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, strdup(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         void                                                                                            release();
 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,
 203                                                                                         bool logAllFiles, bool verboseOptimizationHints);
 204         static bool                                             libLTOisLoaded() { return (::lto_get_version() != NULL); }
 205         static bool                                             optimize(   const std::vector<const ld::Atom*>& allAtoms,
 206                                                                                                 ld::Internal&                                           state,
 207                                                                                                 const OptimizeOptions&                          options,
 208                                                                                                 ld::File::AtomHandler&                          handler,
 209                                                                                                 std::vector<const ld::Atom*>&           newAtoms,
 210                                                                                                 std::vector<const char*>&                       additionalUndefines);
 211
 212         static const char*                      ltoVersion()    { return ::lto_get_version(); }
 213
 214 private:
 215         static const char*                              tripletPrefixForArch(cpu_type_t arch);
 216         static ld::relocatable::File*   parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options);
 217 #if LTO_API_VERSION >= 7
 218         static void ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t, const char*, void*);
 219 #endif
 220
 221         typedef std::unordered_set<const char*, ld::CStringHash, ld::CStringEquals>  CStringSet;
 222         typedef std::unordered_map<const char*, Atom*, ld::CStringHash, ld::CStringEquals> CStringToAtom;
 223
 224         class AtomSyncer : public ld::File::AtomHandler {
 225         public:
 226                                                         AtomSyncer(std::vector<const char*>& a, std::vector<const ld::Atom*>&na,
 227                                                                                 CStringToAtom la, CStringToAtom dla, const OptimizeOptions& options) :
 228                                                                                 _options(options), _additionalUndefines(a), _newAtoms(na), _llvmAtoms(la), _deadllvmAtoms(dla) { }
 229                 virtual void            doAtom(const class ld::Atom&);
 230                 virtual void            doFile(const class ld::File&) { }
 231
 232
 233                 const OptimizeOptions&                  _options;
 234                 std::vector<const char*>&               _additionalUndefines;
 235                 std::vector<const ld::Atom*>&   _newAtoms;
 236                 CStringToAtom                                   _llvmAtoms;
 237                 CStringToAtom                                   _deadllvmAtoms;
 238         };
 239
 240         static std::vector<File*>               _s_files;
 241 };
 242
 243 std::vector<File*> Parser::_s_files;
 244
 245
 246 bool Parser::validFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 247 {
 248         for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 249                 if ( (architecture == t->cpuType) && (!(t->isSubType) || (subarch == t->cpuSubType)) ) {
 250                         bool result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefix);
 251                         if ( !result ) {
 252                                 // <rdar://problem/8434487> LTO only supports thumbv7 not armv7
 253                                 if ( t->llvmTriplePrefixAlt[0] != '\0' ) {
 254                                         result = ::lto_module_is_object_file_in_memory_for_target(fileContent, fileLength, t->llvmTriplePrefixAlt);
 255                                 }
 256                         }
 257                         return result;
 258                 }
 259         }
 260         return false;
 261 }
 262
 263 const char* Parser::fileKind(const uint8_t* p, uint64_t fileLength)
 264 {
 265         if ( (p[0] == 0xDE) && (p[1] == 0xC0) && (p[2] == 0x17) && (p[3] == 0x0B) ) {
 266                 cpu_type_t arch = LittleEndian::get32(*((uint32_t*)(&p[16])));
 267                 for (const ArchInfo* t=archInfoArray; t->archName != NULL; ++t) {
 268                         if ( arch == t->cpuType ) {
 269                                  if ( t->isSubType ) {
 270                                         if ( ::lto_module_is_object_file_in_memory_for_target(p, fileLength, t->llvmTriplePrefix) )
 271                                                 return t->archName;
 272                                 }
 273                                 else {
 274                                         return t->archName;
 275                                 }
 276                         }
 277                 }
 278                 return "unknown bitcode architecture";
 279         }
 280         return NULL;
 281 }
 282
 283 File* Parser::parse(const uint8_t* fileContent, uint64_t fileLength, const char* path, time_t modTime, ld::File::Ordinal ordinal,
