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 <mach-o/dyld.h>
  34 #include <vector>
  35 #include <ext/hash_set>
  36 #include <ext/hash_map>
  37
  38 #include "MachOFileAbstraction.hpp"
  39 #include "Architectures.hpp"
  40 #include "ld.hpp"
  41 #include "macho_relocatable_file.h"
  42 #include "lto_file.h"
  43
  44 // #defines are a work around for <rdar://problem/8760268>
  45 #define __STDC_LIMIT_MACROS 1
  46 #define __STDC_CONSTANT_MACROS 1
  47 #include "llvm-c/lto.h"
  48
  49
  50 namespace lto {
  51
  52
  53 //
  54 // ld64 only tracks non-internal symbols from an llvm bitcode file.
  55 // We model this by having an InternalAtom which represent all internal functions and data.
  56 // All non-interal symbols from a bitcode file are represented by an Atom
  57 // and each Atom has a reference to the InternalAtom.  The InternalAtom
  58 // also has references to each symbol external to the bitcode file.
  59 //
  60 class InternalAtom : public ld::Atom
  61 {
  62 public:
  63                                                                                                 InternalAtom(class File& f);
  64         // overrides of ld::Atom
  65         virtual ld::File*                                                       file() const            { return &_file; }
  66         virtual const char*                                                     name() const            { return "import-atom"; }
  67         virtual uint64_t                                                        size() const            { return 0; }
  68         virtual uint64_t                                                        objectAddress() const { return 0; }
  69         virtual void                                                            copyRawContent(uint8_t buffer[]) const { }
  70         virtual void                                                            setScope(Scope)         { }
  71         virtual ld::Fixup::iterator                                     fixupsBegin() const     { return &_undefs[0]; }
  72         virtual ld::Fixup::iterator                                     fixupsEnd()     const   { return &_undefs[_undefs.size()]; }
  73
  74         // for adding references to symbols outside bitcode file
  75         void                                                                            addReference(const char* nm)
  76                                                                                                                                         { _undefs.push_back(ld::Fixup(0, ld::Fixup::k1of1,
  77                                                                                                                                                                 ld::Fixup::kindNone, false, nm)); }
  78 private:
  79
  80         ld::File&                                                                       _file;
  81         mutable std::vector<ld::Fixup>                          _undefs;
  82 };
  83
  84
  85 //
  86 // LLVM bitcode file
  87 //
  88 class File : public ld::relocatable::File
  89 {
  90 public:
  91                                                                                         File(const char* path, time_t mTime, ld::File::Ordinal ordinal,
  92                                                                                                          const uint8_t* content, uint32_t contentLength, cpu_type_t arch);
  93         virtual                                                                 ~File();
  94
  95         // overrides of ld::File
  96         virtual bool                                                                            forEachAtom(ld::File::AtomHandler&) const;
  97         virtual bool                                                                            justInTimeforEachAtom(const char* name, ld::File::AtomHandler&) const
  98                                                                                                                                                                         { return false; }
  99         virtual uint32_t                                                                        cpuSubType() const                      { return _cpuSubType; }
 100
 101         // overrides of ld::relocatable::File
 102         virtual DebugInfoKind                                                           debugInfo()     const                   { return _debugInfo; }
 103         virtual const char*                                                                     debugInfoPath() const           { return _debugInfoPath; }
 104         virtual time_t                                                                          debugInfoModificationTime() const
 105                                                                                                                                                                         { return _debugInfoModTime; }
 106         virtual const std::vector<ld::relocatable::File::Stab>* stabs() const                   { return NULL; }
 107         virtual bool                                                                            canScatterAtoms() const         { return true; }
 108
 109         lto_module_t                                                                            module()                                        { return _module; }
 110         class InternalAtom&                                                                     internalAtom()                          { return _internalAtom; }
 111         void                                                                                            setDebugInfo(ld::relocatable::File::DebugInfoKind k,
 112                                                                                                                                         const char* pth, time_t modTime, uint32_t subtype)
 113                                                                                                                                                                         {       _debugInfo = k;
 114                                                                                                                                                                                 _debugInfoPath = pth;
 115                                                                                                                                                                                 _debugInfoModTime = modTime;
 116                                                                                                                                                                                 _cpuSubType = subtype;}
 117
 118 private:
 119         friend class Atom;
 120         friend class InternalAtom;
 121         friend class Parser;
 122
 123         cpu_type_t                                                              _architecture;
 124         class InternalAtom                                              _internalAtom;
 125         class Atom*                                                             _atomArray;
 126         uint32_t                                                                _atomArrayCount;
 127         lto_module_t                                                    _module;
 128         const char*                                                             _debugInfoPath;
 129         time_t                                                                  _debugInfoModTime;
 130         ld::Section                                                             _section;
 131         ld::Fixup                                                               _fixupToInternal;
 132         ld::relocatable::File::DebugInfoKind    _debugInfo;
 133         uint32_t                                                                _cpuSubType;
 134 };
 135
 136 //
 137 // Atom acts as a proxy Atom for the symbols that are exported by LLVM bitcode file. Initially,
 138 // Reader creates Atoms to allow linker proceed with usual symbol resolution phase. After
 139 // optimization is performed, real Atoms are created for these symobls. However these real Atoms
 140 // are not inserted into global symbol table. Atom holds real Atom and forwards appropriate
 141 // methods to real atom.
