src/ld/parsers/lto_file.cpp

   1 /* -*- mode: C++; c-basic-offset: 4; tab-width: 4 -*-
   2  *
   3  * Copyright (c) 2006-2010 Apple Inc. All rights reserved.
   4  *
   5  * @APPLE_LICENSE_HEADER_START@
   6  *
   7  * This file contains Original Code and/or Modifications of Original Code
   8  * as defined in and that are subject to the Apple Public Source License
   9  * Version 2.0 (the 'License'). You may not use this file except in
  10  * compliance with the License. Please obtain a copy of the License at
  11  * http://www.opensource.apple.com/apsl/ and read it before using this
  12  * file.
  13  *
  14  * The Original Code and all software distributed under the License are
  15  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  16  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  17  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  18  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  19  * Please see the License for the specific language governing rights and
  20  * limitations under the License.
  21  *
  22  * @APPLE_LICENSE_HEADER_END@
  23  */
  24
  25 #ifndef __LTO_READER_H__
  26 #define __LTO_READER_H__
  27
  28 #include <stdlib.h>
  29 #include <sys/param.h>
  30 #include <sys/fcntl.h>
  31 #include <sys/stat.h>
  32 #include <errno.h>
  33 #include <pthread.h>
  34 #include <mach-o/dyld.h>
  35 #include <vector>
  36 #include <unordered_set>
  37 #include <unordered_map>
  38
  39 #include "MachOFileAbstraction.hpp"
  40 #include "Architectures.hpp"
  41 #include "ld.hpp"
  42 #include "macho_relocatable_file.h"
  43 #include "lto_file.h"
  44
  45 // #defines are a work around for <rdar://problem/8760268>
  46 #define __STDC_LIMIT_MACROS 1
  47 #define __STDC_CONSTANT_MACROS 1
  48 #include "llvm-c/lto.h"
  49
  50
  51 namespace lto {
  52
  53
  54 //
  55 // ld64 only tracks non-internal symbols from an llvm bitcode file.
  56 // We model this by having an InternalAtom which represent all internal functions and data.
  57 // All non-interal symbols from a bitcode file are represented by an Atom
  58 // and each Atom has a reference to the InternalAtom.  The InternalAtom
  59 // also has references to each symbol external to the bitcode file.
  60 //
  61 class InternalAtom : public ld::Atom
  62 {
  63 public:
  64                                                                                                 InternalAtom(class File& f);
  65         // overrides of ld::Atom
  66         virtual ld::File*                                                       file() const            { return &_file; }
  67         virtual const char*                                                     name() const            { return "import-atom"; }
  68         virtual uint64_t                                                        size() const            { return 0; }
  69         virtual uint64_t                                                        objectAddress() const { return 0; }
  70         virtual void                                                            copyRawContent(uint8_t buffer[]) const { }
  71         virtual void                                                            setScope(Scope)         { }
  72         virtual ld::Fixup::iterator                                     fixupsBegin() const     { return &_undefs[0]; }
  73         virtual ld::Fixup::iterator                                     fixupsEnd()     const   { return &_undefs[_undefs.size()]; }
  74
  75         // for adding references to symbols outside bitcode file
  76         void                                                                            addReference(const char* nm)
  77                                                                                                                                         { _undefs.push_back(ld::Fixup(0, ld::Fixup::k1of1,
  78                                                                                                                                                                 ld::Fixup::kindNone, false, nm)); }
  79 private:
  80
  81         ld::File&                                                                       _file;
  82         mutable std::vector<ld::Fixup>                          _undefs;
  83 };
  84
  85
  86 //
  87 // LLVM bitcode file
  88 //
  89 class File : public ld::relocatable::File
  90 {
  91 public:
  92                                                                                         File(const char* path, time_t mTime, ld::File::Ordinal ordinal,
  93                                                                                                          const uint8_t* content, uint32_t contentLength, cpu_type_t arch);
  94         virtual                                                                 ~File();
  95
  96         // overrides of ld::File
  97         virtual bool                                                                            forEachAtom(ld::File::AtomHandler&) const;
  98         virtual bool                                                                            justInTimeforEachAtom(const char* name, ld::File::AtomHandler&) const
  99                                                                                                                                                                         { return false; }
 100         virtual uint32_t                                                                        cpuSubType() const                      { return _cpuSubType; }
 101
 102         // overrides of ld::relocatable::File
 103         virtual DebugInfoKind                                                           debugInfo()     const                   { return _debugInfo; }
 104         virtual const char*                                                                     debugInfoPath() const           { return _debugInfoPath; }
 105         virtual time_t                                                                          debugInfoModificationTime() const
 106                                                                                                                                                                         { return _debugInfoModTime; }
 107         virtual const std::vector<ld::relocatable::File::Stab>* stabs() const                   { return NULL; }
 108         virtual bool                                                                            canScatterAtoms() const         { return true; }
 109         virtual LinkerOptionsList*                                                      linkerOptions() const           { return NULL; }
 110