 284                                                                                                         cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles, bool verboseOptimizationHints)
 285 {
 286         File* f = new File(path, modTime, ordinal, fileContent, fileLength, architecture);
 287         _s_files.push_back(f);
 288         if ( logAllFiles )
 289                 printf("%s\n", path);
 290         return f;
 291 }
 292
 293
 294 ld::relocatable::File* Parser::parseMachOFile(const uint8_t* p, size_t len, const OptimizeOptions& options)
 295 {
 296         mach_o::relocatable::ParserOptions objOpts;
 297         objOpts.architecture            = options.arch;
 298         objOpts.objSubtypeMustMatch = false;
 299         objOpts.logAllFiles                     = false;
 300         objOpts.warnUnwindConversionProblems    = options.needsUnwindInfoSection;
 301         objOpts.keepDwarfUnwind         = options.keepDwarfUnwind;
 302         objOpts.forceDwarfConversion = false;
 303         objOpts.neverConvertDwarf   = false;
 304         objOpts.verboseOptimizationHints = options.verboseOptimizationHints;
 305         objOpts.armUsesZeroCostExceptions = options.armUsesZeroCostExceptions;
 306
 307         objOpts.subType                         = 0;
 308
 309         // mach-o parsing is done in-memory, but need path for debug notes
 310         const char* path = "/tmp/lto.o";
 311         time_t modTime = 0;
 312         if ( options.tmpObjectFilePath != NULL ) {
 313                 path = options.tmpObjectFilePath;
 314                 struct stat statBuffer;
 315                 if ( stat(options.tmpObjectFilePath, &statBuffer) == 0 )
 316                         modTime = statBuffer.st_mtime;
 317         }
 318
 319         ld::relocatable::File* result = mach_o::relocatable::parse(p, len, path, modTime, ld::File::Ordinal::LTOOrdinal(), objOpts);
 320         if ( result != NULL )
 321                 return result;
 322         throw "LLVM LTO, file is not of required architecture";
 323 }
 324
 325
 326
 327 File::File(const char* pth, time_t mTime, ld::File::Ordinal ordinal, const uint8_t* content, uint32_t contentLength, cpu_type_t arch)
 328         : ld::relocatable::File(pth,mTime,ordinal), _architecture(arch), _internalAtom(*this),
 329         _atomArray(NULL), _atomArrayCount(0), _module(NULL), _debugInfoPath(pth),
 330         _section("__TEXT_", "__tmp_lto", ld::Section::typeTempLTO),
 331         _fixupToInternal(0, ld::Fixup::k1of1, ld::Fixup::kindNone, &_internalAtom),
 332         _debugInfo(ld::relocatable::File::kDebugInfoNone), _cpuSubType(0)
 333 {
 334         const bool log = false;
 335
 336         // create llvm module
 337 #if LTO_API_VERSION >= 9
 338         _module = ::lto_module_create_from_memory_with_path(content, contentLength, pth);
 339         if ( _module == NULL )
 340 #endif
 341         _module = ::lto_module_create_from_memory(content, contentLength);
 342     if ( _module == NULL )
 343                 throwf("could not parse object file %s: '%s', using libLTO version '%s'", pth, ::lto_get_error_message(), ::lto_get_version());
 344
 345         if ( log ) fprintf(stderr, "bitcode file: %s\n", pth);
 346
 347         // create atom for each global symbol in module
 348         uint32_t count = ::lto_module_get_num_symbols(_module);
 349         _atomArray = (Atom*)malloc(sizeof(Atom)*count);
 350         for (uint32_t i=0; i < count; ++i) {
 351                 const char* name = ::lto_module_get_symbol_name(_module, i);
 352                 lto_symbol_attributes attr = lto_module_get_symbol_attribute(_module, i);
 353
 354                 // <rdar://problem/6378110> LTO doesn't like dtrace symbols
 355                 // ignore dtrace static probes for now
 356                 // later when codegen is done and a mach-o file is produces the probes will be processed
 357                 if ( (strncmp(name, "___dtrace_probe$", 16) == 0) || (strncmp(name, "___dtrace_isenabled$", 20) == 0) )
 358                         continue;
 359
 360                 ld::Atom::Definition def;
 361                 ld::Atom::Combine combine = ld::Atom::combineNever;
 362                 switch ( attr & LTO_SYMBOL_DEFINITION_MASK ) {
 363                         case LTO_SYMBOL_DEFINITION_REGULAR:
 364                                 def = ld::Atom::definitionRegular;
 365                                 break;
 366                         case LTO_SYMBOL_DEFINITION_TENTATIVE:
 367                                 def = ld::Atom::definitionTentative;
 368                                 break;
 369                         case LTO_SYMBOL_DEFINITION_WEAK:
 370                                 def = ld::Atom::definitionRegular;
 371                                 combine = ld::Atom::combineByName;
 372                                 break;
 373                         case LTO_SYMBOL_DEFINITION_UNDEFINED:
 374                         case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
 375                                 def = ld::Atom::definitionProxy;
 376                                 break;
 377                         default:
 378                                 throwf("unknown definition kind for symbol %s in bitcode file %s", name, pth);
 379                 }
 380
 381                 // make LLVM atoms for definitions and a reference for undefines
 382                 if ( def != ld::Atom::definitionProxy ) {
 383                         ld::Atom::Scope scope;
