Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 1 | #!/usr/bin/python |
| 2 | |
| 3 | import cmd |
| 4 | import dict_utils |
| 5 | import file_extract |
| 6 | import optparse |
| 7 | import re |
| 8 | import struct |
| 9 | import string |
| 10 | import StringIO |
| 11 | import sys |
| 12 | import uuid |
| 13 | |
| 14 | # Mach header "magic" constants |
| 15 | MH_MAGIC = 0xfeedface |
| 16 | MH_CIGAM = 0xcefaedfe |
| 17 | MH_MAGIC_64 = 0xfeedfacf |
| 18 | MH_CIGAM_64 = 0xcffaedfe |
| 19 | FAT_MAGIC = 0xcafebabe |
| 20 | FAT_CIGAM = 0xbebafeca |
| 21 | |
| 22 | # Mach haeder "filetype" constants |
| 23 | MH_OBJECT = 0x00000001 |
| 24 | MH_EXECUTE = 0x00000002 |
| 25 | MH_FVMLIB = 0x00000003 |
| 26 | MH_CORE = 0x00000004 |
| 27 | MH_PRELOAD = 0x00000005 |
| 28 | MH_DYLIB = 0x00000006 |
| 29 | MH_DYLINKER = 0x00000007 |
| 30 | MH_BUNDLE = 0x00000008 |
| 31 | MH_DYLIB_STUB = 0x00000009 |
| 32 | MH_DSYM = 0x0000000a |
| 33 | MH_KEXT_BUNDLE = 0x0000000b |
| 34 | |
| 35 | # Mach haeder "flag" constant bits |
| 36 | MH_NOUNDEFS = 0x00000001 |
| 37 | MH_INCRLINK = 0x00000002 |
| 38 | MH_DYLDLINK = 0x00000004 |
| 39 | MH_BINDATLOAD = 0x00000008 |
| 40 | MH_PREBOUND = 0x00000010 |
| 41 | MH_SPLIT_SEGS = 0x00000020 |
| 42 | MH_LAZY_INIT = 0x00000040 |
| 43 | MH_TWOLEVEL = 0x00000080 |
| 44 | MH_FORCE_FLAT = 0x00000100 |
| 45 | MH_NOMULTIDEFS = 0x00000200 |
| 46 | MH_NOFIXPREBINDING = 0x00000400 |
| 47 | MH_PREBINDABLE = 0x00000800 |
| 48 | MH_ALLMODSBOUND = 0x00001000 |
| 49 | MH_SUBSECTIONS_VIA_SYMBOLS = 0x00002000 |
| 50 | MH_CANONICAL = 0x00004000 |
| 51 | MH_WEAK_DEFINES = 0x00008000 |
| 52 | MH_BINDS_TO_WEAK = 0x00010000 |
| 53 | MH_ALLOW_STACK_EXECUTION = 0x00020000 |
| 54 | MH_ROOT_SAFE = 0x00040000 |
| 55 | MH_SETUID_SAFE = 0x00080000 |
| 56 | MH_NO_REEXPORTED_DYLIBS = 0x00100000 |
| 57 | MH_PIE = 0x00200000 |
| 58 | MH_DEAD_STRIPPABLE_DYLIB = 0x00400000 |
| 59 | MH_HAS_TLV_DESCRIPTORS = 0x00800000 |
| 60 | MH_NO_HEAP_EXECUTION = 0x01000000 |
| 61 | |
| 62 | # Mach load command constants |
| 63 | LC_REQ_DYLD = 0x80000000 |
| 64 | LC_SEGMENT = 0x00000001 |
| 65 | LC_SYMTAB = 0x00000002 |
| 66 | LC_SYMSEG = 0x00000003 |
| 67 | LC_THREAD = 0x00000004 |
| 68 | LC_UNIXTHREAD = 0x00000005 |
| 69 | LC_LOADFVMLIB = 0x00000006 |
| 70 | LC_IDFVMLIB = 0x00000007 |
| 71 | LC_IDENT = 0x00000008 |
| 72 | LC_FVMFILE = 0x00000009 |
| 73 | LC_PREPAGE = 0x0000000a |
| 74 | LC_DYSYMTAB = 0x0000000b |
| 75 | LC_LOAD_DYLIB = 0x0000000c |
| 76 | LC_ID_DYLIB = 0x0000000d |
| 77 | LC_LOAD_DYLINKER = 0x0000000e |
| 78 | LC_ID_DYLINKER = 0x0000000f |
| 79 | LC_PREBOUND_DYLIB = 0x00000010 |
| 80 | LC_ROUTINES = 0x00000011 |
| 81 | LC_SUB_FRAMEWORK = 0x00000012 |
| 82 | LC_SUB_UMBRELLA = 0x00000013 |
| 83 | LC_SUB_CLIENT = 0x00000014 |
| 84 | LC_SUB_LIBRARY = 0x00000015 |
| 85 | LC_TWOLEVEL_HINTS = 0x00000016 |
| 86 | LC_PREBIND_CKSUM = 0x00000017 |
| 87 | LC_LOAD_WEAK_DYLIB = 0x00000018 | LC_REQ_DYLD |
| 88 | LC_SEGMENT_64 = 0x00000019 |
| 89 | LC_ROUTINES_64 = 0x0000001a |
| 90 | LC_UUID = 0x0000001b |
| 91 | LC_RPATH = 0x0000001c | LC_REQ_DYLD |
| 92 | LC_CODE_SIGNATURE = 0x0000001d |
| 93 | LC_SEGMENT_SPLIT_INFO = 0x0000001e |
| 94 | LC_REEXPORT_DYLIB = 0x0000001f | LC_REQ_DYLD |
| 95 | LC_LAZY_LOAD_DYLIB = 0x00000020 |
| 96 | LC_ENCRYPTION_INFO = 0x00000021 |
| 97 | LC_DYLD_INFO = 0x00000022 |
| 98 | LC_DYLD_INFO_ONLY = 0x00000022 | LC_REQ_DYLD |
| 99 | LC_LOAD_UPWARD_DYLIB = 0x00000023 | LC_REQ_DYLD |
| 100 | LC_VERSION_MIN_MACOSX = 0x00000024 |
| 101 | LC_VERSION_MIN_IPHONEOS = 0x00000025 |
| 102 | LC_FUNCTION_STARTS = 0x00000026 |
| 103 | LC_DYLD_ENVIRONMENT = 0x00000027 |
| 104 | |
| 105 | # Mach CPU constants |
| 106 | CPU_ARCH_MASK = 0xff000000 |
| 107 | CPU_ARCH_ABI64 = 0x01000000 |
| 108 | CPU_TYPE_ANY = 0xffffffff |
| 109 | CPU_TYPE_VAX = 1 |
| 110 | CPU_TYPE_MC680x0 = 6 |
| 111 | CPU_TYPE_I386 = 7 |
| 112 | CPU_TYPE_X86_64 = CPU_TYPE_I386 | CPU_ARCH_ABI64 |
| 113 | CPU_TYPE_MIPS = 8 |
| 114 | CPU_TYPE_MC98000 = 10 |
| 115 | CPU_TYPE_HPPA = 11 |
| 116 | CPU_TYPE_ARM = 12 |
| 117 | CPU_TYPE_MC88000 = 13 |
| 118 | CPU_TYPE_SPARC = 14 |
| 119 | CPU_TYPE_I860 = 15 |
| 120 | CPU_TYPE_ALPHA = 16 |
| 121 | CPU_TYPE_POWERPC = 18 |
| 122 | CPU_TYPE_POWERPC64 = CPU_TYPE_POWERPC | CPU_ARCH_ABI64 |
| 123 | |
| 124 | # VM protection constants |
| 125 | VM_PROT_READ = 1 |
| 126 | VM_PROT_WRITE = 2 |
| 127 | VM_PROT_EXECUTE = 4 |
| 128 | |
| 129 | # VM protection constants |
| 130 | N_STAB = 0xe0 |
| 131 | N_PEXT = 0x10 |
| 132 | N_TYPE = 0x0e |
| 133 | N_EXT = 0x01 |
| 134 | |
| 135 | # Values for nlist N_TYPE bits of the "Mach.NList.type" field. |
| 136 | N_UNDF = 0x0 |
| 137 | N_ABS = 0x2 |
| 138 | N_SECT = 0xe |
| 139 | N_PBUD = 0xc |
| 140 | N_INDR = 0xa |
| 141 | |
| 142 | # Section indexes for the "Mach.NList.sect_idx" fields |
| 143 | NO_SECT = 0 |
| 144 | MAX_SECT = 255 |
| 145 | |
| 146 | # Stab defines |
| 147 | N_GSYM = 0x20 |
| 148 | N_FNAME = 0x22 |
| 149 | N_FUN = 0x24 |
| 150 | N_STSYM = 0x26 |
| 151 | N_LCSYM = 0x28 |
| 152 | N_BNSYM = 0x2e |
| 153 | N_OPT = 0x3c |
| 154 | N_RSYM = 0x40 |
| 155 | N_SLINE = 0x44 |
| 156 | N_ENSYM = 0x4e |
| 157 | N_SSYM = 0x60 |
| 158 | N_SO = 0x64 |
| 159 | N_OSO = 0x66 |
| 160 | N_LSYM = 0x80 |
| 161 | N_BINCL = 0x82 |
| 162 | N_SOL = 0x84 |
| 163 | N_PARAMS = 0x86 |
| 164 | N_VERSION = 0x88 |
| 165 | N_OLEVEL = 0x8A |
| 166 | N_PSYM = 0xa0 |
| 167 | N_EINCL = 0xa2 |
| 168 | N_ENTRY = 0xa4 |
| 169 | N_LBRAC = 0xc0 |
| 170 | N_EXCL = 0xc2 |
| 171 | N_RBRAC = 0xe0 |
| 172 | N_BCOMM = 0xe2 |
| 173 | N_ECOMM = 0xe4 |
| 174 | N_ECOML = 0xe8 |
| 175 | N_LENG = 0xfe |
| 176 | |
| 177 | vm_prot_names = [ '---', 'r--', '-w-', 'rw-', '--x', 'r-x', '-wx', 'rwx' ] |
| 178 | |
| 179 | def dump_memory(base_addr, data, hex_bytes_len, num_per_line): |
| 180 | hex_bytes = data.encode('hex') |
| 181 | if hex_bytes_len == -1: |
| 182 | hex_bytes_len = len(hex_bytes) |
| 183 | addr = base_addr |
| 184 | ascii_str = '' |
| 185 | i = 0 |
| 186 | while i < hex_bytes_len: |
| 187 | if ((i/2) % num_per_line) == 0: |
| 188 | if i > 0: |
| 189 | print ' %s' % (ascii_str) |
| 190 | ascii_str = '' |
| 191 | print '0x%8.8x:' % (addr+i), |
| 192 | hex_byte = hex_bytes[i:i+2] |
| 193 | print hex_byte, |
| 194 | int_byte = int (hex_byte, 16) |
| 195 | ascii_char = '%c' % (int_byte) |
| 196 | if int_byte >= 32 and int_byte < 127: |
| 197 | ascii_str += ascii_char |
| 198 | else: |
| 199 | ascii_str += '.' |
| 200 | i = i + 2 |
| 201 | if ascii_str: |
| 202 | if (i/2) % num_per_line: |
| 203 | padding = num_per_line - ((i/2) % num_per_line) |
| 204 | else: |
| 205 | padding = 0 |
| 206 | print '%*s%s' % (padding*3+1,'',ascii_str) |
| 207 | print |
| 208 | |
| 209 | |
| 210 | class TerminalColors: |
| 211 | '''Simple terminal colors class''' |
| 212 | def __init__(self, enabled = True): |
| 213 | # TODO: discover terminal type from "file" and disable if |
| 214 | # it can't handle the color codes |
| 215 | self.enabled = enabled |
| 216 | |
| 217 | def reset(self): |
| 218 | '''Reset all terminal colors and formatting.''' |
| 219 | if self.enabled: |
| 220 | return "\x1b[0m"; |
| 221 | return '' |
| 222 | |
| 223 | def bold(self, on = True): |
Bruce Mitchener | d93c4a3 | 2014-07-01 21:22:11 +0000 | [diff] [blame^] | 224 | '''Enable or disable bold depending on the "on" parameter.''' |
Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 225 | if self.enabled: |
| 226 | if on: |
| 227 | return "\x1b[1m"; |
| 228 | else: |
| 229 | return "\x1b[22m"; |
| 230 | return '' |
| 231 | |
| 232 | def italics(self, on = True): |
Bruce Mitchener | d93c4a3 | 2014-07-01 21:22:11 +0000 | [diff] [blame^] | 233 | '''Enable or disable italics depending on the "on" parameter.''' |
Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 234 | if self.enabled: |