 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 const char*                                     translationUnitSource() const
 155                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource() : NULL); }
 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, ld::File::Ordinal 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                                                                                                 const OptimizeOptions&                          options,
 207                                                                                                 ld::File::AtomHandler&                          handler,
 208                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 209                                                                                                 std::vector<const char*>&                       additionalUndefines);
 210
 211         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 212
 213 private:
 214         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 215         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options);
 216
 217         typedef std::unordered_set<const char*, ld::CStringHash, ld::CStringEquals>  CStringSet;
 218         typedef std::unordered_map<const char*, Atom*, ld::CStringHash, ld::CStringEquals> CStringToAtom;
 219
 220         class AtomSyncer : public ld::File::AtomHandler {
 221         public:
 222                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 223                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 224                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 225                 virtual void            doAtom(const class ld::Atom&);
 226                 virtual void            doFile(const class ld::File&) { }
 227
 228                 const OptimizeOptions&                  _options;
 229                 std::vector<const char*>&               _additionalUndefines;
 230                 std::vector<const ld::Atom*>&   _newAtoms;
 231                 CStringToAtom                                   _llvmAtoms;
 232                 CStringToAtom                                   _deadllvmAtoms;
 233         };
 234
 235         static std::vector<File*>               _s_files;
 236 };
 237
 238 std::vector<File*> Parser::_s_files;
 239
 240
 241 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 242 {
 243         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 244                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 245                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 246                         if ( !result ) {
 247                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 248                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 249                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 250                                 }
 251                         }
 252                         return result;
 253                 }
 254         }
 255         return false;
 256 }
 257
 258 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 259 {
 260         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 261                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 262                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 263                         if ( arch == t->cpuType ) {
 264                                  if ( t->isSubType ) {
 265                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 266                                                 return t->archName;
 267                                 }
 268                                 else {
 269                                         return t->archName;
 270                                 }
 271                         }
 272                 }
 273                 return "unknown bitcode architecture";
 274         }
 275         return NULL;
 276 }
 277
 278 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime, ld::File::Ordinal ordinal,
 279                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 280 {
 281         File* f = new File(path, modTime, ordinal, fileContent, fileLength, architecture);
 282         _s_files.push_back(f);
 283         if ( logAllFiles )
 284                 printf("%s\n", path);
 285         return f;
 286 }
 287
 288
 289 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 290 {
 291         mach_o::relocatable::ParserOptions objOpts;
 292         objOpts.architecture            = options.arch;
 293         objOpts.objSubtypeMustMatch = false;
 294         objOpts.logAllFiles                     = false;
 295         objOpts.warnUnwindConversionProblems    = options.needsUnwindInfoSection;
 296         objOpts.keepDwarfUnwind         = options.keepDwarfUnwind;
 297         objOpts.forceDwarfConversion = false;
 298         objOpts.subType                         = 0;
 299
 300         // mach-o parsing is done in-memory, but need path for debug notes
 301         const char* path = "/tmp/lto.o";
 302         time_t modTime = 0;
 303         if ( options.tmpObjectFilePath != NULL ) {
 304                 path = options.tmpObjectFilePath;
 305                 struct stat statBuffer;
 306                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 307                         modTime = statBuffer.st_mtime;
 308         }
 309
 310         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 311         if ( result != NULL )
 312                 return result;
 313         throw "LLVM LTO, file is not of required architecture";
 314 }
 315
 316
 317
 318 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 319         : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
 320         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 321         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 322         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 323         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 324 {
 325         const bool log = false;
 326
 327         // create llvm module
 328         _module = ::lto_module_create_from_memory(content, contentLength);
 329     if ( _module == NULL )
 330                 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth, ::lto_get_error_message(), ::lto_get_version());
 331
 332         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 333
 334         // create atom for each global symbol in module
 335         uint32_t count = ::lto_module_get_num_symbols(_module);
 336         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 337         for (uint32_t i=0; i < count; ++i) {
 338                 const char* name = ::lto_module_get_symbol_name(_module, i);