 384                         bool autohide = false;
 385                         switch ( attr & LTO_SYMBOL_SCOPE_MASK) {
 386                                 case LTO_SYMBOL_SCOPE_INTERNAL:
 387                                         scope = ld::Atom::scopeTranslationUnit;
 388                                         break;
 389                                 case LTO_SYMBOL_SCOPE_HIDDEN:
 390                                         scope = ld::Atom::scopeLinkageUnit;
 391                                         break;
 392                                 case LTO_SYMBOL_SCOPE_DEFAULT:
 393                                         scope = ld::Atom::scopeGlobal;
 394                                         break;
 395 #if LTO_API_VERSION >= 4
 396                                 case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
 397                                         scope = ld::Atom::scopeGlobal;
 398                                         autohide = true;
 399                                         break;
 400 #endif
 401                                 default:
 402                                         throwf("unknown scope for symbol %s in bitcode file %s", name, pth);
 403                         }
 404                         // only make atoms for non-internal symbols
 405                         if ( scope == ld::Atom::scopeTranslationUnit )
 406                                 continue;
 407                         uint8_t alignment = (attr & LTO_SYMBOL_ALIGNMENT_MASK);
 408                         // make Atom using placement new operator
 409                         new (&_atomArray[_atomArrayCount++]) Atom(*this, name, scope, def, combine, alignment, autohide);
 410                         if ( scope != ld::Atom::scopeTranslationUnit )
 411                                 _internalAtom.addReference(name);
 412                         if ( log ) fprintf(stderr, "\t0x%08X %s\n", attr, name);
 413                 }
 414                 else {
 415                         // add to list of external references
 416                         _internalAtom.addReference(name);
 417                         if ( log ) fprintf(stderr, "\t%s (undefined)\n", name);
 418                 }
 419         }
 420 }
 421
 422 File::~File()
 423 {
 424         this->release();
 425 }
 426
 427 void File::release()
 428 {
 429         if ( _module != NULL )
 430                 ::lto_module_dispose(_module);
 431         _module = NULL;
 432 }
 433
 434 bool File::forEachAtom(ld::File::AtomHandler& handler) const
 435 {
 436         handler.doAtom(_internalAtom);
 437         for(uint32_t i=0; i < _atomArrayCount; ++i) {
 438                 handler.doAtom(_atomArray[i]);
 439         }
 440         return true;
 441 }
 442
 443 InternalAtom::InternalAtom(File& f)
 444         : ld::Atom(f._section, ld::Atom::definitionRegular, ld::Atom::combineNever, ld::Atom::scopeTranslationUnit,
 445                                 ld::Atom::typeLTOtemporary, ld::Atom::symbolTableNotIn, true, false, false, ld::Atom::Alignment(0)),
 446                 _file(f)
 447 {
 448 }
 449
 450 Atom::Atom(File& f, const char* nm, ld::Atom::Scope s, ld::Atom::Definition d, ld::Atom::Combine c,
 451                         ld::Atom::Alignment a, bool ah)
 452         : ld::Atom(f._section, d, c, s, ld::Atom::typeLTOtemporary,
 453                                 ld::Atom::symbolTableIn, false, false, false, a),
 454                 _file(f), _name(strdup(nm)), _compiledAtom(NULL)
 455 {
 456         if ( ah )
 457                 this->setAutoHide();
 458 }
 459
 460 void Atom::setCompiledAtom(const ld::Atom& atom)
 461 {
 462         // set delegate so virtual methods go to it
 463         _compiledAtom = &atom;
 464
 465         //fprintf(stderr, "setting lto atom %p to delegate to mach-o atom %p (%s)\n", this, &atom, atom.name());
 466
 467         // update fields in ld::Atom to match newly constructed mach-o atom
 468         (const_cast<Atom*>(this))->setAttributesFromAtom(atom);
 469 }
 470
 471
 472
 473 // <rdar://problem/12379604> The order that files are merged must match command line order
 474 struct CommandLineOrderFileSorter
 475 {
 476      bool operator()(File* left, File* right)
 477      {
 478         return ( left->ordinal() < right->ordinal() );
 479      }
 480 };
 481
 482
 483 #if LTO_API_VERSION >= 7
 484 void Parser::ltoDiagnosticHandler(lto_codegen_diagnostic_severity_t severity, const char* message, void*)
 485 {
 486         switch ( severity ) {
 487 #if LTO_API_VERSION >= 10
 488                 case LTO_DS_REMARK:
 489 #endif
 490                 case LTO_DS_NOTE:
 491                 case LTO_DS_WARNING:
 492                         warning("%s", message);
 493                         break;
 494                 case LTO_DS_ERROR:
 495                         throwf("%s", message);
 496         }
 497 }
 498 #endif
 499
 500 bool Parser::optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 501                                                 ld::Internal&                                           state,
 502                                                 const OptimizeOptions&                          options,
 503                                                 ld::File::AtomHandler&                          handler,
 504                                                 std::vector<const ld::Atom*>&           newAtoms,
 505                                                 std::vector<const char*>&                       additionalUndefines)
