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 bool                                                            translationUnitSource(const char** dir, const char** nm) const
  67                                                                                                                                         { return false; }
  68         virtual const char*                                                     name() const            { return "import-atom"; }
  69         virtual uint64_t                                                        size() const            { return 0; }
  70         virtual uint64_t                                                        objectAddress() const { return 0; }
  71         virtual void                                                            copyRawContent(uint8_t buffer[]) const { }
  72         virtual void                                                            setScope(Scope)         { }
  73         virtual ld::Fixup::iterator                                     fixupsBegin() const     { return &_undefs[0]; }
  74         virtual ld::Fixup::iterator                                     fixupsEnd()     const   { return &_undefs[_undefs.size()]; }
  75
  76         // for adding references to symbols outside bitcode file
  77         void                                                                            addReference(const char* nm)
  78                                                                                                                                         { _undefs.push_back(ld::Fixup(0, ld::Fixup::k1of1,
  79                                                                                                                                                                 ld::Fixup::kindNone, false, nm)); }
  80 private:
  81
  82         ld::File&                                                                       _file;
  83         mutable std::vector<ld::Fixup>                          _undefs;
  84 };
  85
  86
  87 //
  88 // LLVM bitcode file
  89 //
  90 class File : public ld::relocatable::File
  91 {
  92 public:
  93                                                                                         File(const char* path, time_t mTime, const uint8_t* content,
  94                                                                                                         uint32_t contentLength, uint32_t ordinal, cpu_type_t arch);
  95         virtual                                                                 ~File();
  96
  97         // overrides of ld::File
  98         virtual bool                                                                            forEachAtom(ld::File::AtomHandler&) const;
  99         virtual bool                                                                            justInTimeforEachAtom(const char* name, ld::File::AtomHandler&) const
 100                                                                                                                                                                         { return false; }
 101         virtual uint32_t                                                                        cpuSubType() const                      { return _cpuSubType; }
 102
 103         // overrides of ld::relocatable::File
 104         virtual bool                                                                            objcReplacementClasses() const  { return false; }
 105         virtual DebugInfoKind                                                           debugInfo()     const                   { return _debugInfo; }
 106         virtual const char*                                                                     debugInfoPath() const           { return _debugInfoPath; }
 107         virtual time_t                                                                          debugInfoModificationTime() const
 108                                                                                                                                                                         { return _debugInfoModTime; }
 109         virtual const std::vector<ld::relocatable::File::Stab>* stabs() const                   { return NULL; }
 110         virtual bool                                                                            canScatterAtoms() const         { return true; }
 111
 112         lto_module_t                                                                            module()                                        { return _module; }
 113         class InternalAtom&                                                                     internalAtom()                          { return _internalAtom; }
 114         void                                                                                            setDebugInfo(ld::relocatable::File::DebugInfoKind k,
 115                                                                                                                                         const char* pth, time_t modTime, uint32_t subtype)
 116                                                                                                                                                                         {       _debugInfo = k;
 117                                                                                                                                                                                 _debugInfoPath = pth;
 118                                                                                                                                                                                 _debugInfoModTime = modTime;
 119                                                                                                                                                                                 _cpuSubType = subtype;}
 120
 121 private:
 122         friend class Atom;
 123         friend class InternalAtom;
 124         friend class Parser;
 125
 126         cpu_type_t                                                              _architecture;
 127         class InternalAtom                                              _internalAtom;
 128         class Atom*                                                             _atomArray;
 129         uint32_t                                                                _atomArrayCount;
 130         lto_module_t                                                    _module;
 131         const char*                                                             _debugInfoPath;
 132         time_t                                                                  _debugInfoModTime;
 133         ld::Section                                                             _section;
 134         ld::Fixup                                                               _fixupToInternal;
 135         ld::relocatable::File::DebugInfoKind    _debugInfo;
 136         uint32_t                                                                _cpuSubType;
 137 };
 138
 139 //
 140 // Atom acts as a proxy Atom for the symbols that are exported by LLVM bitcode file. Initially,
 141 // Reader creates Atoms to allow linker proceed with usual symbol resolution phase. After
 142 // optimization is performed, real Atoms are created for these symobls. However these real Atoms
 143 // are not inserted into global symbol table. Atom holds real Atom and forwards appropriate
 144 // methods to real atom.