 339                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 340
 341                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 342                 // ignore dtrace static probes for now
 343                 // later when codegen is done and a mach-o file is produces the probes will be processed
 344                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 345                         continue;
 346
 347                 ld::Atom::Definition def;
 348                 ld::Atom::Combine combine = ld::Atom::combineNever;
 349                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 350                         case LTO_SYMBOL_DEFINITION_REGULAR:
 351                                 def = ld::Atom::definitionRegular;
 352                                 break;
 353                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 354                                 def = ld::Atom::definitionTentative;
 355                                 break;
 356                         case LTO_SYMBOL_DEFINITION_WEAK:
 357                                 def = ld::Atom::definitionRegular;
 358                                 combine = ld::Atom::combineByName;
 359                                 break;
 360                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 361                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 362                                 def = ld::Atom::definitionProxy;
 363                                 break;
 364                         default:
 365                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 366                 }
 367
 368                 // make LLVM atoms for definitions and a reference for undefines
 369                 if ( def != ld::Atom::definitionProxy ) {
 370                         ld::Atom::Scope scope;
 371                         bool autohide = false;
 372                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 373                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 374                                         scope = ld::Atom::scopeTranslationUnit;
 375                                         break;
 376                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 377                                         scope = ld::Atom::scopeLinkageUnit;
 378                                         break;
 379                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 380                                         scope = ld::Atom::scopeGlobal;
 381                                         break;
 382 #if LTO_API_VERSION >= 4
 383                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 384                                         scope = ld::Atom::scopeGlobal;
 385                                         autohide = true;
 386                                         break;
 387 #endif
 388                                 default:
 389                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 390                         }
 391                         // only make atoms for non-internal symbols
 392                         if ( scope == ld::Atom::scopeTranslationUnit )
 393                                 continue;
 394                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 395                         // make Atom using placement new operator
 396                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 397                         if ( scope != ld::Atom::scopeTranslationUnit )
 398                                 _internalAtom.addReference(name);
 399                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 400                 }
 401                 else {
 402                         // add to list of external references
 403                         _internalAtom.addReference(name);
 404                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 405                 }
 406         }
 407 }
 408
 409 File::~File()
 410 {
 411         if ( _module != NULL )
 412                 ::lto_module_dispose(_module);
 413 }
 414
 415 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 416 {
 417         handler.doAtom(_internalAtom);
 418         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 419                 handler.doAtom(_atomArray[i]);
 420         }
 421         return true;
 422 }
 423
 424 InternalAtom::InternalAtom(File& f)
 425         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 426                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 427                 _file(f)
 428 {
 429 }
 430
 431 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 432                         ld::Atom::Alignment a, bool ah)
 433         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 434                                 ld::Atom::symbolTableIn, false, false, false, a),
 435                 _file(f), _name(nm), _compiledAtom(NULL)
 436 {
 437         if ( ah )
 438                 this->setAutoHide();
 439 }
 440
 441 void Atom::setCompiledAtom(const ld::Atom& atom)
 442 {
 443         // set delegate so virtual methods go to it
 444         _compiledAtom = &atom;
 445
 446         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 447
 448         // update fields in ld::Atom to match newly constructed mach-o atom
 449         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 450 }
 451
 452
 453
 454 // <rdar://problem/12379604> The order that files are merged must match command line order
 455 struct CommandLineOrderFileSorter
 456 {
 457      bool operator()(File* left, File* right)
 458      {
 459         return ( left->ordinal() < right->ordinal() );
 460      }
 461 };
 462
 463
 464 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 465                                                 ld::Internal&                                           state,
 466                                                 const OptimizeOptions&                          options,
 467                                                 ld::File::AtomHandler&                          handler,
 468                                                 std::vector<const ld::Atom*>&           newAtoms,
 469                                                 std::vector<const char*>&                       additionalUndefines)
 470 {
 471         const bool logMustPreserve = false;
 472         const bool logExtraOptions = false;
 473         const bool logBitcodeFiles = false;
 474         const bool logAtomsBeforeSync = false;
 475
 476         // exit quickly if nothing to do
 477         if ( _s_files.size() == 0 )
 478                 return false;
