lib/CodeGen/MachO.h

   1 //=== MachO.h - Mach-O structures and constants -----------------*- C++ -*-===//
   2 //
   3 //                     The LLVM Compiler Infrastructure
   4 //
   5 // This file is distributed under the University of Illinois Open Source
   6 // License. See LICENSE.TXT for details.
   7 //
   8 //===----------------------------------------------------------------------===//
   9 //
  10 // This file defines MachO .
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #ifndef MACHO_H
  15 #define MACHO_H
  16
  17 #include "llvm/CodeGen/BinaryObject.h"
  18 #include <string>
  19 #include <vector>
  20
  21 namespace llvm {
  22
  23 class GlobalValue;
  24 class MCAsmInfo;
  25
  26 /// MachOSym - This struct contains information about each symbol that is
  27 /// added to logical symbol table for the module.  This is eventually
  28 /// turned into a real symbol table in the file.
  29 struct MachOSym {
  30   const GlobalValue *GV;    // The global value this corresponds to.
  31   std::string GVName;       // The mangled name of the global value.
  32   uint32_t    n_strx;       // index into the string table
  33   uint8_t     n_type;       // type flag
  34   uint8_t     n_sect;       // section number or NO_SECT
  35   int16_t     n_desc;       // see <mach-o/stab.h>
  36   uint64_t    n_value;      // value for this symbol (or stab offset)
  37
  38   // Constants for the n_sect field
  39   // see <mach-o/nlist.h>
  40   enum { NO_SECT = 0 };   // symbol is not in any section
  41
  42   // Constants for the n_type field
  43   // see <mach-o/nlist.h>
  44   enum { N_UNDF  = 0x0,  // undefined, n_sect == NO_SECT
  45          N_ABS   = 0x2,  // absolute, n_sect == NO_SECT
  46          N_SECT  = 0xe,  // defined in section number n_sect
  47          N_PBUD  = 0xc,  // prebound undefined (defined in a dylib)
  48          N_INDR  = 0xa   // indirect
  49   };
  50   // The following bits are OR'd into the types above. For example, a type
  51   // of 0x0f would be an external N_SECT symbol (0x0e | 0x01).
  52   enum { N_EXT  = 0x01,   // external symbol bit
  53          N_PEXT = 0x10    // private external symbol bit
  54   };
  55
  56   // Constants for the n_desc field
  57   // see <mach-o/loader.h>
  58   enum { REFERENCE_FLAG_UNDEFINED_NON_LAZY          = 0,
  59          REFERENCE_FLAG_UNDEFINED_LAZY              = 1,
  60          REFERENCE_FLAG_DEFINED                     = 2,
  61          REFERENCE_FLAG_PRIVATE_DEFINED             = 3,
  62          REFERENCE_FLAG_PRIVATE_UNDEFINED_NON_LAZY  = 4,
  63          REFERENCE_FLAG_PRIVATE_UNDEFINED_LAZY      = 5
  64   };
  65   enum { N_NO_DEAD_STRIP = 0x0020, // symbol is not to be dead stripped
  66          N_WEAK_REF      = 0x0040, // symbol is weak referenced
  67          N_WEAK_DEF      = 0x0080  // coalesced symbol is a weak definition
  68   };
  69
  70   MachOSym(const GlobalValue *gv, std::string name, uint8_t sect,
  71            const MCAsmInfo *MAI);
  72
  73   struct SymCmp {
  74     // FIXME: this does not appear to be sorting 'f' after 'F'
  75     bool operator()(const MachOSym &LHS, const MachOSym &RHS) {
  76       return LHS.GVName < RHS.GVName;
  77     }
  78   };
  79
  80
  81   /// PartitionByLocal - Simple boolean predicate that returns true if Sym is
  82   /// a local symbol rather than an external symbol.
  83
  84   static inline bool PartitionByLocal(const MachOSym &Sym) {
  85     return (Sym.n_type & (MachOSym::N_EXT | MachOSym::N_PEXT)) == 0;
  86   }
  87
  88   /// PartitionByDefined - Simple boolean predicate that returns true if Sym is
  89   /// defined in this module.