 145 //
 146 class Atom : public ld::Atom
 147 {
 148 public:
 149                                                                                 Atom(File& f, const char* name, ld::Atom::Scope s,
 150                                                                                                 ld::Atom::Definition d, ld::Atom::Combine c, ld::Atom::Alignment a, bool ah);
 151
 152         // overrides of ld::Atom
 153         virtual ld::File*                                       file() const            { return &_file; }
 154         virtual bool                                            translationUnitSource(const char** dir, const char** nm) const
 155                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource(dir, nm) : false); }
 156         virtual const char*                                     name() const            { return _name; }
 157         virtual uint64_t                                        size() const            { return (_compiledAtom ? _compiledAtom->size() : 0); }
 158         virtual uint64_t                                        objectAddress() const { return (_compiledAtom ? _compiledAtom->objectAddress() : 0); }
 159         virtual void                                            copyRawContent(uint8_t buffer[]) const
 160                                                                                                                         { if (_compiledAtom) _compiledAtom->copyRawContent(buffer); }
 161         virtual const uint8_t*                          rawContentPointer() const
 162                                                                                                                         { return (_compiledAtom ? _compiledAtom->rawContentPointer() : NULL);  }
 163         virtual unsigned long                           contentHash(const class ld::IndirectBindingTable& ibt) const
 164                                                                                                                         { return (_compiledAtom ? _compiledAtom->contentHash(ibt) : 0);  }
 165         virtual bool                                            canCoalesceWith(const ld::Atom& rhs, const class ld::IndirectBindingTable& ibt) const
 166                                                                                                                         { return (_compiledAtom ? _compiledAtom->canCoalesceWith(rhs,ibt) : false); }
 167         virtual ld::Fixup::iterator                             fixupsBegin() const
 168                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsBegin() : (ld::Fixup*)&_file._fixupToInternal); }
 169         virtual ld::Fixup::iterator                             fixupsEnd() const
 170                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsEnd() : &((ld::Fixup*)&_file._fixupToInternal)[1]); }
 171         virtual ld::Atom::UnwindInfo::iterator  beginUnwind() const
 172                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginUnwind() : NULL); }
 173         virtual ld::Atom::UnwindInfo::iterator  endUnwind() const
 174                                                                                                                         { return (_compiledAtom ? _compiledAtom->endUnwind() : NULL); }
 175         virtual ld::Atom::LineInfo::iterator    beginLineInfo() const
 176                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginLineInfo() : NULL); }
 177         virtual ld::Atom::LineInfo::iterator    endLineInfo() const
 178                                                                                                                         { return (_compiledAtom ? _compiledAtom->endLineInfo() : NULL); }
 179
 180         const ld::Atom*                                         compiledAtom()          { return _compiledAtom; }
 181         void                                                            setCompiledAtom(const ld::Atom& atom);
 182
 183 private:
 184
 185         File&                                                           _file;
 186         const char*                                                     _name;
 187         const ld::Atom*                                         _compiledAtom;
 188 };
 189
 190
 191
 192
 193
 194
 195
 196 class Parser
 197 {
 198 public:
 199         static bool                                             validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch);
 200         static const char*                              fileKind(const uint8_t* fileContent, uint64_t fileLength);
 201         static File*                                    parse(const uint8_t* fileContent, uint64_t fileLength, const char* path,
 202                                                                                         time_t modTime, uint32_t ordinal, cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles);
 203         static bool                                             libLTOisLoaded() { return (::lto_get_version() != NULL); }
 204         static bool                                             optimize(   const std::vector<const ld::Atom*>& allAtoms,
 205                                                                                                 ld::Internal&                                           state,
 206                                                                                                 uint32_t                                                        nextInputOrdinal,
 207                                                                                                 const OptimizeOptions&                          options,
 208                                                                                                 ld::File::AtomHandler&                          handler,
 209                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 210                                                                                                 std::vector<const char*>&                       additionalUndefines);
 211
 212         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 213
 214 private:
 215         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 216         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, uint32_t nextInputOrdinal, const OptimizeOptions& options);
 217
 218         class CStringEquals
 219         {
 220         public:
 221                 bool operator()(const char* left, const char* right) const { return (strcmp(left, right) == 0); }
 222         };
 223         typedef __gnu_cxx::hash_set<const char*, __gnu_cxx::hash<const char*>, CStringEquals>  CStringSet;
 224         typedef __gnu_cxx::hash_map<const char*, Atom*, __gnu_cxx::hash<const char*>, CStringEquals> CStringToAtom;