| 235 | if on: |
| 236 | return "\x1b[3m"; |
| 237 | else: |
| 238 | return "\x1b[23m"; |
| 239 | return '' |
| 240 | |
| 241 | def underline(self, on = True): |
Bruce Mitchener | d93c4a3 | 2014-07-01 21:22:11 +0000 | [diff] [blame^] | 242 | '''Enable or disable underline depending on the "on" parameter.''' |
Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 243 | if self.enabled: |
| 244 | if on: |
| 245 | return "\x1b[4m"; |
| 246 | else: |
| 247 | return "\x1b[24m"; |
| 248 | return '' |
| 249 | |
| 250 | def inverse(self, on = True): |
Bruce Mitchener | d93c4a3 | 2014-07-01 21:22:11 +0000 | [diff] [blame^] | 251 | '''Enable or disable inverse depending on the "on" parameter.''' |
Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 252 | if self.enabled: |
| 253 | if on: |
| 254 | return "\x1b[7m"; |
| 255 | else: |
| 256 | return "\x1b[27m"; |
| 257 | return '' |
| 258 | |
| 259 | def strike(self, on = True): |
Bruce Mitchener | d93c4a3 | 2014-07-01 21:22:11 +0000 | [diff] [blame^] | 260 | '''Enable or disable strike through depending on the "on" parameter.''' |
Greg Clayton | 6a23d21 | 2013-09-04 17:31:40 +0000 | [diff] [blame] | 261 | if self.enabled: |
| 262 | if on: |
| 263 | return "\x1b[9m"; |
| 264 | else: |
| 265 | return "\x1b[29m"; |
| 266 | return '' |
| 267 | |
| 268 | def black(self, fg = True): |
| 269 | '''Set the foreground or background color to black. |
| 270 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 271 | if self.enabled: |
| 272 | if fg: |
| 273 | return "\x1b[30m"; |
| 274 | else: |
| 275 | return "\x1b[40m"; |
| 276 | return '' |
| 277 | |
| 278 | def red(self, fg = True): |
| 279 | '''Set the foreground or background color to red. |
| 280 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 281 | if self.enabled: |
| 282 | if fg: |
| 283 | return "\x1b[31m"; |
| 284 | else: |
| 285 | return "\x1b[41m"; |
| 286 | return '' |
| 287 | |
| 288 | def green(self, fg = True): |
| 289 | '''Set the foreground or background color to green. |
| 290 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 291 | if self.enabled: |
| 292 | if fg: |
| 293 | return "\x1b[32m"; |
| 294 | else: |
| 295 | return "\x1b[42m"; |
| 296 | return '' |
| 297 | |
| 298 | def yellow(self, fg = True): |
| 299 | '''Set the foreground or background color to yellow. |
| 300 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 301 | if self.enabled: |
| 302 | if fg: |
| 303 | return "\x1b[43m"; |
| 304 | else: |
| 305 | return "\x1b[33m"; |
| 306 | return '' |
| 307 | |
| 308 | def blue(self, fg = True): |
| 309 | '''Set the foreground or background color to blue. |
| 310 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 311 | if self.enabled: |
| 312 | if fg: |
| 313 | return "\x1b[34m"; |
| 314 | else: |
| 315 | return "\x1b[44m"; |
| 316 | return '' |
| 317 | |
| 318 | def magenta(self, fg = True): |
| 319 | '''Set the foreground or background color to magenta. |
| 320 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 321 | if self.enabled: |
| 322 | if fg: |
| 323 | return "\x1b[35m"; |
| 324 | else: |
| 325 | return "\x1b[45m"; |
| 326 | return '' |
| 327 | |
| 328 | def cyan(self, fg = True): |
| 329 | '''Set the foreground or background color to cyan. |
| 330 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 331 | if self.enabled: |
| 332 | if fg: |
| 333 | return "\x1b[36m"; |
| 334 | else: |
| 335 | return "\x1b[46m"; |
| 336 | return '' |
| 337 | |
| 338 | def white(self, fg = True): |
| 339 | '''Set the foreground or background color to white. |
| 340 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 341 | if self.enabled: |
| 342 | if fg: |
| 343 | return "\x1b[37m"; |
| 344 | else: |
| 345 | return "\x1b[47m"; |
| 346 | return '' |
| 347 | |
| 348 | def default(self, fg = True): |
| 349 | '''Set the foreground or background color to the default. |
| 350 | The foreground color will be set if "fg" tests True. The background color will be set if "fg" tests False.''' |
| 351 | if self.enabled: |
| 352 | if fg: |
| 353 | return "\x1b[39m"; |
| 354 | else: |
| 355 | return "\x1b[49m"; |
| 356 | return '' |
| 357 | |
| 358 | def swap_unpack_char(): |
| 359 | """Returns the unpack prefix that will for non-native endian-ness.""" |
| 360 | if struct.pack('H', 1).startswith("\x00"): |
| 361 | return '<' |
| 362 | return '>' |
| 363 | |
| 364 | |
| 365 | def dump_hex_bytes(addr, s, bytes_per_line=16): |
| 366 | i = 0 |
| 367 | line = '' |
| 368 | for ch in s: |
| 369 | if (i % bytes_per_line) == 0: |
| 370 | if line: |
| 371 | print line |
| 372 | line = '%#8.8x: ' % (addr + i) |
| 373 | line += "%02X " % ord(ch) |
| 374 | i += 1 |
| 375 | print line |
| 376 | |
| 377 | def dump_hex_byte_string_diff(addr, a, b, bytes_per_line=16): |
| 378 | i = 0 |
| 379 | line = '' |
| 380 | a_len = len(a) |
| 381 | b_len = len(b) |
| 382 | if a_len < b_len: |
| 383 | max_len = b_len |
| 384 | else: |
| 385 | max_len = a_len |
| 386 | tty_colors = TerminalColors (True) |
| 387 | for i in range(max_len): |
| 388 | ch = None |
| 389 | if i < a_len: |
| 390 | ch_a = a[i] |
| 391 | ch = ch_a |
| 392 | else: |
| 393 | ch_a = None |
| 394 | if i < b_len: |
| 395 | ch_b = b[i] |
| 396 | if not ch: |
| 397 | ch = ch_b |
| 398 | else: |
| 399 | ch_b = None |
| 400 | mismatch = ch_a != ch_b |
| 401 | if (i % bytes_per_line) == 0: |
| 402 | if line: |
| 403 | print line |
| 404 | line = '%#8.8x: ' % (addr + i) |
| 405 | if mismatch: line += tty_colors.red() |
| 406 | line += "%02X " % ord(ch) |
| 407 | if mismatch: line += tty_colors.default() |
| 408 | i += 1 |
| 409 | |
| 410 | print line |
| 411 | |
| 412 | class Mach: |
| 413 | """Class that does everything mach-o related""" |
| 414 | |
| 415 | class Arch: |
| 416 | """Class that implements mach-o architectures""" |
| 417 | |
| 418 | def __init__(self, c=0, s=0): |
| 419 | self.cpu=c |
| 420 | self.sub=s |
| 421 | |
| 422 | def set_cpu_type(self, c): |
| 423 | self.cpu=c |
| 424 | def set_cpu_subtype(self, s): |
| 425 | self.sub=s |
| 426 | def set_arch(self, c, s): |
| 427 | self.cpu=c |
| 428 | self.sub=s |
| 429 | def is_64_bit(self): |
| 430 | return (self.cpu & CPU_ARCH_ABI64) != 0 |
| 431 | |
| 432 | cpu_infos = [ |
| 433 | [ "arm" , CPU_TYPE_ARM , CPU_TYPE_ANY ], |
| 434 | [ "arm" , CPU_TYPE_ARM , 0 ], |
| 435 | [ "armv4" , CPU_TYPE_ARM , 5 ], |
| 436 | [ "armv6" , CPU_TYPE_ARM , 6 ], |
| 437 | [ "armv5" , CPU_TYPE_ARM , 7 ], |
| 438 | [ "xscale" , CPU_TYPE_ARM , 8 ], |
| 439 | [ "armv7" , CPU_TYPE_ARM , 9 ], |
| 440 | [ "armv7f" , CPU_TYPE_ARM , 10 ], |
| 441 | [ "armv7s" , CPU_TYPE_ARM , 11 ], |
| 442 | [ "armv7k" , CPU_TYPE_ARM , 12 ], |
| 443 | [ "armv7m" , CPU_TYPE_ARM , 15 ], |
| 444 | [ "armv7em" , CPU_TYPE_ARM , 16 ], |
| 445 | [ "ppc" , CPU_TYPE_POWERPC , CPU_TYPE_ANY ], |
| 446 | [ "ppc" , CPU_TYPE_POWERPC , 0 ], |
| 447 | [ "ppc601" , CPU_TYPE_POWERPC , 1 ], |
| 448 | [ "ppc602" , CPU_TYPE_POWERPC , 2 ], |
| 449 | [ "ppc603" , CPU_TYPE_POWERPC , 3 ], |
| 450 | [ "ppc603e" , CPU_TYPE_POWERPC , 4 ], |
| 451 | [ "ppc603ev" , CPU_TYPE_POWERPC , 5 ], |
| 452 | [ "ppc604" , CPU_TYPE_POWERPC , 6 ], |
| 453 | [ "ppc604e" , CPU_TYPE_POWERPC , 7 ], |
| 454 | [ "ppc620" , CPU_TYPE_POWERPC , 8 ], |
| 455 | [ "ppc750" , CPU_TYPE_POWERPC , 9 ], |
| 456 | [ "ppc7400" , CPU_TYPE_POWERPC , 10 ], |
| 457 | [ "ppc7450" , CPU_TYPE_POWERPC , 11 ], |
| 458 | [ "ppc970" , CPU_TYPE_POWERPC , 100 ], |
| 459 | [ "ppc64" , CPU_TYPE_POWERPC64 , 0 ], |
| 460 | [ "ppc970-64" , CPU_TYPE_POWERPC64 , 100 ], |
| 461 | [ "i386" , CPU_TYPE_I386 , 3 ], |
| 462 | [ "i486" , CPU_TYPE_I386 , 4 ], |
| 463 | [ "i486sx" , CPU_TYPE_I386 , 0x84 ], |
| 464 | [ "i386" , CPU_TYPE_I386 , CPU_TYPE_ANY ], |
| 465 | [ "x86_64" , CPU_TYPE_X86_64 , 3 ], |
| 466 | [ "x86_64" , CPU_TYPE_X86_64 , CPU_TYPE_ANY ], |