 142 //
 143 class Atom : public ld::Atom
 144 {
 145 public:
 146                                                                                 Atom(File& f, const char* name, ld::Atom::Scope s,
 147                                                                                                 ld::Atom::Definition d, ld::Atom::Combine c, ld::Atom::Alignment a, bool ah);
 148
 149         // overrides of ld::Atom
 150         virtual ld::File*                                       file() const            { return &_file; }
 151         virtual const char*                                     translationUnitSource() const
 152                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource() : NULL); }
 153         virtual const char*                                     name() const            { return _name; }
 154         virtual uint64_t                                        size() const            { return (_compiledAtom ? _compiledAtom->size() : 0); }
 155         virtual uint64_t                                        objectAddress() const { return (_compiledAtom ? _compiledAtom->objectAddress() : 0); }
 156         virtual void                                            copyRawContent(uint8_t buffer[]) const
 157                                                                                                                         { if (_compiledAtom) _compiledAtom->copyRawContent(buffer); }
 158         virtual const uint8_t*                          rawContentPointer() const
 159                                                                                                                         { return (_compiledAtom ? _compiledAtom->rawContentPointer() : NULL);  }
 160         virtual unsigned long                           contentHash(const class ld::IndirectBindingTable& ibt) const
 161                                                                                                                         { return (_compiledAtom ? _compiledAtom->contentHash(ibt) : 0);  }
 162         virtual bool                                            canCoalesceWith(const ld::Atom& rhs, const class ld::IndirectBindingTable& ibt) const
 163                                                                                                                         { return (_compiledAtom ? _compiledAtom->canCoalesceWith(rhs,ibt) : false); }
 164         virtual ld::Fixup::iterator                             fixupsBegin() const
 165                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsBegin() : (ld::Fixup*)&_file._fixupToInternal); }
 166         virtual ld::Fixup::iterator                             fixupsEnd() const
 167                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsEnd() : &((ld::Fixup*)&_file._fixupToInternal)[1]); }
 168         virtual ld::Atom::UnwindInfo::iterator  beginUnwind() const
 169                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginUnwind() : NULL); }
 170         virtual ld::Atom::UnwindInfo::iterator  endUnwind() const
 171                                                                                                                         { return (_compiledAtom ? _compiledAtom->endUnwind() : NULL); }
 172         virtual ld::Atom::LineInfo::iterator    beginLineInfo() const
 173                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginLineInfo() : NULL); }
 174         virtual ld::Atom::LineInfo::iterator    endLineInfo() const
 175                                                                                                                         { return (_compiledAtom ? _compiledAtom->endLineInfo() : NULL); }
 176
 177         const ld::Atom*                                         compiledAtom()          { return _compiledAtom; }
 178         void                                                            setCompiledAtom(const ld::Atom& atom);
 179
 180 private:
 181
 182         File&                                                           _file;
 183         const char*                                                     _name;
 184         const ld::Atom*                                         _compiledAtom;
 185 };
 186
 187
 188
 189
 190
 191
 192
 193 class Parser
 194 {
 195 public:
 196         static bool                                             validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch);
 197         static const char*                              fileKind(const uint8_t* fileContent, uint64_t fileLength);
 198         static File*                                    parse(const uint8_t* fileContent, uint64_t fileLength, const char* path,
 199                                                                                         time_t modTime, ld::File::Ordinal ordinal, cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles);
 200         static bool                                             libLTOisLoaded() { return (::lto_get_version() != NULL); }
 201         static bool                                             optimize(   const std::vector<const ld::Atom*>& allAtoms,
 202                                                                                                 ld::Internal&                                           state,
 203                                                                                                 const OptimizeOptions&                          options,
 204                                                                                                 ld::File::AtomHandler&                          handler,
 205                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 206                                                                                                 std::vector<const char*>&                       additionalUndefines);
 207
 208         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 209
 210 private:
 211         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 212         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options);
 213
 214         class CStringEquals
 215         {
 216         public:
 217                 bool operator()(const char* left, const char* right) const { return (strcmp(left, right) == 0); }
 218         };
 219         typedef __gnu_cxx::hash_set<const char*, __gnu_cxx::hash<const char*>, CStringEquals>  CStringSet;