 225
 226         class AtomSyncer : public ld::File::AtomHandler {
 227         public:
 228                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 229                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 230                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 231                 virtual void            doAtom(const class ld::Atom&);
 232                 virtual void            doFile(const class ld::File&) { }
 233
 234                 const OptimizeOptions&                  _options;
 235                 std::vector<const char*>&               _additionalUndefines;
 236                 std::vector<const ld::Atom*>&   _newAtoms;
 237                 CStringToAtom                                   _llvmAtoms;
 238                 CStringToAtom                                   _deadllvmAtoms;
 239         };
 240
 241         static std::vector<File*>               _s_files;
 242 };
 243
 244 std::vector<File*> Parser::_s_files;
 245
 246
 247 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 248 {
 249         switch (architecture) {
 250                 case CPU_TYPE_I386:
 251                         return ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, "i386-");
 252                 case CPU_TYPE_X86_64:
 253                         return ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, "x86_64-");
 254                 case CPU_TYPE_ARM:
 255                         for (const ARMSubType* t=ARMSubTypes; t->subTypeName != NULL; ++t) {
 256                                 if ( subarch == t->subType )
 257                                         return ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 258                         }
 259                         break;
 260         }
 261         return false;
 262 }
 263
 264 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 265 {
 266         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 267                 uint32_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 268                 switch (arch) {
 269                         case CPU_TYPE_I386:
 270                                 return "i386";
 271                         case CPU_TYPE_X86_64:
 272                                 return "x86_64";
 273                         case CPU_TYPE_ARM:
 274                                 for (const ARMSubType* t=ARMSubTypes; t->subTypeName != NULL; ++t) {
 275                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 276                                                 return t->subTypeName;
 277                                 }
 278                                 return "arm";
 279                 }
 280                 return "unknown bitcode architecture";
 281         }
 282         return NULL;
 283 }
 284
 285 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime,
 286                                                                                                         uint32_t ordinal, cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 287 {
 288         File* f = new File(path, modTime, fileContent, fileLength, ordinal, architecture);
 289         _s_files.push_back(f);
 290         if ( logAllFiles )
 291                 printf("%s\n", path);
 292         return f;
 293 }
 294
 295
 296 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, uint32_t nextInputOrdinal, const OptimizeOptions& options)
 297 {
 298         mach_o::relocatable::ParserOptions objOpts;
 299         objOpts.architecture            = options.arch;
 300         objOpts.objSubtypeMustMatch = false;
 301         objOpts.logAllFiles                     = false;
 302         objOpts.convertUnwindInfo       = true;
 303         objOpts.subType                         = 0;
 304
 305         // mach-o parsing is done in-memory, but need path for debug notes
 306         const char* path = "/tmp/lto.o";
 307         time_t modTime = 0;
 308         if ( options.tmpObjectFilePath != NULL ) {
 309                 path = options.tmpObjectFilePath;
 310                 struct stat statBuffer;
 311                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 312                         modTime = statBuffer.st_mtime;
 313         }
 314
 315         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, nextInputOrdinal, objOpts);
 316         if ( result != NULL )
 317                 return result;
 318         throw "LLVM LTO, file is not of required architecture";
 319 }
 320
 321
 322
 323 File::File(const char* pth, time_t mTime, const uint8_t* content, uint32_t contentLength, uint32_t ord, cpu_type_t arch)
 324         : ld::relocatable::File(pth,mTime,ord), _architecture(arch), _internalAtom(*this),
 325         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 326         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 327         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 328         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 329 {
 330         const bool log = false;
 331
 332         // create llvm module
 333         _module = ::lto_module_create_from_memory(content, contentLength);
 334     if ( _module == NULL )
 335                 throwf("could not parse object file %s: %s", pth, lto_get_error_message());
 336
 337         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 338
 339         // create atom for each global symbol in module
 340         uint32_t count = ::lto_module_get_num_symbols(_module);
 341         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 342         for (uint32_t i=0; i < count; ++i) {
 343                 const char* name = ::lto_module_get_symbol_name(_module, i);
 344                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 345
 346                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 347                 // ignore dtrace static probes for now
 348                 // later when codegen is done and a mach-o file is produces the probes will be processed
 349                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 350                         continue;
 351
 352                 ld::Atom::Definition def;
 353                 ld::Atom::Combine combine = ld::Atom::combineNever;
 354                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 355                         case LTO_SYMBOL_DEFINITION_REGULAR:
 356                                 def = ld::Atom::definitionRegular;
 357                                 break;
 358                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 359                                 def = ld::Atom::definitionTentative;
 360                                 break;
