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, 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 DebugInfoKind                                                           debugInfo()     const                   { return _debugInfo; }
 105         virtual const char*                                                                     debugInfoPath() const           { return _debugInfoPath; }
 106         virtual time_t                                                                          debugInfoModificationTime() const
 107                                                                                                                                                                         { return _debugInfoModTime; }
 108         virtual const std::vector<ld::relocatable::File::Stab>* stabs() const                   { return NULL; }
 109         virtual bool                                                                            canScatterAtoms() const         { return true; }
 110
 111         lto_module_t                                                                            module()                                        { return _module; }
 112         class InternalAtom&                                                                     internalAtom()                          { return _internalAtom; }
 113         void                                                                                            setDebugInfo(ld::relocatable::File::DebugInfoKind k,
 114                                                                                                                                         const char* pth, time_t modTime, uint32_t subtype)
 115                                                                                                                                                                         {       _debugInfo = k;
 116                                                                                                                                                                                 _debugInfoPath = pth;
 117                                                                                                                                                                                 _debugInfoModTime = modTime;
 118                                                                                                                                                                                 _cpuSubType = subtype;}
 119
 120 private:
 121         friend class Atom;
 122         friend class InternalAtom;
 123         friend class Parser;
 124
 125         cpu_type_t                                                              _architecture;
 126         class InternalAtom                                              _internalAtom;
 127         class Atom*                                                             _atomArray;
 128         uint32_t                                                                _atomArrayCount;
 129         lto_module_t                                                    _module;
 130         const char*                                                             _debugInfoPath;
 131         time_t                                                                  _debugInfoModTime;
 132         ld::Section                                                             _section;
 133         ld::Fixup                                                               _fixupToInternal;
 134         ld::relocatable::File::DebugInfoKind    _debugInfo;
 135         uint32_t                                                                _cpuSubType;
 136 };
 137
 138 //
 139 // Atom acts as a proxy Atom for the symbols that are exported by LLVM bitcode file. Initially,
 140 // Reader creates Atoms to allow linker proceed with usual symbol resolution phase. After
 141 // optimization is performed, real Atoms are created for these symobls. However these real Atoms
 142 // are not inserted into global symbol table. Atom holds real Atom and forwards appropriate
 143 // methods to real atom.
 144 //
 145 class Atom : public ld::Atom
 146 {
 147 public:
 148                                                                                 Atom(File& f, const char* name, ld::Atom::Scope s,
 149                                                                                                 ld::Atom::Definition d, ld::Atom::Combine c, ld::Atom::Alignment a, bool ah);
 150
 151         // overrides of ld::Atom
 152         virtual ld::File*                                       file() const            { return &_file; }
 153         virtual bool                                            translationUnitSource(const char** dir, const char** nm) const
 154                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource(dir, nm) : false); }
 155         virtual const char*                                     name() const            { return _name; }
 156         virtual uint64_t                                        size() const            { return (_compiledAtom ? _compiledAtom->size() : 0); }
 157         virtual uint64_t                                        objectAddress() const { return (_compiledAtom ? _compiledAtom->objectAddress() : 0); }
 158         virtual void                                            copyRawContent(uint8_t buffer[]) const
 159                                                                                                                         { if (_compiledAtom) _compiledAtom->copyRawContent(buffer); }
 160         virtual const uint8_t*                          rawContentPointer() const
 161                                                                                                                         { return (_compiledAtom ? _compiledAtom->rawContentPointer() : NULL);  }
 162         virtual unsigned long                           contentHash(const class ld::IndirectBindingTable& ibt) const
 163                                                                                                                         { return (_compiledAtom ? _compiledAtom->contentHash(ibt) : 0);  }
 164         virtual bool                                            canCoalesceWith(const ld::Atom& rhs, const class ld::IndirectBindingTable& ibt) const
 165                                                                                                                         { return (_compiledAtom ? _compiledAtom->canCoalesceWith(rhs,ibt) : false); }
 166         virtual ld::Fixup::iterator                             fixupsBegin() const
 167                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsBegin() : (ld::Fixup*)&_file._fixupToInternal); }
 168         virtual ld::Fixup::iterator                             fixupsEnd() const
 169                                                                                                                         { return (_compiledAtom ? _compiledAtom->fixupsEnd() : &((ld::Fixup*)&_file._fixupToInternal)[1]); }
 170         virtual ld::Atom::UnwindInfo::iterator  beginUnwind() const