 479
 480         // print out LTO version string if -v was used
 481         if ( options.verbose )
 482                 fprintf(stderr, "%s\n", ::lto_get_version());
 483
 484         // create optimizer and add each Reader
 485         lto_code_gen_t generator = ::lto_codegen_create();
 486         // <rdar://problem/12379604> The order that files are merged must match command line order
 487         std::sort(_s_files.begin(), _s_files.end(), CommandLineOrderFileSorter());
 488         ld::File::Ordinal lastOrdinal;
 489         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 490                 File* f = *it;
 491                 assert(f->ordinal() > lastOrdinal);
 492                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", f->path());
 493                 if ( ::lto_codegen_add_module(generator, f->module()) )
 494                         throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", f->path(), ::lto_get_error_message(), ::lto_get_version());
 495                 lastOrdinal = f->ordinal();
 496         }
 497
 498         // add any -mllvm command line options
 499         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 500                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 501                 ::lto_codegen_debug_options(generator, *it);
 502         }
 503
 504         // <rdar://problem/13687397> Need a way for LTO to get cpu variants (until that info is in bitcode)
 505         if ( options.mcpu != NULL )
 506                 ::lto_codegen_set_cpu(generator, options.mcpu);
 507
 508         // The atom graph uses directed edges (references). Collect all references where
 509         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 510         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 511         // defined in llvm bitcode file.
 512         CStringSet nonLLVMRefs;
 513         CStringToAtom llvmAtoms;
 514     bool hasNonllvmAtoms = false;
 515         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 516                 const ld::Atom* atom = *it;
 517                 // only look at references that come from an atom that is not an llvm atom
 518                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 519                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 520                                 hasNonllvmAtoms = true;
 521                         }
 522                         const ld::Atom* target;
 523                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 524                                 switch ( fit->binding ) {
 525                                         case ld::Fixup::bindingDirectlyBound:
 526                                                 // that reference an llvm atom
 527                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 528                                                         nonLLVMRefs.insert(fit->u.target->name());
 529                                                 break;
 530                                         case ld::Fixup::bindingsIndirectlyBound:
 531                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 532                                                 if ( target == NULL )
 533                                                         throwf("'%s' in %s contains undefined reference", atom->name(), atom->file()->path());
 534                                                 assert(target != NULL);
 535                                                 if ( target->contentType() == ld::Atom::typeLTOtemporary )
 536                                                         nonLLVMRefs.insert(target->name());
 537                                         default:
 538                                                 break;
 539                                 }
 540                         }
 541                 }
 542                 else {
 543                         llvmAtoms[atom->name()] = (Atom*)atom;
 544                 }
 545         }
 546         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 547         if ( state.entryPoint!= NULL ) {
 548                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 549                         nonLLVMRefs.insert(state.entryPoint->name());
 550         }
 551
 552         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 553         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 554         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 555         // atom so that the linker can replace it with the mach-o one later.
 556         CStringToAtom deadllvmAtoms;
 557         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 558                 const ld::Atom* atom = *it;
 559                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 560                         const char* name = atom->name();
 561                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 562                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 563                         deadllvmAtoms[name] = (Atom*)atom;
 564                 }
 565         }
 566         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 567                 File* file = *it;
 568                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 569                         Atom* llvmAtom = &file->_atomArray[i];
 570                         if ( llvmAtom->coalescedAway()  ) {
 571                                 const char* name = llvmAtom->name();
 572                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 573                                         if ( logMustPreserve )
 574                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 575                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 576                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 577                                 }
 578                         }
 579                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 580                                 const char* name = llvmAtom->name();
 581                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 582                         }
 583                 }
 584         }
 585
 586         // tell code generator about symbols that must be preserved
 587         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 588                 const char* name = it->first;
 589                 Atom* atom = it->second;
 590                 // Include llvm Symbol in export list if it meets one of following two conditions
 591                 // 1 - atom scope is global (and not linkage unit).
 592                 // 2 - included in nonLLVMRefs set.
