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 namespace lto {
  51
  52
  53 //
  54 // ld64 only tracks non-internal symbols from an llvm bitcode file.
  55 // We model this by having an InternalAtom which represent all internal functions and data.
  56 // All non-interal symbols from a bitcode file are represented by an Atom
  57 // and each Atom has a reference to the InternalAtom.  The InternalAtom
  58 // also has references to each symbol external to the bitcode file.
  59 //
  60 class InternalAtom : public ld::Atom
  61 {
  62 public:
  63                                                                                                 InternalAtom(class File& f);
  64         // overrides of ld::Atom
  65         virtual ld::File*                                                       file() const            { return &_file; }
  66         virtual const char*                                                     name() const            { return "import-atom"; }
  67         virtual uint64_t                                                        size() const            { return 0; }
  68         virtual uint64_t                                                        objectAddress() const { return 0; }
  69         virtual void                                                            copyRawContent(uint8_t buffer[]) const { }
  70         virtual void                                                            setScope(Scope)         { }
  71         virtual ld::Fixup::iterator                                     fixupsBegin() const     { return &_undefs[0]; }
  72         virtual ld::Fixup::iterator                                     fixupsEnd()     const   { return &_undefs[_undefs.size()]; }
  73
  74         // for adding references to symbols outside bitcode file
  75         void                                                                            addReference(const char* nm)
  76                                                                                                                                         { _undefs.push_back(ld::Fixup(0, ld::Fixup::k1of1,
  77                                                                                                                                                                 ld::Fixup::kindNone, false, nm)); }
  78 private:
  79
  80         ld::File&                                                                       _file;
  81         mutable std::vector<ld::Fixup>                          _undefs;
  82 };
  83
  84
  85 //
  86 // LLVM bitcode file
  87 //
  88 class File : public ld::relocatable::File
  89 {
  90 public:
  91                                                                                         File(const char* path, time_t mTime, ld::File::Ordinal ordinal,
  92                                                                                                          const uint8_t* content, uint32_t contentLength, cpu_type_t arch);
  93         virtual                                                                 ~File();
  94
  95         // overrides of ld::File
  96         virtual bool                                                                            forEachAtom(ld::File::AtomHandler&) const;
  97         virtual bool                                                                            justInTimeforEachAtom(const char* name, ld::File::AtomHandler&) const
  98                                                                                                                                                                         { return false; }
  99         virtual uint32_t                                                                        cpuSubType() const                      { return _cpuSubType; }
 100
 101         // overrides of ld::relocatable::File
 102         virtual DebugInfoKind                                                           debugInfo()     const                   { return _debugInfo; }
 103         virtual const char*                                                                     debugInfoPath() const           { return _debugInfoPath; }
 104         virtual time_t                                                                          debugInfoModificationTime() const
 105                                                                                                                                                                         { return _debugInfoModTime; }
 106         virtual const std::vector<ld::relocatable::File::Stab>* stabs() const                   { return NULL; }
 107         virtual bool                                                                            canScatterAtoms() const         { return true; }
 108         virtual LinkerOptionsList*                                                      linkerOptions() const           { return NULL; }
 109
 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 const char*                                     translationUnitSource() const
 154                                                                                                                         { return (_compiledAtom ? _compiledAtom->translationUnitSource() : NULL); }
 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, ld::File::Ordinal ordinal, cpu_type_t architecture, cpu_subtype_t subarch,
 202                                                                                         bool logAllFiles, bool verboseOptimizationHints);
 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 #if LTO_API_VERSION >= 7
 217         static void ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t, const char*, void*);
 218 #endif
 219
 220         typedef std::unordered_set<const char*, ld::CStringHash, ld::CStringEquals>  CStringSet;
 221         typedef std::unordered_map<const char*, Atom*, ld::CStringHash, ld::CStringEquals> CStringToAtom;
 222
 223         class AtomSyncer : public ld::File::AtomHandler {
 224         public:
 225                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 226                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 227                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 228                 virtual void            doAtom(const class ld::Atom&);
 229                 virtual void            doFile(const class ld::File&) { }
 230
 231                 const OptimizeOptions&                  _options;
 232                 std::vector<const char*>&               _additionalUndefines;
 233                 std::vector<const ld::Atom*>&   _newAtoms;
 234                 CStringToAtom                                   _llvmAtoms;
 235                 CStringToAtom                                   _deadllvmAtoms;
 236         };
 237
 238         static std::vector<File*>               _s_files;
 239 };
 240