 506 {
 507         const bool logMustPreserve = false;
 508         const bool logExtraOptions = false;
 509         const bool logBitcodeFiles = false;
 510         const bool logAtomsBeforeSync = false;
 511
 512         // exit quickly if nothing to do
 513         if ( _s_files.size() == 0 )
 514                 return false;
 515
 516         // print out LTO version string if -v was used
 517         if ( options.verbose )
 518                 fprintf(stderr, "%s\n", ::lto_get_version());
 519
 520         // create optimizer and add each Reader
 521         lto_code_gen_t generator = ::lto_codegen_create();
 522 #if LTO_API_VERSION >= 7
 523         lto_codegen_set_diagnostic_handler(generator, ltoDiagnosticHandler, NULL);
 524 #endif
 525
 526         // <rdar://problem/12379604> The order that files are merged must match command line order
 527         std::sort(_s_files.begin(), _s_files.end(), CommandLineOrderFileSorter());
 528         ld::File::Ordinal lastOrdinal;
 529         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 530                 File* f = *it;
 531                 assert(f->ordinal() > lastOrdinal);
 532                 if ( logBitcodeFiles ) fprintf(stderr, "lto_codegen_add_module(%s)\n", f->path());
 533                 if ( ::lto_codegen_add_module(generator, f->module()) )
 534                         throwf("lto: could not merge in %s because '%s', using libLTO version '%s'", f->path(), ::lto_get_error_message(), ::lto_get_version());
 535                 // <rdar://problem/15471128> linker should release module as soon as possible
 536                 f->release();
 537                 lastOrdinal = f->ordinal();
 538         }
 539
 540         // add any -mllvm command line options
 541         for (std::vector<const char*>::const_iterator it=options.llvmOptions->begin(); it != options.llvmOptions->end(); ++it) {
 542                 if ( logExtraOptions ) fprintf(stderr, "passing option to llvm: %s\n", *it);
 543                 ::lto_codegen_debug_options(generator, *it);
 544         }
 545
 546         // <rdar://problem/13687397> Need a way for LTO to get cpu variants (until that info is in bitcode)
 547         if ( options.mcpu != NULL )
 548                 ::lto_codegen_set_cpu(generator, options.mcpu);
 549
 550         // The atom graph uses directed edges (references). Collect all references where
 551         // originating atom is not part of any LTO Reader. This allows optimizer to optimize an
 552         // external (i.e. not originated from same .o file) reference if all originating atoms are also
 553         // defined in llvm bitcode file.
 554         CStringSet nonLLVMRefs;
 555         CStringToAtom llvmAtoms;
 556     bool hasNonllvmAtoms = false;
 557         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 558                 const ld::Atom* atom = *it;
 559                 // only look at references that come from an atom that is not an llvm atom
 560                 if ( atom->contentType() != ld::Atom::typeLTOtemporary ) {
 561                         if ( (atom->section().type() != ld::Section::typeMachHeader) && (atom->definition() != ld::Atom::definitionProxy) ) {
 562                                 hasNonllvmAtoms = true;
 563                         }
 564                         const ld::Atom* target;
 565                         for (ld::Fixup::iterator fit=atom->fixupsBegin(); fit != atom->fixupsEnd(); ++fit) {
 566                                 switch ( fit->binding ) {
 567                                         case ld::Fixup::bindingDirectlyBound:
 568                                                 // that reference an llvm atom
 569                                                 if ( fit->u.target->contentType() == ld::Atom::typeLTOtemporary )
 570                                                         nonLLVMRefs.insert(fit->u.target->name());
 571                                                 break;
 572                                         case ld::Fixup::bindingsIndirectlyBound:
 573                                                 target = state.indirectBindingTable[fit->u.bindingIndex];
 574                                                 if ( (target != NULL) && (target->contentType() == ld::Atom::typeLTOtemporary) )
 575                                                         nonLLVMRefs.insert(target->name());
 576                                         default:
 577                                                 break;
 578                                 }
 579                         }
 580                 }
 581                 else if ( atom->scope() >= ld::Atom::scopeLinkageUnit ) {
 582                         llvmAtoms[atom->name()] = (Atom*)atom;
 583                 }
 584         }
 585         // if entry point is in a llvm bitcode file, it must be preserved by LTO
 586         if ( state.entryPoint!= NULL ) {
 587                 if ( state.entryPoint->contentType() == ld::Atom::typeLTOtemporary )
 588                         nonLLVMRefs.insert(state.entryPoint->name());
 589         }
 590
 591         // deadAtoms are the atoms that the linker coalesced.  For instance weak or tentative definitions
 592         // overriden by another atom.  If any of these deadAtoms are llvm atoms and they were replaced
 593         // with a mach-o atom, we need to tell the lto engine to preserve (not optimize away) its dead
 594         // atom so that the linker can replace it with the mach-o one later.