 361                         case LTO_SYMBOL_DEFINITION_WEAK:
 362                                 def = ld::Atom::definitionRegular;
 363                                 combine = ld::Atom::combineByName;
 364                                 break;
 365                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 366                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 367                                 def = ld::Atom::definitionProxy;
 368                                 break;
 369                         default:
 370                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 371                 }
 372
 373                 // make LLVM atoms for definitions and a reference for undefines
 374                 if ( def != ld::Atom::definitionProxy ) {
 375                         ld::Atom::Scope scope;
 376                         bool autohide = false;
 377                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 378                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 379                                         scope = ld::Atom::scopeTranslationUnit;
 380                                         break;
 381                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 382                                         scope = ld::Atom::scopeLinkageUnit;
 383                                         break;
 384                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 385                                         scope = ld::Atom::scopeGlobal;
 386                                         break;
 387 #if LTO_API_VERSION >= 4
 388                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 389                                         scope = ld::Atom::scopeGlobal;
 390                                         autohide = true;
 391                                         break;
 392 #endif
 393                                 default:
 394                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 395                         }
 396                         // only make atoms for non-internal symbols
 397                         if ( scope == ld::Atom::scopeTranslationUnit )
 398                                 continue;
 399                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 400                         // make Atom using placement new operator
 401                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 402                         if ( scope == ld::Atom::scopeLinkageUnit )
 403                                 _internalAtom.addReference(name);
 404                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 405                 }
 406                 else {
 407                         // add to list of external references
 408                         _internalAtom.addReference(name);
 409                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 410                 }
 411         }
 412 }
 413
 414 File::~File()
 415 {
 416         if ( _module != NULL )
 417                 ::lto_module_dispose(_module);
 418 }
 419
 420 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 421 {
 422         handler.doAtom(_internalAtom);
 423         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 424                 handler.doAtom(_atomArray[i]);
 425         }
 426         return true;
 427 }
 428
 429 InternalAtom::InternalAtom(File& f)
 430         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 431                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 432                 _file(f)
 433 {
 434 }
 435
 436 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 437                         ld::Atom::Alignment a, bool ah)
 438         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 439                                 ld::Atom::symbolTableIn, false, false, false, a),
 440                 _file(f), _name(nm), _compiledAtom(NULL)
 441 {
 442         if ( ah )
 443                 this->setAutoHide();
 444 }
 445
 446 void Atom::setCompiledAtom(const ld::Atom& atom)
 447 {
 448         // set delegate so virtual methods go to it
 449         _compiledAtom = &atom;
 450
 451         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 452
 453         // update fields in ld::Atom to match newly constructed mach-o atom
 454         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 455 }
 456
 457
 458
 459 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 460                                                 ld::Internal&                                           state,
 461                                                 uint32_t                                                        nextInputOrdinal,
 462                                                 const OptimizeOptions&                          options,
 463                                                 ld::File::AtomHandler&                          handler,
 464                                                 std::vector<const ld::Atom*>&           newAtoms,
 465                                                 std::vector<const char*>&                       additionalUndefines)
 466 {
 467         const bool logMustPreserve = false;
 468         const bool logExtraOptions = false;
 469         const bool logBitcodeFiles = false;
 470         const bool logAtomsBeforeSync = false;
 471
 472         // exit quickly if nothing to do
 473         if ( _s_files.size() == 0 )
 474                 return false;
 475
 476         // print out LTO version string if -v was used
 477         if ( options.verbose )
 478                 fprintf(stderr, "%s\n", lto_get_version());
 479
 480         // create optimizer and add each Reader
 481         lto_code_gen_t generator = ::lto_codegen_create();
 482         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 483                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", (*it)->path());
 484                 if ( ::lto_codegen_add_module(generator, (*it)->module()) )
 485                         throwf("lto: could not merge in %s because %s", (*it)->path(), ::lto_get_error_message());
 486         }
 487
 488         // add any -mllvm command line options
 489         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 490                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 491                 ::lto_codegen_debug_options(generator, *it);
 492         }
 493
 494         // The atom graph uses directed edges (references). Collect all references where
 495         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 496         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 497         // defined in llvm bitcode file.