 171                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginUnwind() : NULL); }
 172         virtual ld::Atom::UnwindInfo::iterator  endUnwind() const
 173                                                                                                                         { return (_compiledAtom ? _compiledAtom->endUnwind() : NULL); }
 174         virtual ld::Atom::LineInfo::iterator    beginLineInfo() const
 175                                                                                                                         { return (_compiledAtom ? _compiledAtom->beginLineInfo() : NULL); }
 176         virtual ld::Atom::LineInfo::iterator    endLineInfo() const
 177                                                                                                                         { return (_compiledAtom ? _compiledAtom->endLineInfo() : NULL); }
 178
 179         const ld::Atom*                                         compiledAtom()          { return _compiledAtom; }
 180         void                                                            setCompiledAtom(const ld::Atom& atom);
 181
 182 private:
 183
 184         File&                                                           _file;
 185         const char*                                                     _name;
 186         const ld::Atom*                                         _compiledAtom;
 187 };
 188
 189
 190
 191
 192
 193
 194
 195 class Parser
 196 {
 197 public:
 198         static bool                                             validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch);
 199         static const char*                              fileKind(const uint8_t* fileContent, uint64_t fileLength);
 200         static File*                                    parse(const uint8_t* fileContent, uint64_t fileLength, const char* path,
 201                                                                                         time_t modTime, cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles);
 202         static bool                                             libLTOisLoaded() { return (::lto_get_version() != NULL); }
 203         static bool                                             optimize(   const std::vector<const ld::Atom*>& allAtoms,
 204                                                                                                 ld::Internal&                                           state,
 205                                                                                                 const OptimizeOptions&                          options,
 206                                                                                                 ld::File::AtomHandler&                          handler,
 207                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 208                                                                                                 std::vector<const char*>&                       additionalUndefines);
 209
 210         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 211
 212 private:
 213         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 214         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options);
 215
 216         class CStringEquals
 217         {
 218         public:
 219                 bool operator()(const char* left, const char* right) const { return (strcmp(left, right) == 0); }
 220         };
 221         typedef __gnu_cxx::hash_set<const char*, __gnu_cxx::hash<const char*>, CStringEquals>  CStringSet;
 222         typedef __gnu_cxx::hash_map<const char*, Atom*, __gnu_cxx::hash<const char*>, CStringEquals> CStringToAtom;
 223
 224         class AtomSyncer : public ld::File::AtomHandler {
 225         public:
 226                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 227                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 228                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 229                 virtual void            doAtom(const class ld::Atom&);
 230                 virtual void            doFile(const class ld::File&) { }
 231
 232                 const OptimizeOptions&                  _options;
 233                 std::vector<const char*>&               _additionalUndefines;
 234                 std::vector<const ld::Atom*>&   _newAtoms;
 235                 CStringToAtom                                   _llvmAtoms;
 236                 CStringToAtom                                   _deadllvmAtoms;
 237         };
 238
 239         static std::vector<File*>               _s_files;
 240 };
 241
 242 std::vector<File*> Parser::_s_files;
 243
 244
 245 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 246 {
 247         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 248                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 249                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 250                         if ( !result ) {
 251                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 252                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 253                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 254                                 }
 255                         }
 256                         return result;
 257                 }
 258         }
 259         return false;
 260 }
 261
 262 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 263 {
 264         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 265                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 266                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 267                         if ( arch == t->cpuType ) {
 268                                  if ( t->isSubType ) {
 269                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 270                                                 return t->archName;
 271                                 }
 272                                 else {
 273                                         return t->archName;
 274                                 }
 275                         }
 276                 }
 277                 return "unknown bitcode architecture";
 278         }
 279         return NULL;
 280 }
 281
 282 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime,
 283                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 284 {
 285         File* f = new File(path, modTime, fileContent, fileLength, architecture);
 286         _s_files.push_back(f);
 287         if ( logAllFiles )
 288                 printf("%s\n", path);
 289         return f;
 290 }
 291
 292
 293 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 294 {
 295         mach_o::relocatable::ParserOptions objOpts;
 296         objOpts.architecture            = options.arch;
 297         objOpts.objSubtypeMustMatch = false;
 298         objOpts.logAllFiles                     = false;
 299         objOpts.convertUnwindInfo       = true;
 300         objOpts.subType                         = 0;
 301