 220         typedef __gnu_cxx::hash_map<const char*, Atom*, __gnu_cxx::hash<const char*>, CStringEquals> CStringToAtom;
 221
 222         class AtomSyncer : public ld::File::AtomHandler {
 223         public:
 224                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 225                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 226                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 227                 virtual void            doAtom(const class ld::Atom&);
 228                 virtual void            doFile(const class ld::File&) { }
 229
 230                 const OptimizeOptions&                  _options;
 231                 std::vector<const char*>&               _additionalUndefines;
 232                 std::vector<const ld::Atom*>&   _newAtoms;
 233                 CStringToAtom                                   _llvmAtoms;
 234                 CStringToAtom                                   _deadllvmAtoms;
 235         };
 236
 237         static std::vector<File*>               _s_files;
 238 };
 239
 240 std::vector<File*> Parser::_s_files;
 241
 242
 243 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 244 {
 245         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 246                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 247                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 248                         if ( !result ) {
 249                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 250                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 251                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 252                                 }
 253                         }
 254                         return result;
 255                 }
 256         }
 257         return false;
 258 }
 259
 260 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 261 {
 262         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 263                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 264                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 265                         if ( arch == t->cpuType ) {
 266                                  if ( t->isSubType ) {
 267                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 268                                                 return t->archName;
 269                                 }
 270                                 else {
 271                                         return t->archName;
 272                                 }
 273                         }
 274                 }
 275                 return "unknown bitcode architecture";
 276         }
 277         return NULL;
 278 }
 279
 280 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime, ld::File::Ordinal ordinal,
 281                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 282 {
 283         File* f = new File(path, modTime, ordinal, fileContent, fileLength, architecture);
 284         _s_files.push_back(f);
 285         if ( logAllFiles )
 286                 printf("%s\n", path);
 287         return f;
 288 }
 289
 290
 291 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 292 {
 293         mach_o::relocatable::ParserOptions objOpts;
 294         objOpts.architecture            = options.arch;
 295         objOpts.objSubtypeMustMatch = false;
 296         objOpts.logAllFiles                     = false;
 297         objOpts.convertUnwindInfo       = true;
 298         objOpts.subType                         = 0;
 299
 300         // mach-o parsing is done in-memory, but need path for debug notes
 301         const char* path = "/tmp/lto.o";
 302         time_t modTime = 0;
 303         if ( options.tmpObjectFilePath != NULL ) {
 304                 path = options.tmpObjectFilePath;
 305                 struct stat statBuffer;
 306                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 307                         modTime = statBuffer.st_mtime;
 308         }
 309
 310         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 311         if ( result != NULL )
 312                 return result;
 313         throw "LLVM LTO, file is not of required architecture";
 314 }
 315
 316
 317
 318 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 319         : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
 320         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 321         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 322         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 323         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 324 {
 325         const bool log = false;
 326
 327         // create llvm module
 328         _module = ::lto_module_create_from_memory(content, contentLength);
 329     if ( _module == NULL )
 330                 throwf("could not parse object file %s: %s", pth, lto_get_error_message());
 331
 332         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 333
 334         // create atom for each global symbol in module
 335         uint32_t count = ::lto_module_get_num_symbols(_module);
 336         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 337         for (uint32_t i=0; i < count; ++i) {
 338                 const char* name = ::lto_module_get_symbol_name(_module, i);
 339                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 340
 341                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 342                 // ignore dtrace static probes for now
 343                 // later when codegen is done and a mach-o file is produces the probes will be processed
 344                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 345                         continue;
 346
 347                 ld::Atom::Definition def;
 348                 ld::Atom::Combine combine = ld::Atom::combineNever;
 349                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 350                         case LTO_SYMBOL_DEFINITION_REGULAR:
 351                                 def = ld::Atom::definitionRegular;
 352                                 break;
 353                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 354                                 def = ld::Atom::definitionTentative;
 355                                 break;
 356                         case LTO_SYMBOL_DEFINITION_WEAK:
 357                                 def = ld::Atom::definitionRegular;