 241 std::vector<File*> Parser::_s_files;
 242
 243
 244 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 245 {
 246         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 247                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 248                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 249                         if ( !result ) {
 250                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 251                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 252                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 253                                 }
 254                         }
 255                         return result;
 256                 }
 257         }
 258         return false;
 259 }
 260
 261 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 262 {
 263         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 264                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 265                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 266                         if ( arch == t->cpuType ) {
 267                                  if ( t->isSubType ) {
 268                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 269                                                 return t->archName;
 270                                 }
 271                                 else {
 272                                         return t->archName;
 273                                 }
 274                         }
 275                 }
 276                 return "unknown bitcode architecture";
 277         }
 278         return NULL;
 279 }
 280
 281 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime, ld::File::Ordinal ordinal,
 282                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles, bool verboseOptimizationHints)
 283 {
 284         File* f = new File(path, modTime, ordinal, fileContent, fileLength, architecture);
 285         _s_files.push_back(f);
 286         if ( logAllFiles )
 287                 printf("%s\n", path);
 288         return f;
 289 }
 290
 291
 292 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 293 {
 294         mach_o::relocatable::ParserOptions objOpts;
 295         objOpts.architecture            = options.arch;
 296         objOpts.objSubtypeMustMatch = false;
 297         objOpts.logAllFiles                     = false;
 298         objOpts.warnUnwindConversionProblems    = options.needsUnwindInfoSection;
 299         objOpts.keepDwarfUnwind         = options.keepDwarfUnwind;
 300         objOpts.forceDwarfConversion = false;
 301         objOpts.neverConvertDwarf   = false;
 302         objOpts.verboseOptimizationHints = options.verboseOptimizationHints;
 303         objOpts.subType                         = 0;
 304
 305         // mach-o parsing is done in-memory, but need path for debug notes
 306         const char* path = "/tmp/lto.o";
 307         time_t modTime = 0;
 308         if ( options.tmpObjectFilePath != NULL ) {
 309                 path = options.tmpObjectFilePath;
 310                 struct stat statBuffer;
 311                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 312                         modTime = statBuffer.st_mtime;
 313         }
 314
 315         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 316         if ( result != NULL )
 317                 return result;
 318         throw "LLVM LTO, file is not of required architecture";
 319 }
 320
 321
 322
 323 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 324         : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
 325         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 326         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 327         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 328         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 329 {
 330         const bool log = false;
 331
 332         // create llvm module
 333         _module = ::lto_module_create_from_memory(content, contentLength);
 334     if ( _module == NULL )
 335                 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth, ::lto_get_error_message(), ::lto_get_version());
 336
 337         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 338
 339         // create atom for each global symbol in module
 340         uint32_t count = ::lto_module_get_num_symbols(_module);
 341         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 342         for (uint32_t i=0; i < count; ++i) {
 343                 const char* name = ::lto_module_get_symbol_name(_module, i);
 344                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 345
 346                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 347                 // ignore dtrace static probes for now
 348                 // later when codegen is done and a mach-o file is produces the probes will be processed
 349                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 350                         continue;
 351
 352                 ld::Atom::Definition def;
 353                 ld::Atom::Combine combine = ld::Atom::combineNever;
 354                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 355                         case LTO_SYMBOL_DEFINITION_REGULAR:
 356                                 def = ld::Atom::definitionRegular;
 357                                 break;
 358                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 359                                 def = ld::Atom::definitionTentative;
 360                                 break;
 361                         case LTO_SYMBOL_DEFINITION_WEAK:
 362                                 def = ld::Atom::definitionRegular;
 363                                 combine = ld::Atom::combineByName;
 364                                 break;
 365                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 366                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 367                                 def = ld::Atom::definitionProxy;
 368                                 break;
 369                         default:
 370                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 371                 }
 372
 373                 // make LLVM atoms for definitions and a reference for undefines
 374                 if ( def != ld::Atom::definitionProxy ) {
 375                         ld::Atom::Scope scope;
 376                         bool autohide = false;
 377                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 378                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 379                                         scope = ld::Atom::scopeTranslationUnit;
 380                                         break;