 498         CStringSet nonLLVMRefs;
 499         CStringToAtom llvmAtoms;
 500     bool hasNonllvmAtoms = false;
 501         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 502                 const ld::Atom* atom = *it;
 503                 // only look at references that come from an atom that is not an llvm atom
 504                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 505                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 506                                 hasNonllvmAtoms = true;
 507                         }
 508                         const ld::Atom* target;
 509                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 510                                 switch ( fit->binding ) {
 511                                         case ld::Fixup::bindingDirectlyBound:
 512                                                 // that reference an llvm atom
 513                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 514                                                         nonLLVMRefs.insert(fit->u.target->name());
 515                                                 break;
 516                                         case ld::Fixup::bindingsIndirectlyBound:
 517                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 518                                                 if ( target == NULL )
 519                                                         throwf("'%s' in %s contains undefined reference", atom->name(), atom->file()->path());
 520                                                 assert(target != NULL);
 521                                                 if ( target->contentType() == ld::Atom::typeLTOtemporary )
 522                                                         nonLLVMRefs.insert(target->name());
 523                                         default:
 524                                                 break;
 525                                 }
 526                         }
 527                 }
 528                 else {
 529                         llvmAtoms[atom->name()] = (Atom*)atom;
 530                 }
 531         }
 532         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 533         if ( state.entryPoint!= NULL ) {
 534                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 535                         nonLLVMRefs.insert(state.entryPoint->name());
 536         }
 537
 538         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 539         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 540         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 541         // atom so that the linker can replace it with the mach-o one later.
 542         CStringToAtom deadllvmAtoms;
 543         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 544                 const ld::Atom* atom = *it;
 545                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 546                         const char* name = atom->name();
 547                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 548                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 549                         deadllvmAtoms[name] = (Atom*)atom;
 550                 }
 551         }
 552         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 553                 File* file = *it;
 554                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 555                         Atom* llvmAtom = &file->_atomArray[i];
 556                         if ( llvmAtom->coalescedAway()  ) {
 557                                 const char* name = llvmAtom->name();
 558                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 559                                         if ( logMustPreserve )
 560                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 561                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 562                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 563                                 }
 564                         }
 565                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 566                                 const char* name = llvmAtom->name();
 567                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 568                         }
 569                 }
 570         }
 571
 572         // tell code generator about symbols that must be preserved
 573         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 574                 const char* name = it->first;
 575                 Atom* atom = it->second;
 576                 // Include llvm Symbol in export list if it meets one of following two conditions
 577                 // 1 - atom scope is global (and not linkage unit).
 578                 // 2 - included in nonLLVMRefs set.