 595         CStringToAtom deadllvmAtoms;
 596         for (std::vector<const ld::Atom*>::const_iterator it = allAtoms.begin(); it != allAtoms.end(); ++it) {
 597                 const ld::Atom* atom = *it;
 598                 if ( atom->coalescedAway() && (atom->contentType() == ld::Atom::typeLTOtemporary) ) {
 599                         const char* name = atom->name();
 600                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 601                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 602                         deadllvmAtoms[name] = (Atom*)atom;
 603                 }
 604         }
 605         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 606                 File* file = *it;
 607                 for(uint32_t i=0; i < file->_atomArrayCount; ++i) {
 608                         Atom* llvmAtom = &file->_atomArray[i];
 609                         if ( llvmAtom->coalescedAway()  ) {
 610                                 const char* name = llvmAtom->name();
 611                                 if ( deadllvmAtoms.find(name) == deadllvmAtoms.end() ) {
 612                                         if ( logMustPreserve )
 613                                                 fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because linker coalesce away and replace with a mach-o atom\n", name);
 614                                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 615                                         deadllvmAtoms[name] = (Atom*)llvmAtom;
 616                                 }
 617                         }
 618                         else if ( options.linkerDeadStripping && !llvmAtom->live() ) {
 619                                 const char* name = llvmAtom->name();
 620                                 deadllvmAtoms[name] = (Atom*)llvmAtom;
 621                         }
 622                 }
 623         }
 624
 625         // tell code generator about symbols that must be preserved
 626         for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 627                 const char* name = it->first;
 628                 Atom* atom = it->second;
 629                 // Include llvm Symbol in export list if it meets one of following two conditions
 630                 // 1 - atom scope is global (and not linkage unit).
 631                 // 2 - included in nonLLVMRefs set.
 632                 // If a symbol is not listed in exportList then LTO is free to optimize it away.
 633                 if ( (atom->scope() == ld::Atom::scopeGlobal) && options.preserveAllGlobals ) {
 634                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because global symbol\n", name);
 635                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 636                 }
 637                 else if ( nonLLVMRefs.find(name) != nonLLVMRefs.end() ) {
 638                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because referenced by a mach-o atom\n", name);
 639                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 640                 }
 641                 else if ( options.relocatable && hasNonllvmAtoms ) {
 642                         // <rdar://problem/14334895> ld -r mode but merging in some mach-o files, so need to keep libLTO from optimizing away anything
 643                         if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because -r mode disable LTO dead stripping\n", name);
 644                         ::lto_codegen_add_must_preserve_symbol(generator, name);
 645                 }
 646         }
 647
 648         // <rdar://problem/16165191> tell code generator to preserve initial undefines
 649         for( std::vector<const char*>::const_iterator it=options.initialUndefines->begin(); it != options.initialUndefines->end(); ++it) {
 650                 if ( logMustPreserve ) fprintf(stderr, "lto_codegen_add_must_preserve_symbol(%s) because it is an initial undefine\n", *it);
 651                 ::lto_codegen_add_must_preserve_symbol(generator, *it);
 652         }
 653
 654     // special case running ld -r on all bitcode files to produce another bitcode file (instead of mach-o)
 655     if ( options.relocatable && !hasNonllvmAtoms ) {
 656                 if ( ! ::lto_codegen_write_merged_modules(generator, options.outputFilePath) ) {
 657                         // HACK, no good way to tell linker we are all done, so just quit
 658                         exit(0);
 659                 }
 660                 warning("could not produce merged bitcode file");
 661     }
 662
 663         // set code-gen model
 664         lto_codegen_model model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 665         if ( options.mainExecutable ) {
 666                 if ( options.staticExecutable ) {
 667                         // x86_64 "static" or any "-static -pie" is really dynamic code model
 668                         if ( (options.arch == CPU_TYPE_X86_64) || options.pie )
 669                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 670                         else
 671                                 model = LTO_CODEGEN_PIC_MODEL_STATIC;