  90
  91   static inline bool PartitionByDefined(const MachOSym &Sym) {
  92     // FIXME: Do N_ABS or N_INDR count as defined?
  93     return (Sym.n_type & MachOSym::N_SECT) == MachOSym::N_SECT;
  94   }
  95
  96 }; // end struct MachOSym
  97
  98 /// MachOHeader - This struct contains the header information about a
  99 /// specific architecture type/subtype pair that is emitted to the file.
 100
 101 struct MachOHeader {
 102   uint32_t  magic;      // mach magic number identifier
 103   uint32_t  filetype;   // type of file
 104   uint32_t  ncmds;      // number of load commands
 105   uint32_t  sizeofcmds; // the size of all the load commands
 106   uint32_t  flags;      // flags
 107   uint32_t  reserved;   // 64-bit only
 108
 109   /// HeaderData - The actual data for the header which we are building
 110   /// up for emission to the file.
 111   std::vector<unsigned char> HeaderData;
 112
 113   // Constants for the filetype field
 114   // see <mach-o/loader.h> for additional info on the various types
 115   enum { MH_OBJECT     = 1, // relocatable object file
 116          MH_EXECUTE    = 2, // demand paged executable file
 117          MH_FVMLIB     = 3, // fixed VM shared library file
 118          MH_CORE       = 4, // core file
 119          MH_PRELOAD    = 5, // preloaded executable file
 120          MH_DYLIB      = 6, // dynamically bound shared library
 121          MH_DYLINKER   = 7, // dynamic link editor
 122          MH_BUNDLE     = 8, // dynamically bound bundle file
 123          MH_DYLIB_STUB = 9, // shared library stub for static linking only
 124          MH_DSYM       = 10 // companion file wiht only debug sections
 125   };
 126
 127   // Constants for the flags field
 128   enum { MH_NOUNDEFS                = 1 << 0,
 129             // the object file has no undefined references
 130          MH_INCRLINK                = 1 << 1,
 131             // the object file is the output of an incremental link against
 132             // a base file and cannot be link edited again
 133          MH_DYLDLINK                = 1 << 2,
 134             // the object file is input for the dynamic linker and cannot be
 135             // statically link edited again.
 136          MH_BINDATLOAD              = 1 << 3,
 137             // the object file's undefined references are bound by the
 138             // dynamic linker when loaded.
 139          MH_PREBOUND                = 1 << 4,
 140             // the file has its dynamic undefined references prebound
 141          MH_SPLIT_SEGS              = 1 << 5,
 142             // the file has its read-only and read-write segments split
 143             // see <mach/shared_memory_server.h>
 144          MH_LAZY_INIT               = 1 << 6,
 145             // the shared library init routine is to be run lazily via
 146             // catching memory faults to its writable segments (obsolete)
 147          MH_TWOLEVEL                = 1 << 7,
 148             // the image is using two-level namespace bindings
 149          MH_FORCE_FLAT              = 1 << 8,
 150             // the executable is forcing all images to use flat namespace
 151             // bindings.
 152          MH_NOMULTIDEFS             = 1 << 8,
 153             // this umbrella guarantees no multiple definitions of symbols
 154             // in its sub-images so the two-level namespace hints can
 155             // always be used.
 156          MH_NOFIXPREBINDING         = 1 << 10,
 157             // do not have dyld notify the prebidning agent about this
 158             // executable.
 159          MH_PREBINDABLE             = 1 << 11,
 160             // the binary is not prebound but can have its prebinding
 161             // redone.  only used when MH_PREBOUND is not set.
 162          MH_ALLMODSBOUND            = 1 << 12,
 163             // indicates that this binary binds to all two-level namespace
 164             // modules of its dependent libraries.  Only used when
 165             // MH_PREBINDABLE and MH_TWOLEVEL are both set.
 166          MH_SUBSECTIONS_VIA_SYMBOLS = 1 << 13,
 167             // safe to divide up the sections into sub-sections via symbols
 168             // for dead code stripping.