| 467 | ] |
| 468 | |
| 469 | def __str__(self): |
| 470 | for info in self.cpu_infos: |
| 471 | if self.cpu == info[1] and (self.sub & 0x00ffffff) == info[2]: |
| 472 | return info[0] |
| 473 | return "{0}.{1}".format(self.cpu,self.sub) |
| 474 | |
| 475 | |
| 476 | class Magic(dict_utils.Enum): |
| 477 | |
| 478 | enum = { |
| 479 | 'MH_MAGIC' : MH_MAGIC, |
| 480 | 'MH_CIGAM' : MH_CIGAM, |
| 481 | 'MH_MAGIC_64' : MH_MAGIC_64, |
| 482 | 'MH_CIGAM_64' : MH_CIGAM_64, |
| 483 | 'FAT_MAGIC' : FAT_MAGIC, |
| 484 | 'FAT_CIGAM' : FAT_CIGAM |
| 485 | } |
| 486 | |
| 487 | def __init__(self, initial_value = 0): |
| 488 | dict_utils.Enum.__init__(self, initial_value, self.enum) |
| 489 | |
| 490 | def is_skinny_mach_file(self): |
| 491 | return self.value == MH_MAGIC or self.value == MH_CIGAM or self.value == MH_MAGIC_64 or self.value == MH_CIGAM_64 |
| 492 | |
| 493 | def is_universal_mach_file(self): |
| 494 | return self.value == FAT_MAGIC or self.value == FAT_CIGAM |
| 495 | |
| 496 | def unpack(self, data): |
| 497 | data.set_byte_order('native') |
| 498 | self.value = data.get_uint32(); |
| 499 | |
| 500 | def get_byte_order(self): |
| 501 | if self.value == MH_CIGAM or self.value == MH_CIGAM_64 or self.value == FAT_CIGAM: |
| 502 | return swap_unpack_char() |
| 503 | else: |
| 504 | return '=' |
| 505 | |
| 506 | def is_64_bit(self): |
| 507 | return self.value == MH_MAGIC_64 or self.value == MH_CIGAM_64 |
| 508 | |
| 509 | def __init__(self): |
| 510 | self.magic = Mach.Magic() |
| 511 | self.content = None |
| 512 | self.path = None |
| 513 | |
| 514 | def extract (self, path, extractor): |
| 515 | self.path = path; |
| 516 | self.unpack(extractor) |
| 517 | |
| 518 | def parse(self, path): |
| 519 | self.path = path; |
| 520 | try: |
| 521 | f = open(self.path) |
| 522 | file_extractor = file_extract.FileExtract(f, '=') |
| 523 | self.unpack(file_extractor) |
| 524 | #f.close() |
| 525 | except IOError as (errno, strerror): |
| 526 | print "I/O error({0}): {1}".format(errno, strerror) |
| 527 | except ValueError: |
| 528 | print "Could not convert data to an integer." |
| 529 | except: |
| 530 | print "Unexpected error:", sys.exc_info()[0] |
| 531 | raise |
| 532 | |
| 533 | def compare(self, rhs): |
| 534 | self.content.compare(rhs.content) |
| 535 | |
| 536 | def dump(self, options = None): |
| 537 | self.content.dump(options) |
| 538 | |
| 539 | def dump_header(self, dump_description = True, options = None): |
| 540 | self.content.dump_header(dump_description, options) |
| 541 | |
| 542 | def dump_load_commands(self, dump_description = True, options = None): |
| 543 | self.content.dump_load_commands(dump_description, options) |
| 544 | |
| 545 | def dump_sections(self, dump_description = True, options = None): |
| 546 | self.content.dump_sections(dump_description, options) |
| 547 | |
| 548 | def dump_section_contents(self, options): |
| 549 | self.content.dump_section_contents(options) |
| 550 | |
| 551 | def dump_symtab(self, dump_description = True, options = None): |
| 552 | self.content.dump_symtab(dump_description, options) |
| 553 | |
| 554 | def dump_symbol_names_matching_regex(self, regex, file=None): |
| 555 | self.content.dump_symbol_names_matching_regex(regex, file) |
| 556 | |
| 557 | def description(self): |
| 558 | return self.content.description() |
| 559 | |
| 560 | def unpack(self, data): |
| 561 | self.magic.unpack(data) |
| 562 | if self.magic.is_skinny_mach_file(): |
| 563 | self.content = Mach.Skinny(self.path) |
| 564 | elif self.magic.is_universal_mach_file(): |
| 565 | self.content = Mach.Universal(self.path) |
| 566 | else: |
| 567 | self.content = None |
| 568 | |
| 569 | if self.content != None: |
| 570 | self.content.unpack(data, self.magic) |
| 571 | |
| 572 | def is_valid(self): |
| 573 | return self.content != None |
| 574 | |
| 575 | class Universal: |
| 576 | |
| 577 | def __init__(self, path): |
| 578 | self.path = path |
| 579 | self.type = 'universal' |
| 580 | self.file_off = 0 |
| 581 | self.magic = None |
| 582 | self.nfat_arch = 0 |
| 583 | self.archs = list() |
| 584 | |
| 585 | def description(self): |
| 586 | s = '%#8.8x: %s (' % (self.file_off, self.path) |
| 587 | archs_string = '' |
| 588 | for arch in self.archs: |
| 589 | if len(archs_string): |
| 590 | archs_string += ', ' |
| 591 | archs_string += '%s' % arch.arch |
| 592 | s += archs_string |
| 593 | s += ')' |
| 594 | return s |
| 595 | |
| 596 | def unpack(self, data, magic = None): |
| 597 | self.file_off = data.tell() |
| 598 | if magic is None: |
| 599 | self.magic = Mach.Magic() |
| 600 | self.magic.unpack(data) |
| 601 | else: |
| 602 | self.magic = magic |
| 603 | self.file_off = self.file_off - 4 |
| 604 | # Universal headers are always in big endian |
| 605 | data.set_byte_order('big') |
| 606 | self.nfat_arch = data.get_uint32() |
| 607 | for i in range(self.nfat_arch): |
| 608 | self.archs.append(Mach.Universal.ArchInfo()) |
| 609 | self.archs[i].unpack(data) |
| 610 | for i in range(self.nfat_arch): |
| 611 | self.archs[i].mach = Mach.Skinny(self.path) |
| 612 | data.seek (self.archs[i].offset, 0) |
| 613 | skinny_magic = Mach.Magic() |
| 614 | skinny_magic.unpack (data) |
| 615 | self.archs[i].mach.unpack(data, skinny_magic) |
| 616 | |
| 617 | def compare(self, rhs): |
| 618 | print 'error: comparing two universal files is not supported yet' |
| 619 | return False |
| 620 | |
| 621 | def dump(self, options): |
| 622 | if options.dump_header: |
| 623 | print |
| 624 | print "Universal Mach File: magic = %s, nfat_arch = %u" % (self.magic, self.nfat_arch) |
| 625 | print |
| 626 | if self.nfat_arch > 0: |
| 627 | if options.dump_header: |
| 628 | self.archs[0].dump_header(True, options) |
| 629 | for i in range(self.nfat_arch): |
| 630 | self.archs[i].dump_flat(options) |
| 631 | if options.dump_header: |
| 632 | print |
| 633 | for i in range(self.nfat_arch): |
| 634 | self.archs[i].mach.dump(options) |
| 635 | |
| 636 | def dump_header(self, dump_description = True, options = None): |
| 637 | if dump_description: |
| 638 | print self.description() |
| 639 | for i in range(self.nfat_arch): |
| 640 | self.archs[i].mach.dump_header(True, options) |
| 641 | print |
| 642 | |
| 643 | def dump_load_commands(self, dump_description = True, options = None): |
| 644 | if dump_description: |
| 645 | print self.description() |
| 646 | for i in range(self.nfat_arch): |
| 647 | self.archs[i].mach.dump_load_commands(True, options) |
| 648 | print |
| 649 | |
| 650 | def dump_sections(self, dump_description = True, options = None): |
| 651 | if dump_description: |
| 652 | print self.description() |
| 653 | for i in range(self.nfat_arch): |
| 654 | self.archs[i].mach.dump_sections(True, options) |
| 655 | print |
| 656 | |
| 657 | def dump_section_contents(self, options): |
| 658 | for i in range(self.nfat_arch): |
| 659 | self.archs[i].mach.dump_section_contents(options) |
| 660 | print |
| 661 | |
| 662 | def dump_symtab(self, dump_description = True, options = None): |
| 663 | if dump_description: |
| 664 | print self.description() |
| 665 | for i in range(self.nfat_arch): |
| 666 | self.archs[i].mach.dump_symtab(True, options) |
| 667 | print |
| 668 | |
| 669 | def dump_symbol_names_matching_regex(self, regex, file=None): |
| 670 | for i in range(self.nfat_arch): |
| 671 | self.archs[i].mach.dump_symbol_names_matching_regex(regex, file) |
| 672 | |
| 673 | class ArchInfo: |
| 674 | |
| 675 | def __init__(self): |
| 676 | self.arch = Mach.Arch(0,0) |
| 677 | self.offset = 0 |
| 678 | self.size = 0 |
| 679 | self.align = 0 |
| 680 | self.mach = None |
| 681 | |
| 682 | def unpack(self, data): |
| 683 | # Universal headers are always in big endian |
| 684 | data.set_byte_order('big') |