 579                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 580                 if ( (atom->scope() == ld::Atom::scopeGlobal) ) {
 581                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 582                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 583                 }
 584                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 585                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 586                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 587                 }
 588         }
 589
 590     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 591     if ( options.relocatable && !hasNonllvmAtoms ) {
 592                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 593                         // HACK, no good way to tell linker we are all done, so just quit
 594                         exit(0);
 595                 }
 596                 warning("could not produce merged bitcode file");
 597     }
 598
 599         // set code-gen model
 600         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 601         if ( options.mainExecutable ) {
 602                 if ( options.staticExecutable ) {
 603                         // darwin x86_64 "static" code model is really dynamic code model
 604                         if ( options.arch == CPU_TYPE_X86_64 )
 605                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 606                         else
 607                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 608                 }
 609                 else {
 610                         if ( options.pie )
 611                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 612                         else
 613                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 614                 }
 615         }
 616         else {
 617                 if ( options.allowTextRelocs )
 618                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 619                 else
 620                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 621         }
 622         if ( ::lto_codegen_set_pic_model(generator, model) )
 623                 throwf("could not create set codegen model: %s", lto_get_error_message());
 624
 625     // if requested, save off merged bitcode file
 626     if ( options.saveTemps ) {
 627         char tempBitcodePath[MAXPATHLEN];
 628         strcpy(tempBitcodePath, options.outputFilePath);
 629         strcat(tempBitcodePath, ".lto.bc");
 630         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 631     }
 632
 633 #if LTO_API_VERSION >= 3
 634         // find assembler next to linker
 635         char path[PATH_MAX];
 636         uint32_t bufSize = PATH_MAX;
 637         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 638                 char* lastSlash = strrchr(path, '/');
 639                 if ( lastSlash != NULL ) {
 640                         strcpy(lastSlash+1, "as");
 641                         struct stat statInfo;
 642                         if ( stat(path, &statInfo) == 0 )
 643                                 ::lto_codegen_set_assembler_path(generator, path);
 644                 }
 645         }
 646 #endif
 647         // run code generator
 648         size_t machOFileLen;
 649         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 650         if ( machOFile == NULL )
 651                 throwf("could not do LTO codegen: %s", ::lto_get_error_message());
 652
 653     // if requested, save off temp mach-o file
 654     if ( options.saveTemps ) {
 655         char tempMachoPath[MAXPATHLEN];
 656         strcpy(tempMachoPath, options.outputFilePath);
 657         strcat(tempMachoPath, ".lto.o");
 658         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 659                 if ( fd != -1) {
 660                         ::write(fd, machOFile, machOFileLen);
 661                         ::close(fd);
 662                 }
 663                 //      save off merged bitcode file
 664                 char tempOptBitcodePath[MAXPATHLEN];
 665         strcpy(tempOptBitcodePath, options.outputFilePath);
 666         strcat(tempOptBitcodePath, ".lto.opt.bc");
 667         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 668         }
 669
 670         // if needed, save temp mach-o file to specific location
 671         if ( options.tmpObjectFilePath != NULL ) {
 672                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 673                 if ( fd != -1) {
 674                         ::write(fd, machOFile, machOFileLen);
 675                         ::close(fd);
 676                 }
 677                 else {
 678                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 679                 }
 680         }
 681
 682         // parse generated mach-o file into a MachOReader
 683         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, nextInputOrdinal, options);
 684
 685         // sync generated mach-o atoms with existing atoms ld knows about
 686         if ( logAtomsBeforeSync ) {
 687                 fprintf(stderr, "llvmAtoms:\n");
 688                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 689                         const char* name = it->first;
 690                         //Atom* atom = it->second;
 691                         fprintf(stderr, "\t%s\n", name);
 692                 }
 693                 fprintf(stderr, "deadllvmAtoms:\n");
 694                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 695                         const char* name = it->first;
 696                         //Atom* atom = it->second;
 697                         fprintf(stderr, "\t%s\n", name);
 698                 }
 699         }
 700         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 701         machoFile->forEachAtom(syncer);
 702
 703         // Remove InternalAtoms from ld
 704         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 705                 (*it)->internalAtom().setCoalescedAway();
 706         }
 707         // Remove Atoms from ld if code generator optimized them away
 708         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 709                 // check if setRealAtom() called on this Atom
 710                 if ( li->second->compiledAtom() == NULL ) {
 711                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 712                         li->second->setCoalescedAway();
 713                 }
 714         }
 715
 716         // notify about file level attributes
 717         handler.doFile(*machoFile);
 718
 719         // if final mach-o file has debug info, update original bitcode files to match
 720         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 721                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 722                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 723         }
 724