 169          MH_CANONICAL               = 1 << 14,
 170             // the binary has been canonicalized via the unprebind operation
 171          MH_WEAK_DEFINES            = 1 << 15,
 172             // the final linked image contains external weak symbols
 173          MH_BINDS_TO_WEAK           = 1 << 16,
 174             // the final linked image uses weak symbols
 175          MH_ALLOW_STACK_EXECUTION   = 1 << 17
 176             // When this bit is set, all stacks in the task will be given
 177             // stack execution privilege.  Only used in MH_EXECUTE filetype
 178   };
 179
 180   MachOHeader() : magic(0), filetype(0), ncmds(0), sizeofcmds(0), flags(0),
 181                   reserved(0) {}
 182
 183   /// cmdSize - This routine returns the size of the MachOSection as written
 184   /// to disk, depending on whether the destination is a 64 bit Mach-O file.
 185   unsigned cmdSize(bool is64Bit) const {
 186     if (is64Bit)
 187       return 8 * sizeof(uint32_t);
 188     else
 189       return 7 * sizeof(uint32_t);
 190   }
 191
 192   /// setMagic - This routine sets the appropriate value for the 'magic'
 193   /// field based on pointer size and endianness.
 194   void setMagic(bool isLittleEndian, bool is64Bit) {
 195     if (isLittleEndian)
 196       if (is64Bit) magic = 0xcffaedfe;
 197       else         magic = 0xcefaedfe;
 198     else
 199       if (is64Bit) magic = 0xfeedfacf;
 200       else         magic = 0xfeedface;
 201   }
 202
 203 }; // end struct MachOHeader
 204
 205 /// MachOSegment - This struct contains the necessary information to
 206 /// emit the load commands for each section in the file.
 207 struct MachOSegment {
 208   uint32_t    cmd;      // LC_SEGMENT or LC_SEGMENT_64
 209   uint32_t    cmdsize;  // Total size of this struct and section commands
 210   std::string segname;  // segment name
 211   uint64_t    vmaddr;   // address of this segment
 212   uint64_t    vmsize;   // size of this segment, may be larger than filesize
 213   uint64_t    fileoff;  // offset in file
 214   uint64_t    filesize; // amount to read from file
 215   uint32_t    maxprot;  // maximum VM protection
 216   uint32_t    initprot; // initial VM protection
 217   uint32_t    nsects;   // number of sections in this segment
 218   uint32_t    flags;    // flags
 219
 220   // The following constants are getting pulled in by one of the
 221   // system headers, which creates a neat clash with the enum.
 222 #if !defined(VM_PROT_NONE)
 223 #define VM_PROT_NONE    0x00
 224 #endif
 225 #if !defined(VM_PROT_READ)
 226 #define VM_PROT_READ    0x01
 227 #endif
 228 #if !defined(VM_PROT_WRITE)
 229 #define VM_PROT_WRITE   0x02
 230 #endif
 231 #if !defined(VM_PROT_EXECUTE)
 232 #define VM_PROT_EXECUTE 0x04
 233 #endif
 234 #if !defined(VM_PROT_ALL)
 235 #define VM_PROT_ALL     0x07
 236 #endif
 237
 238   // Constants for the vm protection fields
 239   // see <mach-o/vm_prot.h>
 240   enum { SEG_VM_PROT_NONE     = VM_PROT_NONE,
 241          SEG_VM_PROT_READ     = VM_PROT_READ, // read permission
 242          SEG_VM_PROT_WRITE    = VM_PROT_WRITE, // write permission
 243          SEG_VM_PROT_EXECUTE  = VM_PROT_EXECUTE,
 244          SEG_VM_PROT_ALL      = VM_PROT_ALL
 245   };
 246
 247   // Constants for the cmd field
 248   // see <mach-o/loader.h>
 249   enum { LC_SEGMENT    = 0x01,  // segment of this file to be mapped
 250          LC_SEGMENT_64 = 0x19   // 64-bit segment of this file to be mapped
 251   };
 252
 253   /// cmdSize - This routine returns the size of the MachOSection as written
 254   /// to disk, depending on whether the destination is a 64 bit Mach-O file.