| 685 | self.arch.cpu, self.arch.sub, self.offset, self.size, self.align = data.get_n_uint32(5) |
| 686 | |
| 687 | def dump_header(self, dump_description = True, options = None): |
| 688 | if options.verbose: |
| 689 | print "CPU SUBTYPE OFFSET SIZE ALIGN" |
| 690 | print "---------- ---------- ---------- ---------- ----------" |
| 691 | else: |
| 692 | print "ARCH FILEOFFSET FILESIZE ALIGN" |
| 693 | print "---------- ---------- ---------- ----------" |
| 694 | def dump_flat(self, options): |
| 695 | if options.verbose: |
| 696 | print "%#8.8x %#8.8x %#8.8x %#8.8x %#8.8x" % (self.arch.cpu, self.arch.sub, self.offset, self.size, self.align) |
| 697 | else: |
| 698 | print "%-10s %#8.8x %#8.8x %#8.8x" % (self.arch, self.offset, self.size, self.align) |
| 699 | def dump(self): |
| 700 | print " cputype: %#8.8x" % self.arch.cpu |
| 701 | print "cpusubtype: %#8.8x" % self.arch.sub |
| 702 | print " offset: %#8.8x" % self.offset |
| 703 | print " size: %#8.8x" % self.size |
| 704 | print " align: %#8.8x" % self.align |
| 705 | def __str__(self): |
| 706 | return "Mach.Universal.ArchInfo: %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x" % (self.arch.cpu, self.arch.sub, self.offset, self.size, self.align) |
| 707 | def __repr__(self): |
| 708 | return "Mach.Universal.ArchInfo: %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x" % (self.arch.cpu, self.arch.sub, self.offset, self.size, self.align) |
| 709 | |
| 710 | class Flags: |
| 711 | |
| 712 | def __init__(self, b): |
| 713 | self.bits = b |
| 714 | |
| 715 | def __str__(self): |
| 716 | s = '' |
| 717 | if self.bits & MH_NOUNDEFS: |
| 718 | s += 'MH_NOUNDEFS | ' |
| 719 | if self.bits & MH_INCRLINK: |
| 720 | s += 'MH_INCRLINK | ' |
| 721 | if self.bits & MH_DYLDLINK: |
| 722 | s += 'MH_DYLDLINK | ' |
| 723 | if self.bits & MH_BINDATLOAD: |
| 724 | s += 'MH_BINDATLOAD | ' |
| 725 | if self.bits & MH_PREBOUND: |
| 726 | s += 'MH_PREBOUND | ' |
| 727 | if self.bits & MH_SPLIT_SEGS: |
| 728 | s += 'MH_SPLIT_SEGS | ' |
| 729 | if self.bits & MH_LAZY_INIT: |
| 730 | s += 'MH_LAZY_INIT | ' |
| 731 | if self.bits & MH_TWOLEVEL: |
| 732 | s += 'MH_TWOLEVEL | ' |
| 733 | if self.bits & MH_FORCE_FLAT: |
| 734 | s += 'MH_FORCE_FLAT | ' |
| 735 | if self.bits & MH_NOMULTIDEFS: |
| 736 | s += 'MH_NOMULTIDEFS | ' |
| 737 | if self.bits & MH_NOFIXPREBINDING: |
| 738 | s += 'MH_NOFIXPREBINDING | ' |
| 739 | if self.bits & MH_PREBINDABLE: |
| 740 | s += 'MH_PREBINDABLE | ' |
| 741 | if self.bits & MH_ALLMODSBOUND: |
| 742 | s += 'MH_ALLMODSBOUND | ' |
| 743 | if self.bits & MH_SUBSECTIONS_VIA_SYMBOLS: |
| 744 | s += 'MH_SUBSECTIONS_VIA_SYMBOLS | ' |
| 745 | if self.bits & MH_CANONICAL: |
| 746 | s += 'MH_CANONICAL | ' |
| 747 | if self.bits & MH_WEAK_DEFINES: |
| 748 | s += 'MH_WEAK_DEFINES | ' |
| 749 | if self.bits & MH_BINDS_TO_WEAK: |
| 750 | s += 'MH_BINDS_TO_WEAK | ' |
| 751 | if self.bits & MH_ALLOW_STACK_EXECUTION: |
| 752 | s += 'MH_ALLOW_STACK_EXECUTION | ' |
| 753 | if self.bits & MH_ROOT_SAFE: |
| 754 | s += 'MH_ROOT_SAFE | ' |
| 755 | if self.bits & MH_SETUID_SAFE: |
| 756 | s += 'MH_SETUID_SAFE | ' |
| 757 | if self.bits & MH_NO_REEXPORTED_DYLIBS: |
| 758 | s += 'MH_NO_REEXPORTED_DYLIBS | ' |
| 759 | if self.bits & MH_PIE: |
| 760 | s += 'MH_PIE | ' |
| 761 | if self.bits & MH_DEAD_STRIPPABLE_DYLIB: |
| 762 | s += 'MH_DEAD_STRIPPABLE_DYLIB | ' |
| 763 | if self.bits & MH_HAS_TLV_DESCRIPTORS: |
| 764 | s += 'MH_HAS_TLV_DESCRIPTORS | ' |
| 765 | if self.bits & MH_NO_HEAP_EXECUTION: |
| 766 | s += 'MH_NO_HEAP_EXECUTION | ' |
| 767 | # Strip the trailing " |" if we have any flags |
| 768 | if len(s) > 0: |
| 769 | s = s[0:-2] |
| 770 | return s |
| 771 | |
| 772 | class FileType(dict_utils.Enum): |
| 773 | |
| 774 | enum = { |
| 775 | 'MH_OBJECT' : MH_OBJECT , |
| 776 | 'MH_EXECUTE' : MH_EXECUTE , |
| 777 | 'MH_FVMLIB' : MH_FVMLIB , |
| 778 | 'MH_CORE' : MH_CORE , |
| 779 | 'MH_PRELOAD' : MH_PRELOAD , |
| 780 | 'MH_DYLIB' : MH_DYLIB , |
| 781 | 'MH_DYLINKER' : MH_DYLINKER , |
| 782 | 'MH_BUNDLE' : MH_BUNDLE , |
| 783 | 'MH_DYLIB_STUB' : MH_DYLIB_STUB , |
| 784 | 'MH_DSYM' : MH_DSYM , |
| 785 | 'MH_KEXT_BUNDLE' : MH_KEXT_BUNDLE |
| 786 | } |
| 787 | |
| 788 | def __init__(self, initial_value = 0): |
| 789 | dict_utils.Enum.__init__(self, initial_value, self.enum) |
| 790 | |
| 791 | class Skinny: |
| 792 | |
| 793 | def __init__(self, path): |
| 794 | self.path = path |
| 795 | self.type = 'skinny' |
| 796 | self.data = None |
| 797 | self.file_off = 0 |
| 798 | self.magic = 0 |
| 799 | self.arch = Mach.Arch(0,0) |
| 800 | self.filetype = Mach.FileType(0) |
| 801 | self.ncmds = 0 |
| 802 | self.sizeofcmds = 0 |
| 803 | self.flags = Mach.Flags(0) |
| 804 | self.uuid = None |
| 805 | self.commands = list() |
| 806 | self.segments = list() |
| 807 | self.sections = list() |
| 808 | self.symbols = list() |
| 809 | self.sections.append(Mach.Section()) |
| 810 | |
| 811 | def description(self): |
| 812 | return '%#8.8x: %s (%s)' % (self.file_off, self.path, self.arch) |
| 813 | |
| 814 | def unpack(self, data, magic = None): |
| 815 | self.data = data |
| 816 | self.file_off = data.tell() |
| 817 | if magic is None: |
| 818 | self.magic = Mach.Magic() |
| 819 | self.magic.unpack(data) |
| 820 | else: |
| 821 | self.magic = magic |
| 822 | self.file_off = self.file_off - 4 |
| 823 | data.set_byte_order(self.magic.get_byte_order()) |
| 824 | self.arch.cpu, self.arch.sub, self.filetype.value, self.ncmds, self.sizeofcmds, bits = data.get_n_uint32(6) |
| 825 | self.flags.bits = bits |
| 826 | |
| 827 | if self.is_64_bit(): |
| 828 | data.get_uint32() # Skip reserved word in mach_header_64 |
| 829 | |
| 830 | for i in range(0,self.ncmds): |
| 831 | lc = self.unpack_load_command (data) |
| 832 | self.commands.append (lc) |
| 833 | |
| 834 | def get_data(self): |
| 835 | if self.data: |
| 836 | self.data.set_byte_order(self.magic.get_byte_order()) |
| 837 | return self.data |
| 838 | return None |
| 839 | |
| 840 | def unpack_load_command (self, data): |
| 841 | lc = Mach.LoadCommand() |
| 842 | lc.unpack (self, data) |
| 843 | lc_command = lc.command.get_enum_value(); |
| 844 | if (lc_command == LC_SEGMENT or |
| 845 | lc_command == LC_SEGMENT_64): |
| 846 | lc = Mach.SegmentLoadCommand(lc) |
| 847 | lc.unpack(self, data) |
| 848 | elif (lc_command == LC_LOAD_DYLIB or |
| 849 | lc_command == LC_ID_DYLIB or |
| 850 | lc_command == LC_LOAD_WEAK_DYLIB or |
| 851 | lc_command == LC_REEXPORT_DYLIB): |
| 852 | lc = Mach.DylibLoadCommand(lc) |
| 853 | lc.unpack(self, data) |
| 854 | elif (lc_command == LC_LOAD_DYLINKER or |
| 855 | lc_command == LC_SUB_FRAMEWORK or |
| 856 | lc_command == LC_SUB_CLIENT or |
| 857 | lc_command == LC_SUB_UMBRELLA or |
| 858 | lc_command == LC_SUB_LIBRARY or |
| 859 | lc_command == LC_ID_DYLINKER or |
| 860 | lc_command == LC_RPATH): |
| 861 | lc = Mach.LoadDYLDLoadCommand(lc) |
| 862 | lc.unpack(self, data) |
| 863 | elif (lc_command == LC_DYLD_INFO_ONLY): |
| 864 | lc = Mach.DYLDInfoOnlyLoadCommand(lc) |
| 865 | lc.unpack(self, data) |
| 866 | elif (lc_command == LC_SYMTAB): |
| 867 | lc = Mach.SymtabLoadCommand(lc) |
| 868 | lc.unpack(self, data) |
| 869 | elif (lc_command == LC_DYSYMTAB): |
| 870 | lc = Mach.DYLDSymtabLoadCommand(lc) |
| 871 | lc.unpack(self, data) |
| 872 | elif (lc_command == LC_UUID): |
| 873 | lc = Mach.UUIDLoadCommand(lc) |
| 874 | lc.unpack(self, data) |
| 875 | elif (lc_command == LC_CODE_SIGNATURE or |
| 876 | lc_command == LC_SEGMENT_SPLIT_INFO or |
| 877 | lc_command == LC_FUNCTION_STARTS): |
| 878 | lc = Mach.DataBlobLoadCommand(lc) |
| 879 | lc.unpack(self, data) |
| 880 | elif (lc_command == LC_UNIXTHREAD): |
| 881 | lc = Mach.UnixThreadLoadCommand(lc) |
| 882 | lc.unpack(self, data) |
| 883 | elif (lc_command == LC_ENCRYPTION_INFO): |