 358                                 combine = ld::Atom::combineByName;
 359                                 break;
 360                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 361                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 362                                 def = ld::Atom::definitionProxy;
 363                                 break;
 364                         default:
 365                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 366                 }
 367
 368                 // make LLVM atoms for definitions and a reference for undefines
 369                 if ( def != ld::Atom::definitionProxy ) {
 370                         ld::Atom::Scope scope;
 371                         bool autohide = false;
 372                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 373                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 374                                         scope = ld::Atom::scopeTranslationUnit;
 375                                         break;
 376                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 377                                         scope = ld::Atom::scopeLinkageUnit;
 378                                         break;
 379                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 380                                         scope = ld::Atom::scopeGlobal;
 381                                         break;
 382 #if LTO_API_VERSION >= 4
 383                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 384                                         scope = ld::Atom::scopeGlobal;
 385                                         autohide = true;
 386                                         break;
 387 #endif
 388                                 default:
 389                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 390                         }
 391                         // only make atoms for non-internal symbols
 392                         if ( scope == ld::Atom::scopeTranslationUnit )
 393                                 continue;
 394                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 395                         // make Atom using placement new operator
 396                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 397                         if ( scope != ld::Atom::scopeTranslationUnit )
 398                                 _internalAtom.addReference(name);
 399                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 400                 }
 401                 else {
 402                         // add to list of external references
 403                         _internalAtom.addReference(name);
 404                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 405                 }
 406         }
 407 }
 408
 409 File::~File()
 410 {
 411         if ( _module != NULL )
 412                 ::lto_module_dispose(_module);
 413 }
 414
 415 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 416 {
 417         handler.doAtom(_internalAtom);
 418         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 419                 handler.doAtom(_atomArray[i]);
 420         }
 421         return true;
 422 }
 423
 424 InternalAtom::InternalAtom(File& f)
 425         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 426                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 427                 _file(f)
 428 {
 429 }
 430
 431 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 432                         ld::Atom::Alignment a, bool ah)
 433         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 434                                 ld::Atom::symbolTableIn, false, false, false, a),
 435                 _file(f), _name(nm), _compiledAtom(NULL)
 436 {
 437         if ( ah )
 438                 this->setAutoHide();
 439 }
 440
 441 void Atom::setCompiledAtom(const ld::Atom& atom)
 442 {
 443         // set delegate so virtual methods go to it
 444         _compiledAtom = &atom;
 445
 446         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 447
 448         // update fields in ld::Atom to match newly constructed mach-o atom
 449         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 450 }
 451
 452
 453
 454 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 455                                                 ld::Internal&                                           state,
 456                                                 const OptimizeOptions&                          options,
 457                                                 ld::File::AtomHandler&                          handler,
 458                                                 std::vector<const ld::Atom*>&           newAtoms,
 459                                                 std::vector<const char*>&                       additionalUndefines)
 460 {
 461         const bool logMustPreserve = false;
 462         const bool logExtraOptions = false;
 463         const bool logBitcodeFiles = false;
 464         const bool logAtomsBeforeSync = false;
 465
 466         // exit quickly if nothing to do
 467         if ( _s_files.size() == 0 )
 468                 return false;
 469
 470         // print out LTO version string if -v was used
 471         if ( options.verbose )
 472                 fprintf(stderr, "%s\n", lto_get_version());
 473
 474         // create optimizer and add each Reader
 475         lto_code_gen_t generator = ::lto_codegen_create();
 476         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 477                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", (*it)->path());
 478                 if ( ::lto_codegen_add_module(generator, (*it)->module()) )
 479                         throwf("lto: could not merge in %s because %s", (*it)->path(), ::lto_get_error_message());
 480         }
 481
 482         // add any -mllvm command line options
 483         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 484                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 485                 ::lto_codegen_debug_options(generator, *it);
 486         }
 487
 488         // The atom graph uses directed edges (references). Collect all references where
 489         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 490         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 491         // defined in llvm bitcode file.