 255   unsigned cmdSize(bool is64Bit) const {
 256     if (is64Bit)
 257       return 6 * sizeof(uint32_t) + 4 * sizeof(uint64_t) + 16;
 258     else
 259       return 10 * sizeof(uint32_t) + 16;  // addresses only 32 bits
 260   }
 261
 262   MachOSegment(const std::string &seg, bool is64Bit)
 263     : cmd(is64Bit ? LC_SEGMENT_64 : LC_SEGMENT), cmdsize(0), segname(seg),
 264       vmaddr(0), vmsize(0), fileoff(0), filesize(0), maxprot(VM_PROT_ALL),
 265       initprot(VM_PROT_ALL), nsects(0), flags(0) { }
 266 };
 267
 268 /// MachOSection - This struct contains information about each section in a
 269 /// particular segment that is emitted to the file.  This is eventually
 270 /// turned into the SectionCommand in the load command for a particlar
 271 /// segment.
 272
 273 struct MachOSection : public BinaryObject {
 274   std::string  sectname; // name of this section,
 275   std::string  segname;  // segment this section goes in
 276   uint64_t  addr;        // memory address of this section
 277   uint32_t  offset;      // file offset of this section
 278   uint32_t  align;       // section alignment (power of 2)
 279   uint32_t  reloff;      // file offset of relocation entries
 280   uint32_t  nreloc;      // number of relocation entries
 281   uint32_t  flags;       // flags (section type and attributes)
 282   uint32_t  reserved1;   // reserved (for offset or index)
 283   uint32_t  reserved2;   // reserved (for count or sizeof)
 284   uint32_t  reserved3;   // reserved (64 bit only)
 285
 286   /// A unique number for this section, which will be used to match symbols
 287   /// to the correct section.
 288   uint32_t Index;
 289
 290   /// RelocBuffer - A buffer to hold the mach-o relocations before we write
 291   /// them out at the appropriate location in the file.
 292   std::vector<unsigned char> RelocBuffer;
 293
 294   // Constants for the section types (low 8 bits of flags field)
 295   // see <mach-o/loader.h>
 296   enum { S_REGULAR = 0,
 297             // regular section
 298          S_ZEROFILL = 1,
 299             // zero fill on demand section
 300          S_CSTRING_LITERALS = 2,
 301             // section with only literal C strings
 302          S_4BYTE_LITERALS = 3,
 303             // section with only 4 byte literals
 304          S_8BYTE_LITERALS = 4,
 305             // section with only 8 byte literals
 306          S_LITERAL_POINTERS = 5,
 307             // section with only pointers to literals
 308          S_NON_LAZY_SYMBOL_POINTERS = 6,
 309             // section with only non-lazy symbol pointers
 310          S_LAZY_SYMBOL_POINTERS = 7,
 311             // section with only lazy symbol pointers
 312          S_SYMBOL_STUBS = 8,
 313             // section with only symbol stubs
 314             // byte size of stub in the reserved2 field
 315          S_MOD_INIT_FUNC_POINTERS = 9,
 316             // section with only function pointers for initialization
 317          S_MOD_TERM_FUNC_POINTERS = 10,
 318             // section with only function pointers for termination
 319          S_COALESCED = 11,
 320             // section contains symbols that are coalesced
 321          S_GB_ZEROFILL = 12,
 322             // zero fill on demand section (that can be larger than 4GB)
 323          S_INTERPOSING = 13,
 324             // section with only pairs of function pointers for interposing
 325          S_16BYTE_LITERALS = 14
 326             // section with only 16 byte literals
 327   };
 328
 329   // Constants for the section flags (high 24 bits of flags field)
 330   // see <mach-o/loader.h>
 331   enum { S_ATTR_PURE_INSTRUCTIONS   = 1 << 31,
 332             // section contains only true machine instructions
 333          S_ATTR_NO_TOC              = 1 << 30,
 334             // section contains coalesced symbols that are not to be in a
 335             // ranlib table of contents
 336          S_ATTR_STRIP_STATIC_SYMS   = 1 << 29,
 337             // ok to strip static symbols in this section in files with the
 338             // MY_DYLDLINK flag
 339          S_ATTR_NO_DEAD_STRIP       = 1 << 28,
 340             // no dead stripping
 341          S_ATTR_LIVE_SUPPORT        = 1 << 27,
 342             // blocks are live if they reference live blocks
 343          S_ATTR_SELF_MODIFYING_CODE = 1 << 26,
 344             // used with i386 code stubs written on by dyld
 345          S_ATTR_DEBUG               = 1 << 25,
 346             // a debug section
 347          S_ATTR_SOME_INSTRUCTIONS   = 1 << 10,
 348             // section contains some machine instructions
 349          S_ATTR_EXT_RELOC           = 1 << 9,
 350             // section has external relocation entries
 351          S_ATTR_LOC_RELOC           = 1 << 8
 352             // section has local relocation entries
 353   };
 354
 355   /// cmdSize - This routine returns the size of the MachOSection as written
 356   /// to disk, depending on whether the destination is a 64 bit Mach-O file.