 672                 }
 673                 else {
 674                         if ( options.pie )
 675                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 676                         else
 677                                 model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 678                 }
 679         }
 680         else {
 681                 if ( options.allowTextRelocs )
 682                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
 683                 else
 684                         model = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
 685         }
 686         if ( ::lto_codegen_set_pic_model(generator, model) )
 687                 throwf("could not create set codegen model: %s", lto_get_error_message());
 688
 689     // if requested, save off merged bitcode file
 690     if ( options.saveTemps ) {
 691         char tempBitcodePath[MAXPATHLEN];
 692         strcpy(tempBitcodePath, options.outputFilePath);
 693         strcat(tempBitcodePath, ".lto.bc");
 694         ::lto_codegen_write_merged_modules(generator, tempBitcodePath);
 695     }
 696
 697 #if LTO_API_VERSION >= 3
 698         // find assembler next to linker
 699         char path[PATH_MAX];
 700         uint32_t bufSize = PATH_MAX;
 701         if ( _NSGetExecutablePath(path, &bufSize) != -1 ) {
 702                 char* lastSlash = strrchr(path, '/');
 703                 if ( lastSlash != NULL ) {
 704                         strcpy(lastSlash+1, "as");
 705                         struct stat statInfo;
 706                         if ( stat(path, &statInfo) == 0 )
 707                                 ::lto_codegen_set_assembler_path(generator, path);
 708                 }
 709         }
 710 #endif
 711         // run code generator
 712         size_t machOFileLen;
 713         const uint8_t* machOFile = (uint8_t*)::lto_codegen_compile(generator, &machOFileLen);
 714         if ( machOFile == NULL )
 715                 throwf("could not do LTO codegen: '%s', using libLTO version '%s'", ::lto_get_error_message(), ::lto_get_version());
 716
 717     // if requested, save off temp mach-o file
 718     if ( options.saveTemps ) {
 719         char tempMachoPath[MAXPATHLEN];
 720         strcpy(tempMachoPath, options.outputFilePath);
 721         strcat(tempMachoPath, ".lto.o");
 722         int fd = ::open(tempMachoPath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 723                 if ( fd != -1) {
 724                         ::write(fd, machOFile, machOFileLen);
 725                         ::close(fd);
 726                 }
 727                 //      save off merged bitcode file
 728                 char tempOptBitcodePath[MAXPATHLEN];
 729         strcpy(tempOptBitcodePath, options.outputFilePath);
 730         strcat(tempOptBitcodePath, ".lto.opt.bc");
 731         ::lto_codegen_write_merged_modules(generator, tempOptBitcodePath);
 732         }
 733
 734         // if needed, save temp mach-o file to specific location
 735         if ( options.tmpObjectFilePath != NULL ) {
 736                 int fd = ::open(options.tmpObjectFilePath, O_CREAT | O_WRONLY | O_TRUNC, 0666);
 737                 if ( fd != -1) {
 738                         ::write(fd, machOFile, machOFileLen);
 739                         ::close(fd);
 740                 }
 741                 else {
 742                         warning("could not write LTO temp file '%s', errno=%d", options.tmpObjectFilePath, errno);
 743                 }
 744         }
 745
 746         // parse generated mach-o file into a MachOReader
 747         ld::relocatable::File* machoFile = parseMachOFile(machOFile, machOFileLen, options);
 748
 749         // sync generated mach-o atoms with existing atoms ld knows about
 750         if ( logAtomsBeforeSync ) {
 751                 fprintf(stderr, "llvmAtoms:\n");
 752                 for (CStringToAtom::iterator it = llvmAtoms.begin(); it != llvmAtoms.end(); ++it) {
 753                         const char* name = it->first;
 754                         Atom* atom = it->second;
 755                         fprintf(stderr, "\t%p\t%s\n", atom, name);
 756                 }
 757                 fprintf(stderr, "deadllvmAtoms:\n");
 758                 for (CStringToAtom::iterator it = deadllvmAtoms.begin(); it != deadllvmAtoms.end(); ++it) {
 759                         const char* name = it->first;
 760                         Atom* atom = it->second;
 761                         fprintf(stderr, "\t%p\t%s\n", atom, name);
 762                 }
 763         }
 764         AtomSyncer syncer(additionalUndefines, newAtoms, llvmAtoms, deadllvmAtoms, options);
 765         machoFile->forEachAtom(syncer);
 766
 767         // Remove InternalAtoms from ld
 768         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 769                 (*it)->internalAtom().setCoalescedAway();
 770         }
 771         // Remove Atoms from ld if code generator optimized them away
 772         for (CStringToAtom::iterator li = llvmAtoms.begin(), le = llvmAtoms.end(); li != le; ++li) {
 773                 // check if setRealAtom() called on this Atom