 381                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 382                                         scope = ld::Atom::scopeLinkageUnit;
 383                                         break;
 384                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 385                                         scope = ld::Atom::scopeGlobal;
 386                                         break;
 387 #if LTO_API_VERSION >= 4
 388                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 389                                         scope = ld::Atom::scopeGlobal;
 390                                         autohide = true;
 391                                         break;
 392 #endif
 393                                 default:
 394                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 395                         }
 396                         // only make atoms for non-internal symbols
 397                         if ( scope == ld::Atom::scopeTranslationUnit )
 398                                 continue;
 399                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 400                         // make Atom using placement new operator
 401                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 402                         if ( scope != ld::Atom::scopeTranslationUnit )
 403                                 _internalAtom.addReference(name);
 404                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 405                 }
 406                 else {
 407                         // add to list of external references
 408                         _internalAtom.addReference(name);
 409                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 410                 }
 411         }
 412 }
 413
 414 File::~File()
 415 {
 416         if ( _module != NULL )
 417                 ::lto_module_dispose(_module);
 418 }
 419
 420 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 421 {
 422         handler.doAtom(_internalAtom);
 423         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 424                 handler.doAtom(_atomArray[i]);
 425         }
 426         return true;
 427 }
 428
 429 InternalAtom::InternalAtom(File& f)
 430         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 431                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 432                 _file(f)
 433 {
 434 }
 435
 436 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 437                         ld::Atom::Alignment a, bool ah)
 438         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 439                                 ld::Atom::symbolTableIn, false, false, false, a),
 440                 _file(f), _name(nm), _compiledAtom(NULL)
 441 {
 442         if ( ah )
 443                 this->setAutoHide();
 444 }
 445
 446 void Atom::setCompiledAtom(const ld::Atom& atom)
 447 {
 448         // set delegate so virtual methods go to it
 449         _compiledAtom = &atom;
 450
 451         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 452
 453         // update fields in ld::Atom to match newly constructed mach-o atom
 454         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 455 }
 456
 457
 458
 459 // <rdar://problem/12379604> The order that files are merged must match command line order
 460 struct CommandLineOrderFileSorter
 461 {
 462      bool operator()(File* left, File* right)
 463      {
 464         return ( left->ordinal() < right->ordinal() );
 465      }
 466 };
 467
 468
 469 #if LTO_API_VERSION >= 7
 470 void Parser::ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t severity, const char* message, void*)
 471 {
 472         switch ( severity ) {
 473                 case LTO_DS_NOTE:
 474                 case LTO_DS_WARNING:
 475                         warning("%s", message);
 476                         break;
 477                 case LTO_DS_ERROR:
 478                         throwf("%s", message);
 479         }
 480 }
 481 #endif
 482
 483 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 484                                                 ld::Internal&                                           state,
 485                                                 const OptimizeOptions&                          options,
 486                                                 ld::File::AtomHandler&                          handler,
 487                                                 std::vector<const ld::Atom*>&           newAtoms,
 488                                                 std::vector<const char*>&                       additionalUndefines)
 489 {
 490         const bool logMustPreserve = false;
 491         const bool logExtraOptions = false;
 492         const bool logBitcodeFiles = false;
 493         const bool logAtomsBeforeSync = false;
 494
 495         // exit quickly if nothing to do
 496         if ( _s_files.size() == 0 )
 497                 return false;
 498
 499         // print out LTO version string if -v was used
 500         if ( options.verbose )
 501                 fprintf(stderr, "%s\n", ::lto_get_version());
 502
 503         // create optimizer and add each Reader
 504         lto_code_gen_t generator = ::lto_codegen_create();
 505 #if LTO_API_VERSION >= 7
 506         lto_codegen_set_diagnostic_handler(generator, ltoDiagnosticHandler, NULL);
 507 #endif
 508
 509         // <rdar://problem/12379604> The order that files are merged must match command line order
 510         std::sort(_s_files.begin(), _s_files.end(), CommandLineOrderFileSorter());
 511         ld::File::Ordinal lastOrdinal;
 512         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 513                 File* f = *it;
 514                 assert(f->ordinal() > lastOrdinal);
 515                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", f->path());
 516                 if ( ::lto_codegen_add_module(generator, f->module()) )
 517                         throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", f->path(), ::lto_get_error_message(), ::lto_get_version());
 518                 lastOrdinal = f->ordinal();
 519         }
 520
 521         // add any -mllvm command line options
 522         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 523                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 524                 ::lto_codegen_debug_options(generator, *it);
 525         }
 526
 527         // <rdar://problem/13687397> Need a way for LTO to get cpu variants (until that info is in bitcode)
 528         if ( options.mcpu != NULL )
 529                 ::lto_codegen_set_cpu(generator, options.mcpu);
 530
 531         // The atom graph uses directed edges (references). Collect all references where
 532         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 533         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 534         // defined in llvm bitcode file.