 302         // mach-o parsing is done in-memory, but need path for debug notes
 303         const char* path = "/tmp/lto.o";
 304         time_t modTime = 0;
 305         if ( options.tmpObjectFilePath != NULL ) {
 306                 path = options.tmpObjectFilePath;
 307                 struct stat statBuffer;
 308                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 309                         modTime = statBuffer.st_mtime;
 310         }
 311
 312         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 313         if ( result != NULL )
 314                 return result;
 315         throw "LLVM LTO, file is not of required architecture";
 316 }
 317
 318
 319
 320 File::File(const char* pth, time_t mTime, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 321         : ld::relocatable::File(pth,mTime,ld::File::Ordinal::LTOOrdinal()), _architecture(arch), _internalAtom(*this),
 322         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 323         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 324         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 325         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 326 {
 327         const bool log = false;
 328
 329         // create llvm module
 330         _module = ::lto_module_create_from_memory(content, contentLength);
 331     if ( _module == NULL )
 332                 throwf("could not parse object file %s: %s", pth, lto_get_error_message());
 333
 334         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 335
 336         // create atom for each global symbol in module
 337         uint32_t count = ::lto_module_get_num_symbols(_module);
 338         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 339         for (uint32_t i=0; i < count; ++i) {
 340                 const char* name = ::lto_module_get_symbol_name(_module, i);
 341                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 342
 343                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 344                 // ignore dtrace static probes for now
 345                 // later when codegen is done and a mach-o file is produces the probes will be processed
 346                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 347                         continue;
 348
 349                 ld::Atom::Definition def;
 350                 ld::Atom::Combine combine = ld::Atom::combineNever;
 351                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 352                         case LTO_SYMBOL_DEFINITION_REGULAR:
 353                                 def = ld::Atom::definitionRegular;
 354                                 break;
 355                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 356                                 def = ld::Atom::definitionTentative;
 357                                 break;
 358                         case LTO_SYMBOL_DEFINITION_WEAK:
 359                                 def = ld::Atom::definitionRegular;
 360                                 combine = ld::Atom::combineByName;
 361                                 break;
 362                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 363                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 364                                 def = ld::Atom::definitionProxy;
 365                                 break;
 366                         default:
 367                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 368                 }
 369
 370                 // make LLVM atoms for definitions and a reference for undefines
 371                 if ( def != ld::Atom::definitionProxy ) {
 372                         ld::Atom::Scope scope;
 373                         bool autohide = false;
 374                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 375                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 376                                         scope = ld::Atom::scopeTranslationUnit;
 377                                         break;
 378                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 379                                         scope = ld::Atom::scopeLinkageUnit;
 380                                         break;
 381                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 382                                         scope = ld::Atom::scopeGlobal;
 383                                         break;
 384 #if LTO_API_VERSION >= 4
 385                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 386                                         scope = ld::Atom::scopeGlobal;
 387                                         autohide = true;
 388                                         break;
 389 #endif
 390                                 default:
 391                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 392                         }
 393                         // only make atoms for non-internal symbols
 394                         if ( scope == ld::Atom::scopeTranslationUnit )
 395                                 continue;
 396                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 397                         // make Atom using placement new operator
 398                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 399                         if ( scope != ld::Atom::scopeTranslationUnit )
 400                                 _internalAtom.addReference(name);
 401                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 402                 }
 403                 else {
 404                         // add to list of external references
 405                         _internalAtom.addReference(name);
 406                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 407                 }
 408         }
 409 }
 410
 411 File::~File()
 412 {
 413         if ( _module != NULL )
 414                 ::lto_module_dispose(_module);
 415 }
 416
 417 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 418 {
 419         handler.doAtom(_internalAtom);
 420         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 421                 handler.doAtom(_atomArray[i]);
 422         }
 423         return true;
 424 }
 425
 426 InternalAtom::InternalAtom(File& f)
 427         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 428                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 429                 _file(f)
 430 {
 431 }
 432
 433 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 434                         ld::Atom::Alignment a, bool ah)
 435         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 436                                 ld::Atom::symbolTableIn, false, false, false, a),
 437                 _file(f), _name(nm), _compiledAtom(NULL)
 438 {
 439         if ( ah )
 440                 this->setAutoHide();
 441 }
 442
 443 void Atom::setCompiledAtom(const ld::Atom& atom)
 444 {
 445         // set delegate so virtual methods go to it
 446         _compiledAtom = &atom;
 447
 448         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 449