| 884 | lc = Mach.EncryptionInfoLoadCommand(lc) |
| 885 | lc.unpack(self, data) |
| 886 | lc.skip(data) |
| 887 | return lc |
| 888 | |
| 889 | def compare(self, rhs): |
| 890 | print "\nComparing:" |
| 891 | print "a) %s %s" % (self.arch, self.path) |
| 892 | print "b) %s %s" % (rhs.arch, rhs.path) |
| 893 | result = True |
| 894 | if self.type == rhs.type: |
| 895 | for lhs_section in self.sections[1:]: |
| 896 | rhs_section = rhs.get_section_by_section(lhs_section) |
| 897 | if rhs_section: |
| 898 | print 'comparing %s.%s...' % (lhs_section.segname, lhs_section.sectname), |
| 899 | sys.stdout.flush() |
| 900 | lhs_data = lhs_section.get_contents (self) |
| 901 | rhs_data = rhs_section.get_contents (rhs) |
| 902 | if lhs_data and rhs_data: |
| 903 | if lhs_data == rhs_data: |
| 904 | print 'ok' |
| 905 | else: |
| 906 | lhs_data_len = len(lhs_data) |
| 907 | rhs_data_len = len(rhs_data) |
| 908 | # if lhs_data_len < rhs_data_len: |
| 909 | # if lhs_data == rhs_data[0:lhs_data_len]: |
| 910 | # print 'section data for %s matches the first %u bytes' % (lhs_section.sectname, lhs_data_len) |
| 911 | # else: |
| 912 | # # TODO: check padding |
| 913 | # result = False |
| 914 | # elif lhs_data_len > rhs_data_len: |
| 915 | # if lhs_data[0:rhs_data_len] == rhs_data: |
| 916 | # print 'section data for %s matches the first %u bytes' % (lhs_section.sectname, lhs_data_len) |
| 917 | # else: |
| 918 | # # TODO: check padding |
| 919 | # result = False |
| 920 | # else: |
| 921 | result = False |
| 922 | print 'error: sections differ' |
| 923 | #print 'a) %s' % (lhs_section) |
| 924 | # dump_hex_byte_string_diff(0, lhs_data, rhs_data) |
| 925 | #print 'b) %s' % (rhs_section) |
| 926 | # dump_hex_byte_string_diff(0, rhs_data, lhs_data) |
| 927 | elif lhs_data and not rhs_data: |
| 928 | print 'error: section data missing from b:' |
| 929 | print 'a) %s' % (lhs_section) |
| 930 | print 'b) %s' % (rhs_section) |
| 931 | result = False |
| 932 | elif not lhs_data and rhs_data: |
| 933 | print 'error: section data missing from a:' |
| 934 | print 'a) %s' % (lhs_section) |
| 935 | print 'b) %s' % (rhs_section) |
| 936 | result = False |
| 937 | elif lhs_section.offset or rhs_section.offset: |
| 938 | print 'error: section data missing for both a and b:' |
| 939 | print 'a) %s' % (lhs_section) |
| 940 | print 'b) %s' % (rhs_section) |
| 941 | result = False |
| 942 | else: |
| 943 | print 'ok' |
| 944 | else: |
| 945 | result = False |
| 946 | print 'error: section %s is missing in %s' % (lhs_section.sectname, rhs.path) |
| 947 | else: |
| 948 | print 'error: comaparing a %s mach-o file with a %s mach-o file is not supported' % (self.type, rhs.type) |
| 949 | result = False |
| 950 | if not result: |
| 951 | print 'error: mach files differ' |
| 952 | return result |
| 953 | def dump_header(self, dump_description = True, options = None): |
| 954 | if options.verbose: |
| 955 | print "MAGIC CPU SUBTYPE FILETYPE NUM CMDS SIZE CMDS FLAGS" |
| 956 | print "---------- ---------- ---------- ---------- -------- ---------- ----------" |
| 957 | else: |
| 958 | print "MAGIC ARCH FILETYPE NUM CMDS SIZE CMDS FLAGS" |
| 959 | print "------------ ---------- -------------- -------- ---------- ----------" |
| 960 | |
| 961 | def dump_flat(self, options): |
| 962 | if options.verbose: |
| 963 | print "%#8.8x %#8.8x %#8.8x %#8.8x %#8u %#8.8x %#8.8x" % (self.magic, self.arch.cpu , self.arch.sub, self.filetype.value, self.ncmds, self.sizeofcmds, self.flags.bits) |
| 964 | else: |
| 965 | print "%-12s %-10s %-14s %#8u %#8.8x %s" % (self.magic, self.arch, self.filetype, self.ncmds, self.sizeofcmds, self.flags) |
| 966 | |
| 967 | def dump(self, options): |
| 968 | if options.dump_header: |
| 969 | self.dump_header(True, options) |
| 970 | if options.dump_load_commands: |
| 971 | self.dump_load_commands(False, options) |
| 972 | if options.dump_sections: |
| 973 | self.dump_sections(False, options) |
| 974 | if options.section_names: |
| 975 | self.dump_section_contents(options) |
| 976 | if options.dump_symtab: |
| 977 | self.get_symtab() |
| 978 | if len(self.symbols): |
| 979 | self.dump_sections(False, options) |
| 980 | else: |
| 981 | print "No symbols" |
| 982 | if options.find_mangled: |
| 983 | self.dump_symbol_names_matching_regex (re.compile('^_?_Z')) |
| 984 | |
| 985 | def dump_header(self, dump_description = True, options = None): |
| 986 | if dump_description: |
| 987 | print self.description() |
| 988 | print "Mach Header" |
| 989 | print " magic: %#8.8x %s" % (self.magic.value, self.magic) |
| 990 | print " cputype: %#8.8x %s" % (self.arch.cpu, self.arch) |
| 991 | print " cpusubtype: %#8.8x" % self.arch.sub |
| 992 | print " filetype: %#8.8x %s" % (self.filetype.get_enum_value(), self.filetype.get_enum_name()) |
| 993 | print " ncmds: %#8.8x %u" % (self.ncmds, self.ncmds) |
| 994 | print " sizeofcmds: %#8.8x" % self.sizeofcmds |
| 995 | print " flags: %#8.8x %s" % (self.flags.bits, self.flags) |
| 996 | |
| 997 | def dump_load_commands(self, dump_description = True, options = None): |
| 998 | if dump_description: |
| 999 | print self.description() |
| 1000 | for lc in self.commands: |
| 1001 | print lc |
| 1002 | |
| 1003 | def get_section_by_name (self, name): |
| 1004 | for section in self.sections: |
| 1005 | if section.sectname and section.sectname == name: |
| 1006 | return section |
| 1007 | return None |
| 1008 | |
| 1009 | def get_section_by_section (self, other_section): |
| 1010 | for section in self.sections: |
| 1011 | if section.sectname == other_section.sectname and section.segname == other_section.segname: |
| 1012 | return section |
| 1013 | return None |
| 1014 | |
| 1015 | def dump_sections(self, dump_description = True, options = None): |
| 1016 | if dump_description: |
| 1017 | print self.description() |
| 1018 | num_sections = len(self.sections) |
| 1019 | if num_sections > 1: |
| 1020 | self.sections[1].dump_header() |
| 1021 | for sect_idx in range(1,num_sections): |
| 1022 | print "%s" % self.sections[sect_idx] |
| 1023 | |
| 1024 | def dump_section_contents(self, options): |
| 1025 | saved_section_to_disk = False |
| 1026 | for sectname in options.section_names: |
| 1027 | section = self.get_section_by_name(sectname) |
| 1028 | if section: |
| 1029 | sect_bytes = section.get_contents (self) |
| 1030 | if options.outfile: |
| 1031 | if not saved_section_to_disk: |
| 1032 | outfile = open(options.outfile, 'w') |
| 1033 | if options.extract_modules: |
| 1034 | #print "Extracting modules from mach file..." |
| 1035 | data = file_extract.FileExtract(StringIO.StringIO(sect_bytes), self.data.byte_order) |
| 1036 | version = data.get_uint32() |
| 1037 | num_modules = data.get_uint32() |
| 1038 | #print "version = %u, num_modules = %u" % (version, num_modules) |
| 1039 | for i in range(num_modules): |
| 1040 | data_offset = data.get_uint64() |
| 1041 | data_size = data.get_uint64() |
| 1042 | name_offset = data.get_uint32() |
| 1043 | language = data.get_uint32() |
| 1044 | flags = data.get_uint32() |
| 1045 | data.seek (name_offset) |
| 1046 | module_name = data.get_c_string() |
| 1047 | #print "module[%u] data_offset = %#16.16x, data_size = %#16.16x, name_offset = %#16.16x (%s), language = %u, flags = %#x" % (i, data_offset, data_size, name_offset, module_name, language, flags) |
| 1048 | data.seek (data_offset) |
| 1049 | outfile.write(data.read_size (data_size)) |
| 1050 | else: |
| 1051 | print "Saving section %s to '%s'" % (sectname, options.outfile) |
| 1052 | outfile.write(sect_bytes) |
| 1053 | outfile.close() |
| 1054 | saved_section_to_disk = True |
| 1055 | else: |
| 1056 | print "error: you can only save a single section to disk at a time, skipping section '%s'" % (sectname) |
| 1057 | else: |
| 1058 | print 'section %s:\n' % (sectname) |