 492         CStringSet nonLLVMRefs;
 493         CStringToAtom llvmAtoms;
 494     bool hasNonllvmAtoms = false;
 495         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 496                 const ld::Atom* atom = *it;
 497                 // only look at references that come from an atom that is not an llvm atom
 498                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 499                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 500                                 hasNonllvmAtoms = true;
 501                         }
 502                         const ld::Atom* target;
 503                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 504                                 switch ( fit->binding ) {
 505                                         case ld::Fixup::bindingDirectlyBound:
 506                                                 // that reference an llvm atom
 507                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 508                                                         nonLLVMRefs.insert(fit->u.target->name());
 509                                                 break;
 510                                         case ld::Fixup::bindingsIndirectlyBound:
 511                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 512                                                 if ( target == NULL )
 513                                                         throwf("'%s' in %s contains undefined reference", atom->name(), atom->file()->path());
 514                                                 assert(target != NULL);
 515                                                 if ( target->contentType() == ld::Atom::typeLTOtemporary )
 516                                                         nonLLVMRefs.insert(target->name());
 517                                         default:
 518                                                 break;
 519                                 }
 520                         }
 521                 }
 522                 else {
 523                         llvmAtoms[atom->name()] = (Atom*)atom;
 524                 }
 525         }
 526         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 527         if ( state.entryPoint!= NULL ) {
 528                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 529                         nonLLVMRefs.insert(state.entryPoint->name());
 530         }
 531
 532         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 533         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 534         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 535         // atom so that the linker can replace it with the mach-o one later.
 536         CStringToAtom deadllvmAtoms;
 537         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 538                 const ld::Atom* atom = *it;
 539                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 540                         const char* name = atom->name();
 541                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 542                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 543                         deadllvmAtoms[name] = (Atom*)atom;
 544                 }
 545         }
 546         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 547                 File* file = *it;
 548                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 549                         Atom* llvmAtom = &file->_atomArray[i];
 550                         if ( llvmAtom->coalescedAway()  ) {
 551                                 const char* name = llvmAtom->name();
 552                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 553                                         if ( logMustPreserve )
 554                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 555                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 556                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 557                                 }
 558                         }
 559                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 560                                 const char* name = llvmAtom->name();
 561                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 562                         }
 563                 }
 564         }
 565
 566         // tell code generator about symbols that must be preserved
 567         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 568                 const char* name = it->first;
 569                 Atom* atom = it->second;
 570                 // Include llvm Symbol in export list if it meets one of following two conditions
 571                 // 1 - atom scope is global (and not linkage unit).
 572                 // 2 - included in nonLLVMRefs set.