 450         // update fields in ld::Atom to match newly constructed mach-o atom
 451         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 452 }
 453
 454
 455
 456 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 457                                                 ld::Internal&                                           state,
 458                                                 const OptimizeOptions&                          options,
 459                                                 ld::File::AtomHandler&                          handler,
 460                                                 std::vector<const ld::Atom*>&           newAtoms,
 461                                                 std::vector<const char*>&                       additionalUndefines)
 462 {
 463         const bool logMustPreserve = false;
 464         const bool logExtraOptions = false;
 465         const bool logBitcodeFiles = false;
 466         const bool logAtomsBeforeSync = false;
 467
 468         // exit quickly if nothing to do
 469         if ( _s_files.size() == 0 )
 470                 return false;
 471
 472         // print out LTO version string if -v was used
 473         if ( options.verbose )
 474                 fprintf(stderr, "%s\n", lto_get_version());
 475
 476         // create optimizer and add each Reader
 477         lto_code_gen_t generator = ::lto_codegen_create();
 478         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 479                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", (*it)->path());
 480                 if ( ::lto_codegen_add_module(generator, (*it)->module()) )
 481                         throwf("lto: could not merge in %s because %s", (*it)->path(), ::lto_get_error_message());
 482         }
 483
 484         // add any -mllvm command line options
 485         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 486                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 487                 ::lto_codegen_debug_options(generator, *it);
 488         }
 489
 490         // The atom graph uses directed edges (references). Collect all references where
 491         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 492         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 493         // defined in llvm bitcode file.
 494         CStringSet nonLLVMRefs;
 495         CStringToAtom llvmAtoms;
 496     bool hasNonllvmAtoms = false;
 497         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 498                 const ld::Atom* atom = *it;
 499                 // only look at references that come from an atom that is not an llvm atom
 500                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 501                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 502                                 hasNonllvmAtoms = true;
 503                         }
 504                         const ld::Atom* target;
 505                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 506                                 switch ( fit->binding ) {
 507                                         case ld::Fixup::bindingDirectlyBound:
 508                                                 // that reference an llvm atom
 509                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 510                                                         nonLLVMRefs.insert(fit->u.target->name());
 511                                                 break;
 512                                         case ld::Fixup::bindingsIndirectlyBound:
 513                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 514                                                 if ( target == NULL )
 515                                                         throwf("'%s' in %s contains undefined reference", atom->name(), atom->file()->path());
 516                                                 assert(target != NULL);
 517                                                 if ( target->contentType() == ld::Atom::typeLTOtemporary )
 518                                                         nonLLVMRefs.insert(target->name());
 519                                         default:
 520                                                 break;
 521                                 }
 522                         }
 523                 }
 524                 else {
 525                         llvmAtoms[atom->name()] = (Atom*)atom;
 526                 }
 527         }
 528         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 529         if ( state.entryPoint!= NULL ) {
 530                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 531                         nonLLVMRefs.insert(state.entryPoint->name());
 532         }
 533
 534         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 535         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 536         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 537         // atom so that the linker can replace it with the mach-o one later.
 538         CStringToAtom deadllvmAtoms;
 539         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 540                 const ld::Atom* atom = *it;
 541                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 542                         const char* name = atom->name();
 543                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 544                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 545                         deadllvmAtoms[name] = (Atom*)atom;
 546                 }
 547         }
 548         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 549                 File* file = *it;
 550                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 551                         Atom* llvmAtom = &file->_atomArray[i];
 552                         if ( llvmAtom->coalescedAway()  ) {
 553                                 const char* name = llvmAtom->name();
 554                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 555                                         if ( logMustPreserve )
 556                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 557                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 558                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 559                                 }
 560                         }
 561                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 562                                 const char* name = llvmAtom->name();
 563                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 564                         }
 565                 }
 566         }
 567
 568         // tell code generator about symbols that must be preserved
 569         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 570                 const char* name = it->first;
 571                 Atom* atom = it->second;
 572                 // Include llvm Symbol in export list if it meets one of following two conditions
 573                 // 1 - atom scope is global (and not linkage unit).
 574                 // 2 - included in nonLLVMRefs set.