| 1059 | section.dump_header() |
| 1060 | print '%s\n' % (section) |
| 1061 | dump_memory (0, sect_bytes, options.max_count, 16) |
| 1062 | else: |
| 1063 | print 'error: no section named "%s" was found' % (sectname) |
| 1064 | |
| 1065 | def get_segment(self, segname): |
| 1066 | if len(self.segments) == 1 and self.segments[0].segname == '': |
| 1067 | return self.segments[0] |
| 1068 | for segment in self.segments: |
| 1069 | if segment.segname == segname: |
| 1070 | return segment |
| 1071 | return None |
| 1072 | |
| 1073 | def get_first_load_command(self, lc_enum_value): |
| 1074 | for lc in self.commands: |
| 1075 | if lc.command.value == lc_enum_value: |
| 1076 | return lc |
| 1077 | return None |
| 1078 | |
| 1079 | def get_symtab(self): |
| 1080 | if self.data and not self.symbols: |
| 1081 | lc_symtab = self.get_first_load_command (LC_SYMTAB) |
| 1082 | if lc_symtab: |
| 1083 | symtab_offset = self.file_off |
| 1084 | if self.data.is_in_memory(): |
| 1085 | linkedit_segment = self.get_segment('__LINKEDIT') |
| 1086 | if linkedit_segment: |
| 1087 | linkedit_vmaddr = linkedit_segment.vmaddr |
| 1088 | linkedit_fileoff = linkedit_segment.fileoff |
| 1089 | symtab_offset = linkedit_vmaddr + lc_symtab.symoff - linkedit_fileoff |
| 1090 | symtab_offset = linkedit_vmaddr + lc_symtab.stroff - linkedit_fileoff |
| 1091 | else: |
| 1092 | symtab_offset += lc_symtab.symoff |
| 1093 | |
| 1094 | self.data.seek (symtab_offset) |
| 1095 | is_64 = self.is_64_bit() |
| 1096 | for i in range(lc_symtab.nsyms): |
| 1097 | nlist = Mach.NList() |
| 1098 | nlist.unpack (self, self.data, lc_symtab) |
| 1099 | self.symbols.append(nlist) |
| 1100 | else: |
| 1101 | print "no LC_SYMTAB" |
| 1102 | |
| 1103 | def dump_symtab(self, dump_description = True, options = None): |
| 1104 | self.get_symtab() |
| 1105 | if dump_description: |
| 1106 | print self.description() |
| 1107 | for i, symbol in enumerate(self.symbols): |
| 1108 | print '[%5u] %s' % (i, symbol) |
| 1109 | |
| 1110 | def dump_symbol_names_matching_regex(self, regex, file=None): |
| 1111 | self.get_symtab() |
| 1112 | for symbol in self.symbols: |
| 1113 | if symbol.name and regex.search (symbol.name): |
| 1114 | print symbol.name |
| 1115 | if file: |
| 1116 | file.write('%s\n' % (symbol.name)) |
| 1117 | |
| 1118 | def is_64_bit(self): |
| 1119 | return self.magic.is_64_bit() |
| 1120 | |
| 1121 | class LoadCommand: |
| 1122 | class Command(dict_utils.Enum): |
| 1123 | enum = { |
| 1124 | 'LC_SEGMENT' : LC_SEGMENT, |
| 1125 | 'LC_SYMTAB' : LC_SYMTAB, |
| 1126 | 'LC_SYMSEG' : LC_SYMSEG, |
| 1127 | 'LC_THREAD' : LC_THREAD, |
| 1128 | 'LC_UNIXTHREAD' : LC_UNIXTHREAD, |
| 1129 | 'LC_LOADFVMLIB' : LC_LOADFVMLIB, |
| 1130 | 'LC_IDFVMLIB' : LC_IDFVMLIB, |
| 1131 | 'LC_IDENT' : LC_IDENT, |
| 1132 | 'LC_FVMFILE' : LC_FVMFILE, |
| 1133 | 'LC_PREPAGE' : LC_PREPAGE, |
| 1134 | 'LC_DYSYMTAB' : LC_DYSYMTAB, |
| 1135 | 'LC_LOAD_DYLIB' : LC_LOAD_DYLIB, |
| 1136 | 'LC_ID_DYLIB' : LC_ID_DYLIB, |
| 1137 | 'LC_LOAD_DYLINKER' : LC_LOAD_DYLINKER, |
| 1138 | 'LC_ID_DYLINKER' : LC_ID_DYLINKER, |
| 1139 | 'LC_PREBOUND_DYLIB' : LC_PREBOUND_DYLIB, |
| 1140 | 'LC_ROUTINES' : LC_ROUTINES, |
| 1141 | 'LC_SUB_FRAMEWORK' : LC_SUB_FRAMEWORK, |
| 1142 | 'LC_SUB_UMBRELLA' : LC_SUB_UMBRELLA, |
| 1143 | 'LC_SUB_CLIENT' : LC_SUB_CLIENT, |
| 1144 | 'LC_SUB_LIBRARY' : LC_SUB_LIBRARY, |
| 1145 | 'LC_TWOLEVEL_HINTS' : LC_TWOLEVEL_HINTS, |
| 1146 | 'LC_PREBIND_CKSUM' : LC_PREBIND_CKSUM, |
| 1147 | 'LC_LOAD_WEAK_DYLIB' : LC_LOAD_WEAK_DYLIB, |
| 1148 | 'LC_SEGMENT_64' : LC_SEGMENT_64, |
| 1149 | 'LC_ROUTINES_64' : LC_ROUTINES_64, |
| 1150 | 'LC_UUID' : LC_UUID, |
| 1151 | 'LC_RPATH' : LC_RPATH, |
| 1152 | 'LC_CODE_SIGNATURE' : LC_CODE_SIGNATURE, |
| 1153 | 'LC_SEGMENT_SPLIT_INFO' : LC_SEGMENT_SPLIT_INFO, |
| 1154 | 'LC_REEXPORT_DYLIB' : LC_REEXPORT_DYLIB, |
| 1155 | 'LC_LAZY_LOAD_DYLIB' : LC_LAZY_LOAD_DYLIB, |
| 1156 | 'LC_ENCRYPTION_INFO' : LC_ENCRYPTION_INFO, |
| 1157 | 'LC_DYLD_INFO' : LC_DYLD_INFO, |
| 1158 | 'LC_DYLD_INFO_ONLY' : LC_DYLD_INFO_ONLY, |
| 1159 | 'LC_LOAD_UPWARD_DYLIB' : LC_LOAD_UPWARD_DYLIB, |
| 1160 | 'LC_VERSION_MIN_MACOSX' : LC_VERSION_MIN_MACOSX, |
| 1161 | 'LC_VERSION_MIN_IPHONEOS' : LC_VERSION_MIN_IPHONEOS, |
| 1162 | 'LC_FUNCTION_STARTS' : LC_FUNCTION_STARTS, |
| 1163 | 'LC_DYLD_ENVIRONMENT' : LC_DYLD_ENVIRONMENT |
| 1164 | } |
| 1165 | |
| 1166 | def __init__(self, initial_value = 0): |
| 1167 | dict_utils.Enum.__init__(self, initial_value, self.enum) |
| 1168 | |
| 1169 | |
| 1170 | def __init__(self, c=None, l=0,o=0): |
| 1171 | if c != None: |
| 1172 | self.command = c |
| 1173 | else: |
| 1174 | self.command = Mach.LoadCommand.Command(0) |
| 1175 | self.length = l |
| 1176 | self.file_off = o |
| 1177 | |
| 1178 | def unpack(self, mach_file, data): |
| 1179 | self.file_off = data.tell() |
| 1180 | self.command.value, self.length = data.get_n_uint32(2) |
| 1181 | |
| 1182 | def skip(self, data): |
| 1183 | data.seek (self.file_off + self.length, 0) |
| 1184 | |
| 1185 | def __str__(self): |
| 1186 | lc_name = self.command.get_enum_name() |
| 1187 | return '%#8.8x: <%#4.4x> %-24s' % (self.file_off, self.length, lc_name) |
| 1188 | |
| 1189 | class Section: |
| 1190 | |
| 1191 | def __init__(self): |
| 1192 | self.index = 0 |
| 1193 | self.is_64 = False |
| 1194 | self.sectname = None |
| 1195 | self.segname = None |
| 1196 | self.addr = 0 |
| 1197 | self.size = 0 |
| 1198 | self.offset = 0 |
| 1199 | self.align = 0 |
| 1200 | self.reloff = 0 |
| 1201 | self.nreloc = 0 |
| 1202 | self.flags = 0 |
| 1203 | self.reserved1 = 0 |
| 1204 | self.reserved2 = 0 |
| 1205 | self.reserved3 = 0 |
| 1206 | |
| 1207 | def unpack(self, is_64, data): |
| 1208 | self.is_64 = is_64 |
| 1209 | self.sectname = data.get_fixed_length_c_string (16, '', True) |
| 1210 | self.segname = data.get_fixed_length_c_string (16, '', True) |
| 1211 | if self.is_64: |
| 1212 | self.addr, self.size = data.get_n_uint64(2) |
| 1213 | self.offset, self.align, self.reloff, self.nreloc, self.flags, self.reserved1, self.reserved2, self.reserved3 = data.get_n_uint32(8) |
| 1214 | else: |
| 1215 | self.addr, self.size = data.get_n_uint32(2) |
| 1216 | self.offset, self.align, self.reloff, self.nreloc, self.flags, self.reserved1, self.reserved2 = data.get_n_uint32(7) |
| 1217 | |
| 1218 | def dump_header(self): |
| 1219 | if self.is_64: |
| 1220 | print "INDEX ADDRESS SIZE OFFSET ALIGN RELOFF NRELOC FLAGS RESERVED1 RESERVED2 RESERVED3 NAME"; |
| 1221 | print "===== ------------------ ------------------ ---------- ---------- ---------- ---------- ---------- ---------- ---------- ---------- ----------------------"; |
| 1222 | else: |
| 1223 | print "INDEX ADDRESS SIZE OFFSET ALIGN RELOFF NRELOC FLAGS RESERVED1 RESERVED2 NAME"; |
| 1224 | print "===== ---------- ---------- ---------- ---------- ---------- ---------- ---------- ---------- ---------- ----------------------"; |
| 1225 | |
| 1226 | def __str__(self): |
| 1227 | if self.is_64: |
| 1228 | return "[%3u] %#16.16x %#16.16x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %s.%s" % (self.index, self.addr, self.size, self.offset, self.align, self.reloff, self.nreloc, self.flags, self.reserved1, self.reserved2, self.reserved3, self.segname, self.sectname) |
| 1229 | else: |
| 1230 | return "[%3u] %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %#8.8x %s.%s" % (self.index, self.addr, self.size, self.offset, self.align, self.reloff, self.nreloc, self.flags, self.reserved1, self.reserved2, self.segname, self.sectname) |
| 1231 | |
| 1232 | def get_contents(self, mach_file): |
| 1233 | '''Get the section contents as a python string''' |
| 1234 | if self.size > 0 and mach_file.get_segment(self.segname).filesize > 0: |
| 1235 | data = mach_file.get_data() |