 573                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 574                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 575                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 576                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 577                 }
 578                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 579                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 580                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 581                 }
 582         }
 583
 584     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 585     if ( options.relocatable && !hasNonllvmAtoms ) {
 586                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 587                         // HACK, no good way to tell linker we are all done, so just quit
 588                         exit(0);
 589                 }
 590                 warning("could not produce merged bitcode file");
 591     }
 592
 593         // set code-gen model
 594         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 595         if ( options.mainExecutable ) {
 596                 if ( options.staticExecutable ) {
 597                         // darwin x86_64 "static" code model is really dynamic code model
 598                         if ( options.arch == CPU_TYPE_X86_64 )
 599                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 600                         else
 601                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 602                 }
 603                 else {
 604                         if ( options.pie )
 605                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 606                         else
 607                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 608                 }
 609         }
 610         else {
 611                 if ( options.allowTextRelocs )
 612                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 613                 else
 614                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 615         }
 616         if ( ::lto_codegen_set_pic_model(generator, model) )
 617                 throwf("could not create set codegen model: %s", lto_get_error_message());
 618
 619     // if requested, save off merged bitcode file
 620     if ( options.saveTemps ) {
 621         char tempBitcodePath[MAXPATHLEN];
 622         strcpy(tempBitcodePath, options.outputFilePath);
 623         strcat(tempBitcodePath, ".lto.bc");
 624         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 625     }
 626
 627 #if LTO_API_VERSION >= 3
 628         // find assembler next to linker
 629         char path[PATH_MAX];
 630         uint32_t bufSize = PATH_MAX;
 631         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 632                 char* lastSlash = strrchr(path, '/');
 633                 if ( lastSlash != NULL ) {
 634                         strcpy(lastSlash+1, "as");
 635                         struct stat statInfo;
 636                         if ( stat(path, &statInfo) == 0 )
 637                                 ::lto_codegen_set_assembler_path(generator, path);
 638                 }
 639         }
 640 #endif
 641         // run code generator
 642         size_t machOFileLen;
 643         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 644         if ( machOFile == NULL )
 645                 throwf("could not do LTO codegen: %s", ::lto_get_error_message());
 646
 647     // if requested, save off temp mach-o file
 648     if ( options.saveTemps ) {
 649         char tempMachoPath[MAXPATHLEN];
 650         strcpy(tempMachoPath, options.outputFilePath);
 651         strcat(tempMachoPath, ".lto.o");
 652         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 653                 if ( fd != -1) {
 654                         ::write(fd, machOFile, machOFileLen);
 655                         ::close(fd);
 656                 }
 657                 //      save off merged bitcode file
 658                 char tempOptBitcodePath[MAXPATHLEN];
 659         strcpy(tempOptBitcodePath, options.outputFilePath);
 660         strcat(tempOptBitcodePath, ".lto.opt.bc");
 661         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 662         }
 663
 664         // if needed, save temp mach-o file to specific location
 665         if ( options.tmpObjectFilePath != NULL ) {
 666                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 667                 if ( fd != -1) {
 668                         ::write(fd, machOFile, machOFileLen);
 669                         ::close(fd);
 670                 }
 671                 else {
 672                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 673                 }
 674         }
 675
 676         // parse generated mach-o file into a MachOReader
 677         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 678
 679         // sync generated mach-o atoms with existing atoms ld knows about
 680         if ( logAtomsBeforeSync ) {
 681                 fprintf(stderr, "llvmAtoms:\n");
 682                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 683                         const char* name = it->first;
 684                         //Atom* atom = it->second;
 685                         fprintf(stderr, "\t%s\n", name);
 686                 }
 687                 fprintf(stderr, "deadllvmAtoms:\n");
 688                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 689                         const char* name = it->first;
 690                         //Atom* atom = it->second;
 691                         fprintf(stderr, "\t%s\n", name);
 692                 }
 693         }
 694         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 695         machoFile->forEachAtom(syncer);
 696
 697         // Remove InternalAtoms from ld
 698         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 699                 (*it)->internalAtom().setCoalescedAway();
 700         }
 701         // Remove Atoms from ld if code generator optimized them away
 702         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 703                 // check if setRealAtom() called on this Atom
 704                 if ( li->second->compiledAtom() == NULL ) {
 705                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 706                         li->second->setCoalescedAway();
 707                 }
 708         }
 709
 710         // notify about file level attributes
 711         handler.doFile(*machoFile);
 712
 713         // if final mach-o file has debug info, update original bitcode files to match
 714         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 715                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 716                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 717         }
 718
 719         return true;
 720 }
 721
 722
 723 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 724 {