 535         CStringSet nonLLVMRefs;
 536         CStringToAtom llvmAtoms;
 537     bool hasNonllvmAtoms = false;
 538         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 539                 const ld::Atom* atom = *it;
 540                 // only look at references that come from an atom that is not an llvm atom
 541                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 542                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 543                                 hasNonllvmAtoms = true;
 544                         }
 545                         const ld::Atom* target;
 546                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 547                                 switch ( fit->binding ) {
 548                                         case ld::Fixup::bindingDirectlyBound:
 549                                                 // that reference an llvm atom
 550                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 551                                                         nonLLVMRefs.insert(fit->u.target->name());
 552                                                 break;
 553                                         case ld::Fixup::bindingsIndirectlyBound:
 554                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 555                                                 if ( (target != NULL) && (target->contentType() == ld::Atom::typeLTOtemporary) )
 556                                                         nonLLVMRefs.insert(target->name());
 557                                         default:
 558                                                 break;
 559                                 }
 560                         }
 561                 }
 562                 else {
 563                         llvmAtoms[atom->name()] = (Atom*)atom;
 564                 }
 565         }
 566         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 567         if ( state.entryPoint!= NULL ) {
 568                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 569                         nonLLVMRefs.insert(state.entryPoint->name());
 570         }
 571
 572         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 573         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 574         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 575         // atom so that the linker can replace it with the mach-o one later.
 576         CStringToAtom deadllvmAtoms;
 577         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 578                 const ld::Atom* atom = *it;
 579                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 580                         const char* name = atom->name();
 581                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 582                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 583                         deadllvmAtoms[name] = (Atom*)atom;
 584                 }
 585         }
 586         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 587                 File* file = *it;
 588                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 589                         Atom* llvmAtom = &file->_atomArray[i];
 590                         if ( llvmAtom->coalescedAway()  ) {
 591                                 const char* name = llvmAtom->name();
 592                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 593                                         if ( logMustPreserve )
 594                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 595                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 596                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 597                                 }
 598                         }
 599                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 600                                 const char* name = llvmAtom->name();
 601                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 602                         }
 603                 }
 604         }
 605
 606         // tell code generator about symbols that must be preserved
 607         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 608                 const char* name = it->first;
 609                 Atom* atom = it->second;
 610                 // Include llvm Symbol in export list if it meets one of following two conditions
 611                 // 1 - atom scope is global (and not linkage unit).
 612                 // 2 - included in nonLLVMRefs set.
 613                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 614                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 615                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 616                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 617                 }
 618                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 619                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 620                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 621                 }
 622         }
 623
 624     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 625     if ( options.relocatable && !hasNonllvmAtoms ) {
 626                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 627                         // HACK, no good way to tell linker we are all done, so just quit
 628                         exit(0);
 629                 }
 630                 warning("could not produce merged bitcode file");
 631     }
 632
 633         // set code-gen model
 634         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 635         if ( options.mainExecutable ) {
 636                 if ( options.staticExecutable ) {
 637                         // x86_64 "static" or any "-static -pie" is really dynamic code model
 638                         if ( (options.arch == CPU_TYPE_X86_64) || options.pie )
 639                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 640                         else
 641                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 642                 }
 643                 else {
 644                         if ( options.pie )
 645                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 646                         else
 647                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 648                 }
 649         }
 650         else {
 651                 if ( options.allowTextRelocs )
 652                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 653                 else
