         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 523         if ( state.entryPoint!= NULL ) {
 524                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 525                         nonLLVMRefs.insert(state.entryPoint->name());
 526         }
 527
 528         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 529         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 530         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 531         // atom so that the linker can replace it with the mach-o one later.
 532         CStringToAtom deadllvmAtoms;
 533         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 534                 const ld::Atom* atom = *it;
 535                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 536                         const char* name = atom->name();
 537                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 538                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 539                         deadllvmAtoms[name] = (Atom*)atom;
 540                 }
 541         }
 542         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 543                 File* file = *it;
 544                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 545                         Atom* llvmAtom = &file->_atomArray[i];
 546                         if ( llvmAtom->coalescedAway()  ) {
 547                                 const char* name = llvmAtom->name();
 548                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 549                                         if ( logMustPreserve )
 550                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 551                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 552                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 553                                 }
 554                         }
 555                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 556                                 const char* name = llvmAtom->name();
 557                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 558                         }
 559                 }
 560         }
 561
 562         // tell code generator about symbols that must be preserved
 563         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 564                 const char* name = it->first;
 565                 Atom* atom = it->second;
 566                 // Include llvm Symbol in export list if it meets one of following two conditions
 567                 // 1 - atom scope is global (and not linkage unit).
 568                 // 2 - included in nonLLVMRefs set.
 569                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 570                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 571                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 572                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 573                 }
 574                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 575                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 576                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 577                 }
 578         }
 579
 580     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 581     if ( options.relocatable && !hasNonllvmAtoms ) {
 582                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 583                         // HACK, no good way to tell linker we are all done, so just quit
 584                         exit(0);
 585                 }
 586                 warning("could not produce merged bitcode file");
 587     }
 588
 589         // set code-gen model
 590         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 591         if ( options.mainExecutable ) {
 592                 if ( options.staticExecutable ) {
 593                         // x86_64 "static" or any "-static -pie" is really dynamic code model
 594                         if ( (options.arch == CPU_TYPE_X86_64) || options.pie )
 595                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 596                         else
 597                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 598                 }
 599                 else {
 600                         if ( options.pie )
 601                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 602                         else
 603                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 604                 }
 605         }
 606         else {
 607                 if ( options.allowTextRelocs )
 608                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 609                 else
 610                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 611         }
 612         if ( ::lto_codegen_set_pic_model(generator, model) )
 613                 throwf("could not create set codegen model: %s", lto_get_error_message());
 614
 615     // if requested, save off merged bitcode file
 616     if ( options.saveTemps ) {
 617         char tempBitcodePath[MAXPATHLEN];
 618         strcpy(tempBitcodePath, options.outputFilePath);
 619         strcat(tempBitcodePath, ".lto.bc");
 620         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 621     }
 622
 623 #if LTO_API_VERSION >= 3
 624         // find assembler next to linker
 625         char path[PATH_MAX];
 626         uint32_t bufSize = PATH_MAX;
 627         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 628                 char* lastSlash = strrchr(path, '/');
 629                 if ( lastSlash != NULL ) {
 630                         strcpy(lastSlash+1, "as");
 631                         struct stat statInfo;
 632                         if ( stat(path, &statInfo) == 0 )
 633                                 ::lto_codegen_set_assembler_path(generator, path);
 634                 }
 635         }
 636 #endif
 637         // run code generator
 638         size_t machOFileLen;
 639         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 640         if ( machOFile == NULL )
 641                 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
 642
 643     // if requested, save off temp mach-o file
 644     if ( options.saveTemps ) {
 645         char tempMachoPath[MAXPATHLEN];
 646         strcpy(tempMachoPath, options.outputFilePath);
 647         strcat(tempMachoPath, ".lto.o");
 648         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 649                 if ( fd != -1) {
 650                         ::write(fd, machOFile, machOFileLen);
 651                         ::close(fd);
 652                 }
 653                 //      save off merged bitcode file
 654                 char tempOptBitcodePath[MAXPATHLEN];
 655         strcpy(tempOptBitcodePath, options.outputFilePath);
 656         strcat(tempOptBitcodePath, ".lto.opt.bc");
 657         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 658         }
 659
 660         // if needed, save temp mach-o file to specific location
 661         if ( options.tmpObjectFilePath != NULL ) {
 662                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 663                 if ( fd != -1) {
 664                         ::write(fd, machOFile, machOFileLen);
 665                         ::close(fd);
 666                 }
 667                 else {
 668                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 669                 }
 670         }
 671
 672         // parse generated mach-o file into a MachOReader
 673         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 674
 675         // sync generated mach-o atoms with existing atoms ld knows about
 676         if ( logAtomsBeforeSync ) {
 677                 fprintf(stderr, "llvmAtoms:\n");
 678                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 679                         const char* name = it->first;
 680                         //Atom* atom = it->second;
 681                         fprintf(stderr, "\t%s\n", name);
 682                 }
 683                 fprintf(stderr, "deadllvmAtoms:\n");
 684                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 685                         const char* name = it->first;
 686                         //Atom* atom = it->second;
 687                         fprintf(stderr, "\t%s\n", name);
 688                 }