| 1236 | if data: |
| 1237 | section_data_offset = mach_file.file_off + self.offset |
| 1238 | #print '%s.%s is at offset 0x%x with size 0x%x' % (self.segname, self.sectname, section_data_offset, self.size) |
| 1239 | data.push_offset_and_seek (section_data_offset) |
| 1240 | bytes = data.read_size(self.size) |
| 1241 | data.pop_offset_and_seek() |
| 1242 | return bytes |
| 1243 | return None |
| 1244 | |
| 1245 | class DylibLoadCommand(LoadCommand): |
| 1246 | def __init__(self, lc): |
| 1247 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1248 | self.name = None |
| 1249 | self.timestamp = 0 |
| 1250 | self.current_version = 0 |
| 1251 | self.compatibility_version = 0 |
| 1252 | |
| 1253 | def unpack(self, mach_file, data): |
| 1254 | byte_order_char = mach_file.magic.get_byte_order() |
| 1255 | name_offset, self.timestamp, self.current_version, self.compatibility_version = data.get_n_uint32(4) |
| 1256 | data.seek(self.file_off + name_offset, 0) |
| 1257 | self.name = data.get_fixed_length_c_string(self.length - 24) |
| 1258 | |
| 1259 | def __str__(self): |
| 1260 | s = Mach.LoadCommand.__str__(self); |
| 1261 | s += "%#8.8x %#8.8x %#8.8x " % (self.timestamp, self.current_version, self.compatibility_version) |
| 1262 | s += self.name |
| 1263 | return s |
| 1264 | |
| 1265 | class LoadDYLDLoadCommand(LoadCommand): |
| 1266 | def __init__(self, lc): |
| 1267 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1268 | self.name = None |
| 1269 | |
| 1270 | def unpack(self, mach_file, data): |
| 1271 | data.get_uint32() |
| 1272 | self.name = data.get_fixed_length_c_string(self.length - 12) |
| 1273 | |
| 1274 | def __str__(self): |
| 1275 | s = Mach.LoadCommand.__str__(self); |
| 1276 | s += "%s" % self.name |
| 1277 | return s |
| 1278 | |
| 1279 | class UnixThreadLoadCommand(LoadCommand): |
| 1280 | class ThreadState: |
| 1281 | def __init__(self): |
| 1282 | self.flavor = 0 |
| 1283 | self.count = 0 |
| 1284 | self.register_values = list() |
| 1285 | |
| 1286 | def unpack(self, data): |
| 1287 | self.flavor, self.count = data.get_n_uint32(2) |
| 1288 | self.register_values = data.get_n_uint32(self.count) |
| 1289 | |
| 1290 | def __str__(self): |
| 1291 | s = "flavor = %u, count = %u, regs =" % (self.flavor, self.count) |
| 1292 | i = 0 |
| 1293 | for register_value in self.register_values: |
| 1294 | if i % 8 == 0: |
| 1295 | s += "\n " |
| 1296 | s += " %#8.8x" % register_value |
| 1297 | i += 1 |
| 1298 | return s |
| 1299 | |
| 1300 | def __init__(self, lc): |
| 1301 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1302 | self.reg_sets = list() |
| 1303 | |
| 1304 | def unpack(self, mach_file, data): |
| 1305 | reg_set = Mach.UnixThreadLoadCommand.ThreadState() |
| 1306 | reg_set.unpack (data) |
| 1307 | self.reg_sets.append(reg_set) |
| 1308 | |
| 1309 | def __str__(self): |
| 1310 | s = Mach.LoadCommand.__str__(self); |
| 1311 | for reg_set in self.reg_sets: |
| 1312 | s += "%s" % reg_set |
| 1313 | return s |
| 1314 | |
| 1315 | class DYLDInfoOnlyLoadCommand(LoadCommand): |
| 1316 | def __init__(self, lc): |
| 1317 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1318 | self.rebase_off = 0 |
| 1319 | self.rebase_size = 0 |
| 1320 | self.bind_off = 0 |
| 1321 | self.bind_size = 0 |
| 1322 | self.weak_bind_off = 0 |
| 1323 | self.weak_bind_size = 0 |
| 1324 | self.lazy_bind_off = 0 |
| 1325 | self.lazy_bind_size = 0 |
| 1326 | self.export_off = 0 |
| 1327 | self.export_size = 0 |
| 1328 | |
| 1329 | def unpack(self, mach_file, data): |
| 1330 | byte_order_char = mach_file.magic.get_byte_order() |
| 1331 | self.rebase_off, self.rebase_size, self.bind_off, self.bind_size, self.weak_bind_off, self.weak_bind_size, self.lazy_bind_off, self.lazy_bind_size, self.export_off, self.export_size = data.get_n_uint32(10) |
| 1332 | |
| 1333 | def __str__(self): |
| 1334 | s = Mach.LoadCommand.__str__(self); |
| 1335 | s += "rebase_off = %#8.8x, rebase_size = %u, " % (self.rebase_off, self.rebase_size) |
| 1336 | s += "bind_off = %#8.8x, bind_size = %u, " % (self.bind_off, self.bind_size) |
| 1337 | s += "weak_bind_off = %#8.8x, weak_bind_size = %u, " % (self.weak_bind_off, self.weak_bind_size) |
| 1338 | s += "lazy_bind_off = %#8.8x, lazy_bind_size = %u, " % (self.lazy_bind_off, self.lazy_bind_size) |
| 1339 | s += "export_off = %#8.8x, export_size = %u, " % (self.export_off, self.export_size) |
| 1340 | return s |
| 1341 | |
| 1342 | class DYLDSymtabLoadCommand(LoadCommand): |
| 1343 | def __init__(self, lc): |
| 1344 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1345 | self.ilocalsym = 0 |
| 1346 | self.nlocalsym = 0 |
| 1347 | self.iextdefsym = 0 |
| 1348 | self.nextdefsym = 0 |
| 1349 | self.iundefsym = 0 |
| 1350 | self.nundefsym = 0 |
| 1351 | self.tocoff = 0 |
| 1352 | self.ntoc = 0 |
| 1353 | self.modtaboff = 0 |
| 1354 | self.nmodtab = 0 |
| 1355 | self.extrefsymoff = 0 |
| 1356 | self.nextrefsyms = 0 |
| 1357 | self.indirectsymoff = 0 |
| 1358 | self.nindirectsyms = 0 |
| 1359 | self.extreloff = 0 |
| 1360 | self.nextrel = 0 |
| 1361 | self.locreloff = 0 |
| 1362 | self.nlocrel = 0 |
| 1363 | |
| 1364 | def unpack(self, mach_file, data): |
| 1365 | byte_order_char = mach_file.magic.get_byte_order() |
| 1366 | self.ilocalsym, self.nlocalsym, self.iextdefsym, self.nextdefsym, self.iundefsym, self.nundefsym, self.tocoff, self.ntoc, self.modtaboff, self.nmodtab, self.extrefsymoff, self.nextrefsyms, self.indirectsymoff, self.nindirectsyms, self.extreloff, self.nextrel, self.locreloff, self.nlocrel = data.get_n_uint32(18) |
| 1367 | |
| 1368 | def __str__(self): |
| 1369 | s = Mach.LoadCommand.__str__(self); |
| 1370 | # s += "ilocalsym = %u, nlocalsym = %u, " % (self.ilocalsym, self.nlocalsym) |
| 1371 | # s += "iextdefsym = %u, nextdefsym = %u, " % (self.iextdefsym, self.nextdefsym) |
| 1372 | # s += "iundefsym %u, nundefsym = %u, " % (self.iundefsym, self.nundefsym) |
| 1373 | # s += "tocoff = %#8.8x, ntoc = %u, " % (self.tocoff, self.ntoc) |
| 1374 | # s += "modtaboff = %#8.8x, nmodtab = %u, " % (self.modtaboff, self.nmodtab) |
| 1375 | # s += "extrefsymoff = %#8.8x, nextrefsyms = %u, " % (self.extrefsymoff, self.nextrefsyms) |
| 1376 | # s += "indirectsymoff = %#8.8x, nindirectsyms = %u, " % (self.indirectsymoff, self.nindirectsyms) |
| 1377 | # s += "extreloff = %#8.8x, nextrel = %u, " % (self.extreloff, self.nextrel) |
| 1378 | # s += "locreloff = %#8.8x, nlocrel = %u" % (self.locreloff, self.nlocrel) |
| 1379 | s += "ilocalsym = %-10u, nlocalsym = %u\n" % (self.ilocalsym, self.nlocalsym) |
| 1380 | s += " iextdefsym = %-10u, nextdefsym = %u\n" % (self.iextdefsym, self.nextdefsym) |
| 1381 | s += " iundefsym = %-10u, nundefsym = %u\n" % (self.iundefsym, self.nundefsym) |
| 1382 | s += " tocoff = %#8.8x, ntoc = %u\n" % (self.tocoff, self.ntoc) |
| 1383 | s += " modtaboff = %#8.8x, nmodtab = %u\n" % (self.modtaboff, self.nmodtab) |
| 1384 | s += " extrefsymoff = %#8.8x, nextrefsyms = %u\n" % (self.extrefsymoff, self.nextrefsyms) |
| 1385 | s += " indirectsymoff = %#8.8x, nindirectsyms = %u\n" % (self.indirectsymoff, self.nindirectsyms) |
| 1386 | s += " extreloff = %#8.8x, nextrel = %u\n" % (self.extreloff, self.nextrel) |
| 1387 | s += " locreloff = %#8.8x, nlocrel = %u" % (self.locreloff, self.nlocrel) |
| 1388 | return s |
| 1389 | |
| 1390 | class SymtabLoadCommand(LoadCommand): |
| 1391 | def __init__(self, lc): |
| 1392 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1393 | self.symoff = 0 |
| 1394 | self.nsyms = 0 |
| 1395 | self.stroff = 0 |
| 1396 | self.strsize = 0 |
| 1397 | |
| 1398 | def unpack(self, mach_file, data): |
| 1399 | byte_order_char = mach_file.magic.get_byte_order() |
| 1400 | self.symoff, self.nsyms, self.stroff, self.strsize = data.get_n_uint32(4) |
| 1401 | |