 575                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 576                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 577                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 578                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 579                 }
 580                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 581                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 582                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 583                 }
 584         }
 585
 586     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 587     if ( options.relocatable && !hasNonllvmAtoms ) {
 588                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 589                         // HACK, no good way to tell linker we are all done, so just quit
 590                         exit(0);
 591                 }
 592                 warning("could not produce merged bitcode file");
 593     }
 594
 595         // set code-gen model
 596         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 597         if ( options.mainExecutable ) {
 598                 if ( options.staticExecutable ) {
 599                         // darwin x86_64 "static" code model is really dynamic code model
 600                         if ( options.arch == CPU_TYPE_X86_64 )
 601                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 602                         else
 603                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 604                 }
 605                 else {
 606                         if ( options.pie )
 607                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 608                         else
 609                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 610                 }
 611         }
 612         else {
 613                 if ( options.allowTextRelocs )
 614                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 615                 else
 616                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 617         }
 618         if ( ::lto_codegen_set_pic_model(generator, model) )
 619                 throwf("could not create set codegen model: %s", lto_get_error_message());
 620
 621     // if requested, save off merged bitcode file
 622     if ( options.saveTemps ) {
 623         char tempBitcodePath[MAXPATHLEN];
 624         strcpy(tempBitcodePath, options.outputFilePath);
 625         strcat(tempBitcodePath, ".lto.bc");
 626         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 627     }
 628
 629 #if LTO_API_VERSION >= 3
 630         // find assembler next to linker
 631         char path[PATH_MAX];
 632         uint32_t bufSize = PATH_MAX;
 633         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 634                 char* lastSlash = strrchr(path, '/');
 635                 if ( lastSlash != NULL ) {
 636                         strcpy(lastSlash+1, "as");
 637                         struct stat statInfo;
 638                         if ( stat(path, &statInfo) == 0 )
 639                                 ::lto_codegen_set_assembler_path(generator, path);
 640                 }
 641         }
 642 #endif
 643         // run code generator
 644         size_t machOFileLen;
 645         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 646         if ( machOFile == NULL )
 647                 throwf("could not do LTO codegen: %s", ::lto_get_error_message());
 648
 649     // if requested, save off temp mach-o file
 650     if ( options.saveTemps ) {
 651         char tempMachoPath[MAXPATHLEN];
 652         strcpy(tempMachoPath, options.outputFilePath);
 653         strcat(tempMachoPath, ".lto.o");
 654         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 655                 if ( fd != -1) {
 656                         ::write(fd, machOFile, machOFileLen);
 657                         ::close(fd);
 658                 }
 659                 //      save off merged bitcode file
 660                 char tempOptBitcodePath[MAXPATHLEN];
 661         strcpy(tempOptBitcodePath, options.outputFilePath);
 662         strcat(tempOptBitcodePath, ".lto.opt.bc");
 663         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 664         }
 665
 666         // if needed, save temp mach-o file to specific location
 667         if ( options.tmpObjectFilePath != NULL ) {
 668                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 669                 if ( fd != -1) {
 670                         ::write(fd, machOFile, machOFileLen);
 671                         ::close(fd);
 672                 }
 673                 else {
 674                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 675                 }
 676         }
 677
 678         // parse generated mach-o file into a MachOReader
 679         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 680
 681         // sync generated mach-o atoms with existing atoms ld knows about
 682         if ( logAtomsBeforeSync ) {
 683                 fprintf(stderr, "llvmAtoms:\n");
 684                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 685                         const char* name = it->first;
 686                         //Atom* atom = it->second;
 687                         fprintf(stderr, "\t%s\n", name);
 688                 }
 689                 fprintf(stderr, "deadllvmAtoms:\n");
 690                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 691                         const char* name = it->first;
 692                         //Atom* atom = it->second;
 693                         fprintf(stderr, "\t%s\n", name);
 694                 }
 695         }
 696         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 697         machoFile->forEachAtom(syncer);
 698
 699         // Remove InternalAtoms from ld
 700         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 701                 (*it)->internalAtom().setCoalescedAway();
 702         }
 703         // Remove Atoms from ld if code generator optimized them away
 704         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 705                 // check if setRealAtom() called on this Atom
 706                 if ( li->second->compiledAtom() == NULL ) {
 707                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 708                         li->second->setCoalescedAway();
 709                 }
 710         }
 711
 712         // notify about file level attributes
 713         handler.doFile(*machoFile);
 714
 715         // if final mach-o file has debug info, update original bitcode files to match
 716         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 717                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 718                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 719         }
 720
 721         return true;
 722 }
 723
 724
 725 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 726 {
 727         // update proxy atoms to point to real atoms and find new atoms