 725         return true;
 726 }
 727
 728
 729 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 730 {
 731         // update proxy atoms to point to real atoms and find new atoms
 732         const char* name = machoAtom.name();
 733         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 734                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 735                 if ( pos != _llvmAtoms.end() ) {
 736                         // turn Atom into a proxy for this mach-o atom
 737                         pos->second->setCompiledAtom(machoAtom);
 738                 }
 739                 else {
 740                         // an atom of this name was not in the allAtoms list the linker gave us
 741                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 742                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 743                                 if ( _options.linkerDeadStripping ) {
 744                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 745                                         Atom* llvmAtom = _deadllvmAtoms[name];
 746                                         llvmAtom->setCompiledAtom(machoAtom);
 747                                         _newAtoms.push_back(&machoAtom);
 748                                 }
 749                                 else {
 750                                         // Don't pass it back as a new atom
 751                                 }
 752                         }
 753                         else
 754                         {
 755                                 // this is something new that lto conjured up, tell ld its new
 756                                 _newAtoms.push_back(&machoAtom);
 757                         }
 758                 }
 759         }
 760         else {
 761                 // ld only knew about non-static atoms, so this one must be new
 762                 _newAtoms.push_back(&machoAtom);
 763         }
 764
 765         // adjust fixups to go through proxy atoms
 766         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 767         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 768                 switch ( fit->binding ) {
 769                         case ld::Fixup::bindingNone:
 770                                 break;
 771                         case ld::Fixup::bindingByNameUnbound:
 772                                 // don't know if this target has been seen by linker before or if it is new
 773                                 // be conservative and tell linker it is new
 774                                 _additionalUndefines.push_back(fit->u.name);
 775                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 776                                 break;
 777                         case ld::Fixup::bindingDirectlyBound:
 778                                 // If mach-o atom is referencing another mach-o atom then
 779                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 780                                 // llvm symbol references always go through Atom proxy.
 781                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 782                                         const char* targetName = fit->u.target->name();
 783                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 784                                         if ( pos != _llvmAtoms.end() ) {
 785                                                 fit->u.target = pos->second;
 786                                         }
 787                                         else {
 788                                                 if ( _deadllvmAtoms.find(targetName) != _deadllvmAtoms.end() ) {
 789                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 790                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 791                                                         fit->u.name = targetName;
 792                                                 }
 793                                         }
 794                                 }
 795                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 796                                 break;
 797                         case ld::Fixup::bindingByContentBound:
 798                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 799                                 break;
 800                         case ld::Fixup::bindingsIndirectlyBound:
 801                                 assert(0 && "indirect binding found in initial mach-o file?");
 802                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 803                                 break;
 804                 }
 805         }
 806
 807 }
 808
 809
 810 //
 811 // Used by archive reader to see if member is an llvm bitcode file
 812 //
 813 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 814 {
 815         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 816 }
 817
 818
 819 //
 820 // main function used by linker to instantiate ld::Files
 821 //
 822 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 823                                                                 const char* path, time_t modTime, uint32_t ordinal,
 824                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 825 {
 826         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 827                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles);
 828         else
 829                 return NULL;
 830 }
 831
 832 //
 833 // used by "ld -v" to report version of libLTO.dylib being used
 834 //
 835 const char* version()
 836 {
 837         return ::lto_get_version();
 838 }
 839
 840
 841 //
 842 // used by ld for error reporting
 843 //
 844 bool libLTOisLoaded()
 845 {
 846         return (::lto_get_version() != NULL);
 847 }
 848
 849 //
 850 // used by ld for error reporting
 851 //
 852 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 853 {
 854         return Parser::fileKind(fileContent, fileLength);
 855 }
 856
 857 //
 858 // used by ld for doing link time optimization
 859 //
 860 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 861                                 ld::Internal&                                           state,
 862                                 uint32_t                                                        nextInputOrdinal,
 863                                 const OptimizeOptions&                          options,
 864                                 ld::File::AtomHandler&                          handler,
 865                                 std::vector<const ld::Atom*>&           newAtoms,
 866                                 std::vector<const char*>&                       additionalUndefines)
 867 {
 868         return Parser::optimize(allAtoms, state, nextInputOrdinal, options, handler, newAtoms, additionalUndefines);
 869 }
 870
 871
 872
 873 }; // namespace lto
 874
 875
 876 #endif
 877