 593                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 594                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 595                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 596                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 597                 }
 598                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 599                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 600                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 601                 }
 602         }
 603
 604     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 605     if ( options.relocatable && !hasNonllvmAtoms ) {
 606                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 607                         // HACK, no good way to tell linker we are all done, so just quit
 608                         exit(0);
 609                 }
 610                 warning("could not produce merged bitcode file");
 611     }
 612
 613         // set code-gen model
 614         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 615         if ( options.mainExecutable ) {
 616                 if ( options.staticExecutable ) {
 617                         // x86_64 "static" or any "-static -pie" is really dynamic code model
 618                         if ( (options.arch == CPU_TYPE_X86_64) || options.pie )
 619                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 620                         else
 621                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 622                 }
 623                 else {
 624                         if ( options.pie )
 625                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 626                         else
 627                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 628                 }
 629         }
 630         else {
 631                 if ( options.allowTextRelocs )
 632                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 633                 else
 634                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 635         }
 636         if ( ::lto_codegen_set_pic_model(generator, model) )
 637                 throwf("could not create set codegen model: %s", lto_get_error_message());
 638
 639     // if requested, save off merged bitcode file
 640     if ( options.saveTemps ) {
 641         char tempBitcodePath[MAXPATHLEN];
 642         strcpy(tempBitcodePath, options.outputFilePath);
 643         strcat(tempBitcodePath, ".lto.bc");
 644         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 645     }
 646
 647 #if LTO_API_VERSION >= 3
 648         // find assembler next to linker
 649         char path[PATH_MAX];
 650         uint32_t bufSize = PATH_MAX;
 651         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 652                 char* lastSlash = strrchr(path, '/');
 653                 if ( lastSlash != NULL ) {
 654                         strcpy(lastSlash+1, "as");
 655                         struct stat statInfo;
 656                         if ( stat(path, &statInfo) == 0 )
 657                                 ::lto_codegen_set_assembler_path(generator, path);
 658                 }
 659         }
 660 #endif
 661         // run code generator
 662         size_t machOFileLen;
 663         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 664         if ( machOFile == NULL )
 665                 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
 666
 667     // if requested, save off temp mach-o file
 668     if ( options.saveTemps ) {
 669         char tempMachoPath[MAXPATHLEN];
 670         strcpy(tempMachoPath, options.outputFilePath);
 671         strcat(tempMachoPath, ".lto.o");
 672         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 673                 if ( fd != -1) {
 674                         ::write(fd, machOFile, machOFileLen);
 675                         ::close(fd);
 676                 }
 677                 //      save off merged bitcode file
 678                 char tempOptBitcodePath[MAXPATHLEN];
 679         strcpy(tempOptBitcodePath, options.outputFilePath);
 680         strcat(tempOptBitcodePath, ".lto.opt.bc");
 681         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 682         }
 683
 684         // if needed, save temp mach-o file to specific location
 685         if ( options.tmpObjectFilePath != NULL ) {
 686                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 687                 if ( fd != -1) {
 688                         ::write(fd, machOFile, machOFileLen);
 689                         ::close(fd);
 690                 }
 691                 else {
 692                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 693                 }
 694         }
 695
 696         // parse generated mach-o file into a MachOReader
 697         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 698
 699         // sync generated mach-o atoms with existing atoms ld knows about
 700         if ( logAtomsBeforeSync ) {
 701                 fprintf(stderr, "llvmAtoms:\n");
 702                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 703                         const char* name = it->first;
 704                         //Atom* atom = it->second;
 705                         fprintf(stderr, "\t%s\n", name);
 706                 }
 707                 fprintf(stderr, "deadllvmAtoms:\n");
 708                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 709                         const char* name = it->first;
 710                         //Atom* atom = it->second;
 711                         fprintf(stderr, "\t%s\n", name);
 712                 }
 713         }
 714         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 715         machoFile->forEachAtom(syncer);
 716
 717         // Remove InternalAtoms from ld
 718         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 719                 (*it)->internalAtom().setCoalescedAway();
 720         }
 721         // Remove Atoms from ld if code generator optimized them away
 722         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 723                 // check if setRealAtom() called on this Atom
 724                 if ( li->second->compiledAtom() == NULL ) {
 725                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 726                         li->second->setCoalescedAway();
 727                 }
 728         }
 729
 730         // notify about file level attributes
 731         handler.doFile(*machoFile);
 732
 733         // if final mach-o file has debug info, update original bitcode files to match
 734         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 735                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 736                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 737         }
 738
 739         return true;
 740 }
 741
 742
 743 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 744 {
 745         // update proxy atoms to point to real atoms and find new atoms
 746         const char* name = machoAtom.name();