 654                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 655         }
 656         if ( ::lto_codegen_set_pic_model(generator, model) )
 657                 throwf("could not create set codegen model: %s", lto_get_error_message());
 658
 659     // if requested, save off merged bitcode file
 660     if ( options.saveTemps ) {
 661         char tempBitcodePath[MAXPATHLEN];
 662         strcpy(tempBitcodePath, options.outputFilePath);
 663         strcat(tempBitcodePath, ".lto.bc");
 664         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 665     }
 666
 667 #if LTO_API_VERSION >= 3
 668         // find assembler next to linker
 669         char path[PATH_MAX];
 670         uint32_t bufSize = PATH_MAX;
 671         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 672                 char* lastSlash = strrchr(path, '/');
 673                 if ( lastSlash != NULL ) {
 674                         strcpy(lastSlash+1, "as");
 675                         struct stat statInfo;
 676                         if ( stat(path, &statInfo) == 0 )
 677                                 ::lto_codegen_set_assembler_path(generator, path);
 678                 }
 679         }
 680 #endif
 681         // run code generator
 682         size_t machOFileLen;
 683         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 684         if ( machOFile == NULL )
 685                 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
 686
 687     // if requested, save off temp mach-o file
 688     if ( options.saveTemps ) {
 689         char tempMachoPath[MAXPATHLEN];
 690         strcpy(tempMachoPath, options.outputFilePath);
 691         strcat(tempMachoPath, ".lto.o");
 692         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 693                 if ( fd != -1) {
 694                         ::write(fd, machOFile, machOFileLen);
 695                         ::close(fd);
 696                 }
 697                 //      save off merged bitcode file
 698                 char tempOptBitcodePath[MAXPATHLEN];
 699         strcpy(tempOptBitcodePath, options.outputFilePath);
 700         strcat(tempOptBitcodePath, ".lto.opt.bc");
 701         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 702         }
 703
 704         // if needed, save temp mach-o file to specific location
 705         if ( options.tmpObjectFilePath != NULL ) {
 706                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 707                 if ( fd != -1) {
 708                         ::write(fd, machOFile, machOFileLen);
 709                         ::close(fd);
 710                 }
 711                 else {
 712                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 713                 }
 714         }
 715
 716         // parse generated mach-o file into a MachOReader
 717         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 718
 719         // sync generated mach-o atoms with existing atoms ld knows about
 720         if ( logAtomsBeforeSync ) {
 721                 fprintf(stderr, "llvmAtoms:\n");
 722                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 723                         const char* name = it->first;
 724                         //Atom* atom = it->second;
 725                         fprintf(stderr, "\t%s\n", name);
 726                 }
 727                 fprintf(stderr, "deadllvmAtoms:\n");
 728                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 729                         const char* name = it->first;
 730                         //Atom* atom = it->second;
 731                         fprintf(stderr, "\t%s\n", name);
 732                 }
 733         }
 734         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 735         machoFile->forEachAtom(syncer);
 736
 737         // Remove InternalAtoms from ld
 738         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 739                 (*it)->internalAtom().setCoalescedAway();
 740         }
 741         // Remove Atoms from ld if code generator optimized them away
 742         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 743                 // check if setRealAtom() called on this Atom
 744                 if ( li->second->compiledAtom() == NULL ) {
 745                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 746                         li->second->setCoalescedAway();
 747                 }
 748         }
 749
 750         // notify about file level attributes
 751         handler.doFile(*machoFile);
 752
 753         // if final mach-o file has debug info, update original bitcode files to match
 754         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 755                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 756                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 757         }
 758
 759         return true;
 760 }
 761
 762
 763 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 764 {
 765         static const ld::Atom* lastProxiedAtom = NULL;
 766         static const ld::File* lastProxiedFile = NULL;
 767         // update proxy atoms to point to real atoms and find new atoms
 768         const char* name = machoAtom.name();
 769         if ( machoAtom.scope() >= ld::Atom::scopeLinkageUnit ) {
 770                 CStringToAtom::iterator pos = _llvmAtoms.find(name);
 771                 if ( pos != _llvmAtoms.end() ) {
 772                         // turn Atom into a proxy for this mach-o atom
 773                         pos->second->setCompiledAtom(machoAtom);
 774                         lastProxiedAtom = &machoAtom;
 775                         lastProxiedFile = pos->second->file();
 776                 }
 777                 else {
 778                         // an atom of this name was not in the allAtoms list the linker gave us
 779                         if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 780                                 // this corresponding to an atom that the linker coalesced away or marked not-live
 781                                 if ( _options.linkerDeadStripping ) {
 782                                         // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 783                                         Atom* llvmAtom = _deadllvmAtoms[name];
 784                                         llvmAtom->setCompiledAtom(machoAtom);
 785                                         _newAtoms.push_back(&machoAtom);
 786                                 }
 787                                 else {
 788                                         // Don't pass it back as a new atom