 689         }
 690         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 691         machoFile->forEachAtom(syncer);
 692
 693         // Remove InternalAtoms from ld
 694         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 695                 (*it)->internalAtom().setCoalescedAway();
 696         }
 697         // Remove Atoms from ld if code generator optimized them away
 698         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 699                 // check if setRealAtom() called on this Atom
 700                 if ( li->second->compiledAtom() == NULL ) {
 701                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 702                         li->second->setCoalescedAway();
 703                 }
 704         }
 705
 706         // notify about file level attributes
 707         handler.doFile(*machoFile);
 708
 709         // if final mach-o file has debug info, update original bitcode files to match
 710         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 711                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 712                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 713         }
 714
 715         return true;
 716 }
 717
 718
 719 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 720 {
 721         // update proxy atoms to point to real atoms and find new atoms
 722         const char* name = machoAtom.name();
 723         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 724                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 725                 if ( pos != _llvmAtoms.end() ) {
 726                         // turn Atom into a proxy for this mach-o atom
 727                         pos->second->setCompiledAtom(machoAtom);
 728                 }
 729                 else {
 730                         // an atom of this name was not in the allAtoms list the linker gave us
 731                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 732                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 733                                 if ( _options.linkerDeadStripping ) {
 734                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 735                                         Atom* llvmAtom = _deadllvmAtoms[name];
 736                                         llvmAtom->setCompiledAtom(machoAtom);
 737                                         _newAtoms.push_back(&machoAtom);
 738                                 }
 739                                 else {
 740                                         // Don't pass it back as a new atom
 741                                 }
 742                         }
 743                         else
 744                         {
 745                                 // this is something new that lto conjured up, tell ld its new
 746                                 _newAtoms.push_back(&machoAtom);
 747                         }
 748                 }
 749         }
 750         else {
 751                 // ld only knew about non-static atoms, so this one must be new
 752                 _newAtoms.push_back(&machoAtom);
 753         }
 754
 755         // adjust fixups to go through proxy atoms
 756         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 757         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 758                 switch ( fit->binding ) {
 759                         case ld::Fixup::bindingNone:
 760                                 break;
 761                         case ld::Fixup::bindingByNameUnbound:
 762                                 // don't know if this target has been seen by linker before or if it is new
 763                                 // be conservative and tell linker it is new
 764                                 _additionalUndefines.push_back(fit->u.name);
 765                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 766                                 break;
 767                         case ld::Fixup::bindingDirectlyBound:
 768                                 // If mach-o atom is referencing another mach-o atom then
 769                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 770                                 // llvm symbol references always go through Atom proxy.
 771                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 772                                         const char* targetName = fit->u.target->name();
 773                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 774                                         if ( pos != _llvmAtoms.end() ) {
 775                                                 fit->u.target = pos->second;
 776                                         }
 777                                         else {
 778                                                 if ( _deadllvmAtoms.find(targetName) != _deadllvmAtoms.end() ) {
 779                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 780                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 781                                                         fit->u.name = targetName;
 782                                                 }
 783                                         }
 784                                 }
 785                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 786                                 break;
 787                         case ld::Fixup::bindingByContentBound:
 788                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 789                                 break;
 790                         case ld::Fixup::bindingsIndirectlyBound:
 791                                 assert(0 && "indirect binding found in initial mach-o file?");
 792                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 793                                 break;
 794                 }
 795         }
 796
 797 }
 798
 799 class Mutex {
 800         static pthread_mutex_t lto_lock;
 801 public:
 802         Mutex() { pthread_mutex_lock(&lto_lock); }
 803         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 804 };
 805 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 806
 807 //
 808 // Used by archive reader to see if member is an llvm bitcode file
 809 //
 810 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 811 {
 812         Mutex lock;
 813         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 814 }
 815
 816
 817 //
 818 // main function used by linker to instantiate ld::Files
 819 //
 820 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 821                                                                 const char* path, time_t modTime, ld::File::Ordinal ordinal,
 822                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 823 {
 824         Mutex lock;
 825         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 826                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles);
 827         else
 828                 return NULL;
 829 }
 830
 831 //
 832 // used by "ld -v" to report version of libLTO.dylib being used
 833 //
 834 const char* version()
 835 {
 836         Mutex lock;
 837         return ::lto_get_version();
 838 }
 839
 840
 841 //
 842 // used by ld for error reporting
 843 //
 844 bool libLTOisLoaded()
 845 {
 846         Mutex lock;
 847         return (::lto_get_version() != NULL);
 848 }
 849
 850 //
 851 // used by ld for error reporting
 852 //
 853 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 854 {
 855         Mutex lock;
 856         return Parser::fileKind(fileContent, fileLength);
 857 }
 858
 859 //
 860 // used by ld for doing link time optimization
 861 //
 862 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 863                                 ld::Internal&                                           state,
 864                                 const OptimizeOptions&                          options,
 865                                 ld::File::AtomHandler&                          handler,
 866                                 std::vector<const ld::Atom*>&           newAtoms,
 867                                 std::vector<const char*>&                       additionalUndefines)
 868 {
 869         Mutex lock;
 870         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 871 }
 872
 873
 874
 875 }; // namespace lto
 876
 877
 878 #endif
 879