 774                 if ( li->second->compiledAtom() == NULL ) {
 775                         //fprintf(stderr, "llvm optimized away %p %s\n", li->second, li->second->name());
 776                         li->second->setCoalescedAway();
 777                 }
 778         }
 779
 780         // notify about file level attributes
 781         handler.doFile(*machoFile);
 782
 783         // if final mach-o file has debug info, update original bitcode files to match
 784         for (std::vector<File*>::iterator it=_s_files.begin(); it != _s_files.end(); ++it) {
 785                 (*it)->setDebugInfo(machoFile->debugInfo(), machoFile->path(),
 786                                                         machoFile->modificationTime(), machoFile->cpuSubType());
 787         }
 788
 789         return true;
 790 }
 791
 792
 793 void Parser::AtomSyncer::doAtom(const ld::Atom& machoAtom)
 794 {
 795         static const bool log = false;
 796         static const ld::Atom* lastProxiedAtom = NULL;
 797         static const ld::File* lastProxiedFile = NULL;
 798         // update proxy atoms to point to real atoms and find new atoms
 799         const char* name = machoAtom.name();
 800         CStringToAtom::iterator pos = _llvmAtoms.find(name);
 801         if ( pos != _llvmAtoms.end() ) {
 802                 // turn Atom into a proxy for this mach-o atom
 803                 pos->second->setCompiledAtom(machoAtom);
 804                 lastProxiedAtom = &machoAtom;
 805                 lastProxiedFile = pos->second->file();
 806                 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p synced to lto atom %p (name=%s)\n", &machoAtom, pos->second, machoAtom.name());
 807         }
 808         else {
 809                 // an atom of this name was not in the allAtoms list the linker gave us
 810                 if ( _deadllvmAtoms.find(name) != _deadllvmAtoms.end() ) {
 811                         // this corresponding to an atom that the linker coalesced away or marked not-live
 812                         if ( _options.linkerDeadStripping ) {
 813                                 // llvm seems to want this atom and -dead_strip is enabled, so it will be deleted if not needed, so add back
 814                                 Atom* llvmAtom = _deadllvmAtoms[name];
 815                                 llvmAtom->setCompiledAtom(machoAtom);
 816                                 _newAtoms.push_back(&machoAtom);
 817                                 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p matches dead lto atom %p but adding back (name=%s)\n", &machoAtom, llvmAtom, machoAtom.name());
 818                         }
 819                         else {
 820                                 // Don't pass it back as a new atom
 821                                 if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p matches dead lto atom %p (name=%s)\n", &machoAtom, _deadllvmAtoms[name], machoAtom.name());
 822                         }
 823                 }
 824                 else
 825                 {
 826                         // this is something new that lto conjured up, tell ld its new
 827                         _newAtoms.push_back(&machoAtom);
 828                         // <rdar://problem/15469363> if new static atom in same section as previous non-static atom, assign to same file as previous
 829                         if ( (lastProxiedAtom != NULL) && (lastProxiedAtom->section() == machoAtom.section()) ) {
 830                                 ld::Atom* ma = const_cast<ld::Atom*>(&machoAtom);
 831                                 ma->setFile(lastProxiedFile);
 832                         }
 833                         if (log) fprintf(stderr, "AtomSyncer, mach-o atom %p is totally new (name=%s)\n", &machoAtom, machoAtom.name());
 834                 }
 835         }
 836
 837         // adjust fixups to go through proxy atoms
 838         if (log) fprintf(stderr, "  adjusting fixups in atom: %s\n", machoAtom.name());
 839         for (ld::Fixup::iterator fit=machoAtom.fixupsBegin(); fit != machoAtom.fixupsEnd(); ++fit) {
 840                 switch ( fit->binding ) {
 841                         case ld::Fixup::bindingNone:
 842                                 break;
 843                         case ld::Fixup::bindingByNameUnbound:
 844                                 // don't know if this target has been seen by linker before or if it is new
 845                                 // be conservative and tell linker it is new
 846                                 _additionalUndefines.push_back(fit->u.name);
 847                                 if (log) fprintf(stderr, "    adding by-name symbol %s\n", fit->u.name);
 848                                 break;
 849                         case ld::Fixup::bindingDirectlyBound:
 850                                 // If mach-o atom is referencing another mach-o atom then
 851                                 // reference is not going through Atom proxy. Fix it here to ensure that all
 852                                 // llvm symbol references always go through Atom proxy.