 747         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 748                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 749                 if ( pos != _llvmAtoms.end() ) {
 750                         // turn Atom into a proxy for this mach-o atom
 751                         pos->second->setCompiledAtom(machoAtom);
 752                 }
 753                 else {
 754                         // an atom of this name was not in the allAtoms list the linker gave us
 755                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 756                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 757                                 if ( _options.linkerDeadStripping ) {
 758                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 759                                         Atom* llvmAtom = _deadllvmAtoms[name];
 760                                         llvmAtom->setCompiledAtom(machoAtom);
 761                                         _newAtoms.push_back(&machoAtom);
 762                                 }
 763                                 else {
 764                                         // Don't pass it back as a new atom
 765                                 }
 766                         }
 767                         else
 768                         {
 769                                 // this is something new that lto conjured up, tell ld its new
 770                                 _newAtoms.push_back(&machoAtom);
 771                         }
 772                 }
 773         }
 774         else {
 775                 // ld only knew about non-static atoms, so this one must be new
 776                 _newAtoms.push_back(&machoAtom);
 777         }
 778
 779         // adjust fixups to go through proxy atoms
 780         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 781         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 782                 switch ( fit->binding ) {
 783                         case ld::Fixup::bindingNone:
 784                                 break;
 785                         case ld::Fixup::bindingByNameUnbound:
 786                                 // don't know if this target has been seen by linker before or if it is new
 787                                 // be conservative and tell linker it is new
 788                                 _additionalUndefines.push_back(fit->u.name);
 789                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 790                                 break;
 791                         case ld::Fixup::bindingDirectlyBound:
 792                                 // If mach-o atom is referencing another mach-o atom then
 793                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 794                                 // llvm symbol references always go through Atom proxy.
 795                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 796                                         const char* targetName = fit->u.target->name();
 797                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 798                                         if ( pos != _llvmAtoms.end() ) {
 799                                                 fit->u.target = pos->second;
 800                                         }
 801                                         else {
 802                                                 // <rdar://problem/12859831> Don't unbind follow-on reference into by-name reference
 803                                                 if ( (_deadllvmAtoms.find(targetName) != _deadllvmAtoms.end()) && (fit->kind != ld::Fixup::kindNoneFollowOn) ) {
 804                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 805                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 806                                                         fit->u.name = targetName;
 807                                                 }
 808                                         }
 809                                 }
 810                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 811                                 break;
 812                         case ld::Fixup::bindingByContentBound:
 813                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 814                                 break;
 815                         case ld::Fixup::bindingsIndirectlyBound:
 816                                 assert(0 && "indirect binding found in initial mach-o file?");
 817                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 818                                 break;
 819                 }
 820         }
 821
 822 }
 823
 824 class Mutex {
 825         static pthread_mutex_t lto_lock;
 826 public:
 827         Mutex() { pthread_mutex_lock(&lto_lock); }
 828         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 829 };
 830 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 831
 832 //
 833 // Used by archive reader to see if member is an llvm bitcode file
 834 //
 835 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 836 {
 837         Mutex lock;
 838         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 839 }
 840
 841
 842 //
 843 // main function used by linker to instantiate ld::Files
 844 //
 845 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 846                                                                 const char* path, time_t modTime, ld::File::Ordinal ordinal,
 847                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles)
 848 {
 849         Mutex lock;
 850         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 851                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles);
 852         else
 853                 return NULL;
 854 }
 855
 856 //
 857 // used by "ld -v" to report version of libLTO.dylib being used
 858 //
 859 const char* version()
 860 {
 861         Mutex lock;
 862         return ::lto_get_version();
 863 }
 864
 865
 866 //
 867 // used by ld for error reporting
 868 //
 869 bool libLTOisLoaded()
 870 {
 871         Mutex lock;
 872         return (::lto_get_version() != NULL);
 873 }
 874
 875 //
 876 // used by ld for error reporting
 877 //
 878 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 879 {
 880         Mutex lock;
 881         return Parser::fileKind(fileContent, fileLength);
 882 }
 883
 884 //
 885 // used by ld for doing link time optimization
 886 //
 887 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 888                                 ld::Internal&                                           state,
 889                                 const OptimizeOptions&                          options,
 890                                 ld::File::AtomHandler&                          handler,
 891                                 std::vector<const ld::Atom*>&           newAtoms,
 892                                 std::vector<const char*>&                       additionalUndefines)
 893 {
 894         Mutex lock;
 895         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 896 }
 897
 898
 899
 900 }; // namespace lto
 901
 902
 903 #endif
 904