 728         const char* name = machoAtom.name();
 729         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 730                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 731                 if ( pos != _llvmAtoms.end() ) {
 732                         // turn Atom into a proxy for this mach-o atom
 733                         pos->second->setCompiledAtom(machoAtom);
 734                 }
 735                 else {
 736                         // an atom of this name was not in the allAtoms list the linker gave us
 737                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 738                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 739                                 if ( _options.linkerDeadStripping ) {
 740                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 741                                         Atom* llvmAtom = _deadllvmAtoms[name];
 742                                         llvmAtom->setCompiledAtom(machoAtom);
 743                                         _newAtoms.push_back(&machoAtom);
 744                                 }
 745                                 else {
 746                                         // Don't pass it back as a new atom
 747                                 }
 748                         }
 749                         else
 750                         {
 751                                 // this is something new that lto conjured up, tell ld its new
 752                                 _newAtoms.push_back(&machoAtom);
 753                         }
 754                 }
 755         }
 756         else {
 757                 // ld only knew about non-static atoms, so this one must be new
 758                 _newAtoms.push_back(&machoAtom);
 759         }
 760
 761         // adjust fixups to go through proxy atoms
 762         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 763         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 764                 switch ( fit->binding ) {
 765                         case ld::Fixup::bindingNone:
 766                                 break;
 767                         case ld::Fixup::bindingByNameUnbound:
 768                                 // don't know if this target has been seen by linker before or if it is new
 769                                 // be conservative and tell linker it is new
 770                                 _additionalUndefines.push_back(fit->u.name);
 771                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 772                                 break;
 773                         case ld::Fixup::bindingDirectlyBound:
 774                                 // If mach-o atom is referencing another mach-o atom then
 775                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 776                                 // llvm symbol references always go through Atom proxy.
 777                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 778                                         const char* targetName = fit->u.target->name();
 779                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 780                                         if ( pos != _llvmAtoms.end() ) {
 781                                                 fit->u.target = pos->second;
 782                                         }
 783                                         else {
 784                                                 if ( _deadllvmAtoms.find(targetName) != _deadllvmAtoms.end() ) {
 785                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 786                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 787                                                         fit->u.name = targetName;
 788                                                 }
 789                                         }
 790                                 }
 791                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 792                                 break;
 793                         case ld::Fixup::bindingByContentBound:
 794                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 795                                 break;
 796                         case ld::Fixup::bindingsIndirectlyBound:
 797                                 assert(0 && "indirect binding found in initial mach-o file?");
 798                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 799                                 break;
 800                 }
 801         }
 802
 803 }
 804
 805 class Mutex {
 806         static pthread_mutex_t lto_lock;
 807 public:
 808         Mutex() { pthread_mutex_lock(&lto_lock); }
 809         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 810 };
 811 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 812
 813 //
 814 // Used by archive reader to see if member is an llvm bitcode file
 815 //
 816 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 817 {
 818         Mutex lock;
 819         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 820 }
 821
 822
 823 //
 824 // main function used by linker to instantiate ld::Files
 825 //
 826 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 827                                                                 const char* path, time_t modTime,
 828                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 829 {
 830         Mutex lock;
 831         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 832                 return Parser::parse(fileContent, fileLength, path, modTime, architecture, subarch, logAllFiles);
 833         else
 834                 return NULL;
 835 }
 836
 837 //
 838 // used by "ld -v" to report version of libLTO.dylib being used
 839 //
 840 const char* version()
 841 {
 842         Mutex lock;
 843         return ::lto_get_version();
 844 }
 845
 846
 847 //
 848 // used by ld for error reporting
 849 //
 850 bool libLTOisLoaded()
 851 {
 852         Mutex lock;
 853         return (::lto_get_version() != NULL);
 854 }
 855
 856 //
 857 // used by ld for error reporting
 858 //
 859 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 860 {
 861         Mutex lock;
 862         return Parser::fileKind(fileContent, fileLength);
 863 }
 864
 865 //
 866 // used by ld for doing link time optimization
 867 //
 868 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 869                                 ld::Internal&                                           state,
 870                                 const OptimizeOptions&                          options,
 871                                 ld::File::AtomHandler&                          handler,
 872                                 std::vector<const ld::Atom*>&           newAtoms,
 873                                 std::vector<const char*>&                       additionalUndefines)
 874 {
 875         Mutex lock;
 876         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 877 }
 878
 879
 880
 881 }; // namespace lto
 882
 883
 884 #endif
 885