 853                                 {
 854                                         const char* targetName = fit->u.target->name();
 855                                         CStringToAtom::iterator post = _llvmAtoms.find(targetName);
 856                                         if ( post != _llvmAtoms.end() ) {
 857                                                 const ld::Atom* t = post->second;
 858                                                 if (log) fprintf(stderr, "    updating direct reference to %p to be ref to %p: %s\n", fit->u.target, t, targetName);
 859                                                 fit->u.target = t;
 860                                         }
 861                                         else {
 862                                                 // <rdar://problem/12859831> Don't unbind follow-on reference into by-name reference
 863                                                 if ( (_deadllvmAtoms.find(targetName) != _deadllvmAtoms.end()) && (fit->kind != ld::Fixup::kindNoneFollowOn) ) {
 864                                                         // target was coalesed away and replace by mach-o atom from a non llvm .o file
 865                                                         fit->binding = ld::Fixup::bindingByNameUnbound;
 866                                                         fit->u.name = targetName;
 867                                                 }
 868                                         }
 869                                 }
 870                                 //fprintf(stderr, "    direct ref to: %s (scope=%d)\n", fit->u.target->name(), fit->u.target->scope());
 871                                 break;
 872                         case ld::Fixup::bindingByContentBound:
 873                                 //fprintf(stderr, "    direct by content to: %s\n", fit->u.target->name());
 874                                 break;
 875                         case ld::Fixup::bindingsIndirectlyBound:
 876                                 assert(0 && "indirect binding found in initial mach-o file?");
 877                                 //fprintf(stderr, "    indirect by content to: %u\n", fit->u.bindingIndex);
 878                                 break;
 879                 }
 880         }
 881
 882 }
 883
 884 class Mutex {
 885         static pthread_mutex_t lto_lock;
 886 public:
 887         Mutex() { pthread_mutex_lock(&lto_lock); }
 888         ~Mutex() { pthread_mutex_unlock(&lto_lock); }
 889 };
 890 pthread_mutex_t Mutex::lto_lock = PTHREAD_MUTEX_INITIALIZER;
 891
 892 //
 893 // Used by archive reader to see if member is an llvm bitcode file
 894 //
 895 bool isObjectFile(const uint8_t* fileContent, uint64_t fileLength, cpu_type_t architecture, cpu_subtype_t subarch)
 896 {
 897         Mutex lock;
 898         return Parser::validFile(fileContent, fileLength, architecture, subarch);
 899 }
 900
 901
 902 //
 903 // main function used by linker to instantiate ld::Files
 904 //
 905 ld::relocatable::File* parse(const uint8_t* fileContent, uint64_t fileLength,
 906                                                                 const char* path, time_t modTime, ld::File::Ordinal ordinal,
 907                                                                 cpu_type_t architecture, cpu_subtype_t subarch, bool logAllFiles,
 908                                                                 bool verboseOptimizationHints)
 909 {
 910         Mutex lock;
 911         if ( Parser::validFile(fileContent, fileLength, architecture, subarch) )
 912                 return Parser::parse(fileContent, fileLength, path, modTime, ordinal, architecture, subarch, logAllFiles, verboseOptimizationHints);
 913         else
 914                 return NULL;
 915 }
 916
 917 //
 918 // used by "ld -v" to report version of libLTO.dylib being used
 919 //
 920 const char* version()
 921 {
 922         Mutex lock;
 923         return ::lto_get_version();
 924 }
 925
 926
 927 //
 928 // used by ld for error reporting
 929 //
 930 bool libLTOisLoaded()
 931 {
 932         Mutex lock;
 933         return (::lto_get_version() != NULL);
 934 }
 935
 936 //
 937 // used by ld for error reporting
 938 //
 939 const char* archName(const uint8_t* fileContent, uint64_t fileLength)
 940 {
 941         Mutex lock;
 942         return Parser::fileKind(fileContent, fileLength);
 943 }
 944
 945 //
 946 // used by ld for doing link time optimization
 947 //
 948 bool optimize(  const std::vector<const ld::Atom*>&     allAtoms,
 949                                 ld::Internal&                                           state,
 950                                 const OptimizeOptions&                          options,
 951                                 ld::File::AtomHandler&                          handler,
 952                                 std::vector<const ld::Atom*>&           newAtoms,
 953                                 std::vector<const char*>&                       additionalUndefines)
 954 {
 955         Mutex lock;
 956         return Parser::optimize(allAtoms, state, options, handler, newAtoms, additionalUndefines);
 957 }
 958
 959
 960
 961 }; // namespace lto
 962
 963
 964 #endif
 965