 725         // update proxy atoms to point to real atoms and find new atoms
 726         const char* name = machoAtom.name();
 727         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 728                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 729                 if ( pos != _llvmAtoms.end() ) {
 730                         // turn Atom into a proxy for this mach-o atom
 731                         pos->second->setCompiledAtom(machoAtom);
 732                 }
 733                 else {
 734                         // an atom of this name was not in the allAtoms list the linker gave us
 735                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 736                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 737                                 if ( _options.linkerDeadStripping ) {
 738                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 739                                         Atom* llvmAtom = _deadllvmAtoms[name];
 740                                         llvmAtom->setCompiledAtom(machoAtom);
 741                                         _newAtoms.push_back(&machoAtom);
 742                                 }
 743                                 else {
 744                                         // Don't pass it back as a new atom
 745                                 }
 746                         }
 747                         else
 748                         {
 749                                 // this is something new that lto conjured up, tell ld its new
 750                                 _newAtoms.push_back(&machoAtom);
 751                         }
 752                 }
 753         }
 754         else {
 755                 // ld only knew about non-static atoms, so this one must be new
 756                 _newAtoms.push_back(&machoAtom);
 757         }
 758
 759         // adjust fixups to go through proxy atoms
 760         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 761         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 762                 switch ( fit->binding ) {
 763                         case ld::Fixup::bindingNone:
 764                                 break;
 765                         case ld::Fixup::bindingByNameUnbound:
 766                                 // don't know if this target has been seen by linker before or if it is new
 767                                 // be conservative and tell linker it is new
 768                                 _additionalUndefines.push_back(fit->u.name);
 769                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 770                                 break;
 771                         case ld::Fixup::bindingDirectlyBound:
 772                                 // If mach-o atom is referencing another mach-o atom then
 773                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 774                                 // llvm symbol references always go through Atom proxy.
 775                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 776                                         const char* targetName = fit->u.target->name();
 777                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 778                                         if ( pos != _llvmAtoms.end() ) {
 779                                                 fit->u.target = pos->second;
 780                                         }
 781                                         else {
 782                                                 if ( _deadllvmAtoms.find(targetName) != _deadllvmAtoms.end() ) {
 783                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 784                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 785                                                         fit->u.name = targetName;
 786                                                 }
 787                                         }
 788                                 }
 789                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 790                                 break;
 791                         case ld::Fixup::bindingByContentBound:
 792                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 793                                 break;
 794                         case ld::Fixup::bindingsIndirectlyBound:
 795                                 assert(0 && "indirect binding found in initial mach-o file?");
 796                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 797                                 break;
 798                 }
 799         }
 800
 801 }
 802
 803 class Mutex {
 804         static pthread_mutex_t lto_lock;
 805 public:
 806         Mutex() { pthread_mutex_lock(&lto_lock); }
 807         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 808 };
 809 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 810
 811 //
 812 // Used by archive reader to see if member is an llvm bitcode file
 813 //
 814 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 815 {
 816         Mutex lock;
 817         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 818 }
 819
 820
 821 //
 822 // main function used by linker to instantiate ld::Files
 823 //
 824 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 825                                                                 const char* path, time_t modTime, ld::File::Ordinal ordinal,
 826                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 827 {
 828         Mutex lock;
 829         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 830                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles);
 831         else
 832                 return NULL;
 833 }
 834
 835 //
 836 // used by "ld -v" to report version of libLTO.dylib being used
 837 //
 838 const char* version()
 839 {
 840         Mutex lock;
 841         return ::lto_get_version();
 842 }
 843
 844
 845 //
 846 // used by ld for error reporting
 847 //
 848 bool libLTOisLoaded()
 849 {
 850         Mutex lock;
 851         return (::lto_get_version() != NULL);
 852 }
 853
 854 //
 855 // used by ld for error reporting
 856 //
 857 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 858 {
 859         Mutex lock;
 860         return Parser::fileKind(fileContent, fileLength);
 861 }
 862
 863 //
 864 // used by ld for doing link time optimization
 865 //
 866 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 867                                 ld::Internal&                                           state,
 868                                 const OptimizeOptions&                          options,
 869                                 ld::File::AtomHandler&                          handler,
 870                                 std::vector<const ld::Atom*>&           newAtoms,
 871                                 std::vector<const char*>&                       additionalUndefines)
 872 {
 873         Mutex lock;
 874         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 875 }
 876
 877
 878
 879 }; // namespace lto
 880
 881
 882 #endif
 883