 789                                 }
 790                         }
 791                         else
 792                         {
 793                                 // this is something new that lto conjured up, tell ld its new
 794                                 _newAtoms.push_back(&machoAtom);
 795                         }
 796                 }
 797         }
 798         else {
 799                 // ld only knew about non-static atoms, so this one must be new
 800                 _newAtoms.push_back(&machoAtom);
 801                 // <rdar://problem/15469363> if new static atom in same section as previous non-static atom, assign to same file as previous
 802                 if ( (lastProxiedAtom != NULL) && (lastProxiedAtom->section() == machoAtom.section()) ) {
 803                         ld::Atom* ma = const_cast<ld::Atom*>(&machoAtom);
 804                         ma->setFile(lastProxiedFile);
 805                 }
 806         }
 807
 808         // adjust fixups to go through proxy atoms
 809         //fprintf(stderr, "adjusting fixups in atom: %s\n", machoAtom.name());
 810         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 811                 switch ( fit->binding ) {
 812                         case ld::Fixup::bindingNone:
 813                                 break;
 814                         case ld::Fixup::bindingByNameUnbound:
 815                                 // don't know if this target has been seen by linker before or if it is new
 816                                 // be conservative and tell linker it is new
 817                                 _additionalUndefines.push_back(fit->u.name);
 818                                 //fprintf(stderr, "    by name ref to: %s\n", fit->u.name);
 819                                 break;
 820                         case ld::Fixup::bindingDirectlyBound:
 821                                 // If mach-o atom is referencing another mach-o atom then
 822                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 823                                 // llvm symbol references always go through Atom proxy.
 824                                 if (  fit->u.target->scope() != ld::Atom::scopeTranslationUnit ) {
 825                                         const char* targetName = fit->u.target->name();
 826                                         CStringToAtom::iterator pos = _llvmAtoms.find(targetName);
 827                                         if ( pos != _llvmAtoms.end() ) {
 828                                                 fit->u.target = pos->second;
 829                                         }
 830                                         else {
 831                                                 // <rdar://problem/12859831> Don't unbind follow-on reference into by-name reference
 832                                                 if ( (_deadllvmAtoms.find(targetName) != _deadllvmAtoms.end()) && (fit->kind != ld::Fixup::kindNoneFollowOn) ) {
 833                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 834                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 835                                                         fit->u.name = targetName;
 836                                                 }
 837                                         }
 838                                 }
 839                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 840                                 break;
 841                         case ld::Fixup::bindingByContentBound:
 842                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 843                                 break;
 844                         case ld::Fixup::bindingsIndirectlyBound:
 845                                 assert(0 && "indirect binding found in initial mach-o file?");
 846                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 847                                 break;
 848                 }
 849         }
 850
 851 }
 852
 853 class Mutex {
 854         static pthread_mutex_t lto_lock;
 855 public:
 856         Mutex() { pthread_mutex_lock(&lto_lock); }
 857         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 858 };
 859 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 860
 861 //
 862 // Used by archive reader to see if member is an llvm bitcode file
 863 //
 864 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 865 {
 866         Mutex lock;
 867         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 868 }
 869
 870
 871 //
 872 // main function used by linker to instantiate ld::Files
 873 //
 874 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 875                                                                 const char* path, time_t modTime, ld::File::Ordinal ordinal,
 876                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles,
 877                                                                 bool verboseOptimizationHints)
 878 {
 879         Mutex lock;
 880         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 881                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles, verboseOptimizationHints);
 882         else
 883                 return NULL;
 884 }
 885
 886 //
 887 // used by "ld -v" to report version of libLTO.dylib being used
 888 //
 889 const char* version()
 890 {
 891         Mutex lock;
 892         return ::lto_get_version();
 893 }
 894
 895
 896 //
 897 // used by ld for error reporting
 898 //
 899 bool libLTOisLoaded()
 900 {
 901         Mutex lock;
 902         return (::lto_get_version() != NULL);
 903 }
 904
 905 //
 906 // used by ld for error reporting
 907 //
 908 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 909 {
 910         Mutex lock;
 911         return Parser::fileKind(fileContent, fileLength);
 912 }
 913
 914 //
 915 // used by ld for doing link time optimization
 916 //
 917 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 918                                 ld::Internal&                                           state,
 919                                 const OptimizeOptions&                          options,
 920                                 ld::File::AtomHandler&                          handler,
 921                                 std::vector<const ld::Atom*>&           newAtoms,
 922                                 std::vector<const char*>&                       additionalUndefines)
 923 {
 924         Mutex lock;
 925         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 926 }
 927
 928
 929
 930 }; // namespace lto
 931
 932
 933 #endif
 934