 357   unsigned cmdSize(bool is64Bit) const {
 358     if (is64Bit)
 359       return 7 * sizeof(uint32_t) + 2 * sizeof(uint64_t) + 32;
 360     else
 361       return 9 * sizeof(uint32_t) + 32;  // addresses only 32 bits
 362   }
 363
 364   MachOSection(const std::string &seg, const std::string &sect)
 365     : BinaryObject(), sectname(sect), segname(seg), addr(0), offset(0),
 366       align(2), reloff(0), nreloc(0), flags(0), reserved1(0), reserved2(0),
 367       reserved3(0) { }
 368
 369 }; // end struct MachOSection
 370
 371 /// MachOSymTab - This struct contains information about the offsets and
 372 /// size of symbol table information.
 373 /// segment.
 374 struct MachODySymTab {
 375   uint32_t cmd;             // LC_DYSYMTAB
 376   uint32_t cmdsize;         // sizeof(MachODySymTab)
 377   uint32_t ilocalsym;       // index to local symbols
 378   uint32_t nlocalsym;       // number of local symbols
 379   uint32_t iextdefsym;      // index to externally defined symbols
 380   uint32_t nextdefsym;      // number of externally defined symbols
 381   uint32_t iundefsym;       // index to undefined symbols
 382   uint32_t nundefsym;       // number of undefined symbols
 383   uint32_t tocoff;          // file offset to table of contents
 384   uint32_t ntoc;            // number of entries in table of contents
 385   uint32_t modtaboff;       // file offset to module table
 386   uint32_t nmodtab;         // number of module table entries
 387   uint32_t extrefsymoff;    // offset to referenced symbol table
 388   uint32_t nextrefsyms;     // number of referenced symbol table entries
 389   uint32_t indirectsymoff;  // file offset to the indirect symbol table
 390   uint32_t nindirectsyms;   // number of indirect symbol table entries
 391   uint32_t extreloff;       // offset to external relocation entries
 392   uint32_t nextrel;         // number of external relocation entries
 393   uint32_t locreloff;       // offset to local relocation entries
 394   uint32_t nlocrel;         // number of local relocation entries
 395
 396   // Constants for the cmd field
 397   // see <mach-o/loader.h>
 398   enum { LC_DYSYMTAB = 0x0B  // dynamic link-edit symbol table info
 399   };
 400
 401   MachODySymTab() : cmd(LC_DYSYMTAB), cmdsize(20 * sizeof(uint32_t)),
 402     ilocalsym(0), nlocalsym(0), iextdefsym(0), nextdefsym(0),
 403     iundefsym(0), nundefsym(0), tocoff(0), ntoc(0), modtaboff(0),
 404     nmodtab(0), extrefsymoff(0), nextrefsyms(0), indirectsymoff(0),
 405     nindirectsyms(0), extreloff(0), nextrel(0), locreloff(0), nlocrel(0) {}
 406
 407 }; // end struct MachODySymTab
 408
 409 } // end namespace llvm
 410
 411 #endif
 412