| 1402 | def __str__(self): |
| 1403 | s = Mach.LoadCommand.__str__(self); |
| 1404 | s += "symoff = %#8.8x, nsyms = %u, stroff = %#8.8x, strsize = %u" % (self.symoff, self.nsyms, self.stroff, self.strsize) |
| 1405 | return s |
| 1406 | |
| 1407 | |
| 1408 | class UUIDLoadCommand(LoadCommand): |
| 1409 | def __init__(self, lc): |
| 1410 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1411 | self.uuid = None |
| 1412 | |
| 1413 | def unpack(self, mach_file, data): |
| 1414 | uuid_data = data.get_n_uint8(16) |
| 1415 | uuid_str = '' |
| 1416 | for byte in uuid_data: |
| 1417 | uuid_str += '%2.2x' % byte |
| 1418 | self.uuid = uuid.UUID(uuid_str) |
| 1419 | mach_file.uuid = self.uuid |
| 1420 | |
| 1421 | def __str__(self): |
| 1422 | s = Mach.LoadCommand.__str__(self); |
| 1423 | s += self.uuid.__str__() |
| 1424 | return s |
| 1425 | |
| 1426 | class DataBlobLoadCommand(LoadCommand): |
| 1427 | def __init__(self, lc): |
| 1428 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1429 | self.dataoff = 0 |
| 1430 | self.datasize = 0 |
| 1431 | |
| 1432 | def unpack(self, mach_file, data): |
| 1433 | byte_order_char = mach_file.magic.get_byte_order() |
| 1434 | self.dataoff, self.datasize = data.get_n_uint32(2) |
| 1435 | |
| 1436 | def __str__(self): |
| 1437 | s = Mach.LoadCommand.__str__(self); |
| 1438 | s += "dataoff = %#8.8x, datasize = %u" % (self.dataoff, self.datasize) |
| 1439 | return s |
| 1440 | |
| 1441 | class EncryptionInfoLoadCommand(LoadCommand): |
| 1442 | def __init__(self, lc): |
| 1443 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1444 | self.cryptoff = 0 |
| 1445 | self.cryptsize = 0 |
| 1446 | self.cryptid = 0 |
| 1447 | |
| 1448 | def unpack(self, mach_file, data): |
| 1449 | byte_order_char = mach_file.magic.get_byte_order() |
| 1450 | self.cryptoff, self.cryptsize, self.cryptid = data.get_n_uint32(3) |
| 1451 | |
| 1452 | def __str__(self): |
| 1453 | s = Mach.LoadCommand.__str__(self); |
| 1454 | s += "file-range = [%#8.8x - %#8.8x), cryptsize = %u, cryptid = %u" % (self.cryptoff, self.cryptoff + self.cryptsize, self.cryptsize, self.cryptid) |
| 1455 | return s |
| 1456 | |
| 1457 | class SegmentLoadCommand(LoadCommand): |
| 1458 | |
| 1459 | def __init__(self, lc): |
| 1460 | Mach.LoadCommand.__init__(self, lc.command, lc.length, lc.file_off) |
| 1461 | self.segname = None |
| 1462 | self.vmaddr = 0 |
| 1463 | self.vmsize = 0 |
| 1464 | self.fileoff = 0 |
| 1465 | self.filesize = 0 |
| 1466 | self.maxprot = 0 |
| 1467 | self.initprot = 0 |
| 1468 | self.nsects = 0 |
| 1469 | self.flags = 0 |
| 1470 | |
| 1471 | def unpack(self, mach_file, data): |
| 1472 | is_64 = self.command.get_enum_value() == LC_SEGMENT_64; |
| 1473 | self.segname = data.get_fixed_length_c_string (16, '', True) |
| 1474 | if is_64: |
| 1475 | self.vmaddr, self.vmsize, self.fileoff, self.filesize = data.get_n_uint64(4) |
| 1476 | else: |
| 1477 | self.vmaddr, self.vmsize, self.fileoff, self.filesize = data.get_n_uint32(4) |
| 1478 | self.maxprot, self.initprot, self.nsects, self.flags = data.get_n_uint32(4) |
| 1479 | mach_file.segments.append(self) |
| 1480 | for i in range(self.nsects): |
| 1481 | section = Mach.Section() |
| 1482 | section.unpack(is_64, data) |
| 1483 | section.index = len (mach_file.sections) |
| 1484 | mach_file.sections.append(section) |
| 1485 | |
| 1486 | |
| 1487 | def __str__(self): |
| 1488 | s = Mach.LoadCommand.__str__(self); |
| 1489 | if self.command.get_enum_value() == LC_SEGMENT: |
| 1490 | s += "%#8.8x %#8.8x %#8.8x %#8.8x " % (self.vmaddr, self.vmsize, self.fileoff, self.filesize) |
| 1491 | else: |
| 1492 | s += "%#16.16x %#16.16x %#16.16x %#16.16x " % (self.vmaddr, self.vmsize, self.fileoff, self.filesize) |
| 1493 | s += "%s %s %3u %#8.8x" % (vm_prot_names[self.maxprot], vm_prot_names[self.initprot], self.nsects, self.flags) |
| 1494 | s += ' ' + self.segname |
| 1495 | return s |
| 1496 | |
| 1497 | class NList: |
| 1498 | class Type: |
| 1499 | class Stab(dict_utils.Enum): |
| 1500 | enum = { |
| 1501 | 'N_GSYM' : N_GSYM , |
| 1502 | 'N_FNAME' : N_FNAME , |
| 1503 | 'N_FUN' : N_FUN , |
| 1504 | 'N_STSYM' : N_STSYM , |
| 1505 | 'N_LCSYM' : N_LCSYM , |
| 1506 | 'N_BNSYM' : N_BNSYM , |
| 1507 | 'N_OPT' : N_OPT , |
| 1508 | 'N_RSYM' : N_RSYM , |
| 1509 | 'N_SLINE' : N_SLINE , |
| 1510 | 'N_ENSYM' : N_ENSYM , |
| 1511 | 'N_SSYM' : N_SSYM , |
| 1512 | 'N_SO' : N_SO , |
| 1513 | 'N_OSO' : N_OSO , |
| 1514 | 'N_LSYM' : N_LSYM , |
| 1515 | 'N_BINCL' : N_BINCL , |
| 1516 | 'N_SOL' : N_SOL , |
| 1517 | 'N_PARAMS' : N_PARAMS , |
| 1518 | 'N_VERSION' : N_VERSION , |
| 1519 | 'N_OLEVEL' : N_OLEVEL , |
| 1520 | 'N_PSYM' : N_PSYM , |
| 1521 | 'N_EINCL' : N_EINCL , |
| 1522 | 'N_ENTRY' : N_ENTRY , |
| 1523 | 'N_LBRAC' : N_LBRAC , |
| 1524 | 'N_EXCL' : N_EXCL , |
| 1525 | 'N_RBRAC' : N_RBRAC , |
| 1526 | 'N_BCOMM' : N_BCOMM , |
| 1527 | 'N_ECOMM' : N_ECOMM , |
| 1528 | 'N_ECOML' : N_ECOML , |
| 1529 | 'N_LENG' : N_LENG |
| 1530 | } |
| 1531 | |
| 1532 | def __init__(self, magic = 0): |
| 1533 | dict_utils.Enum.__init__(self, magic, self.enum) |
| 1534 | |
| 1535 | def __init__(self, t = 0): |
| 1536 | self.value = t |
| 1537 | |
| 1538 | def __str__(self): |
| 1539 | n_type = self.value |
| 1540 | if n_type & N_STAB: |
| 1541 | stab = Mach.NList.Type.Stab(self.value) |
| 1542 | return '%s' % stab |
| 1543 | else: |
| 1544 | type = self.value & N_TYPE |
| 1545 | type_str = '' |
| 1546 | if type == N_UNDF: |
| 1547 | type_str = 'N_UNDF' |
| 1548 | elif type == N_ABS: |
| 1549 | type_str = 'N_ABS ' |
| 1550 | elif type == N_SECT: |
| 1551 | type_str = 'N_SECT' |
| 1552 | elif type == N_PBUD: |
| 1553 | type_str = 'N_PBUD' |
| 1554 | elif type == N_INDR: |
| 1555 | type_str = 'N_INDR' |
| 1556 | else: |
| 1557 | type_str = "??? (%#2.2x)" % type |
| 1558 | if n_type & N_PEXT: |
| 1559 | type_str += ' | PEXT' |
| 1560 | if n_type & N_EXT: |
| 1561 | type_str += ' | EXT ' |
| 1562 | return type_str |
| 1563 | |
| 1564 | |
| 1565 | def __init__(self): |
| 1566 | self.index = 0 |
| 1567 | self.name_offset = 0 |
| 1568 | self.name = 0 |
| 1569 | self.type = Mach.NList.Type() |
| 1570 | self.sect_idx = 0 |
| 1571 | self.desc = 0 |
| 1572 | self.value = 0 |
| 1573 | |
| 1574 | def unpack(self, mach_file, data, symtab_lc): |
| 1575 | self.index = len(mach_file.symbols) |
| 1576 | self.name_offset = data.get_uint32() |
| 1577 | self.type.value, self.sect_idx = data.get_n_uint8(2) |
| 1578 | self.desc = data.get_uint16() |
| 1579 | if mach_file.is_64_bit(): |
| 1580 | self.value = data.get_uint64() |
| 1581 | else: |
| 1582 | self.value = data.get_uint32() |
| 1583 | data.push_offset_and_seek (mach_file.file_off + symtab_lc.stroff + self.name_offset) |
| 1584 | #print "get string for symbol[%u]" % self.index |
| 1585 | self.name = data.get_c_string() |
| 1586 | data.pop_offset_and_seek() |
| 1587 | |
| 1588 | def __str__(self): |
| 1589 | name_display = '' |
| 1590 | if len(self.name): |
| 1591 | name_display = ' "%s"' % self.name |
| 1592 | return '%#8.8x %#2.2x (%-20s) %#2.2x %#4.4x %16.16x%s' % (self.name_offset, self.type.value, self.type, self.sect_idx, self.desc, self.value, name_display) |
| 1593 | |
| 1594 | |
| 1595 | class Interactive(cmd.Cmd): |
| 1596 | '''Interactive command interpreter to mach-o files.''' |
| 1597 | |
| 1598 | def __init__(self, mach, options): |
| 1599 | cmd.Cmd.__init__(self) |
| 1600 | self.intro = 'Interactive mach-o command interpreter' |
| 1601 | self.prompt = 'mach-o: %s %% ' % mach.path |
| 1602 | self.mach = mach |
| 1603 | self.options = options |
| 1604 | |
| 1605 | def default(self, line): |
| 1606 | '''Catch all for unknown command, which will exit the interpreter.''' |
| 1607 | print "uknown command: %s" % line |
| 1608 | return True |
| 1609 | |
| 1610 | def do_q(self, line): |
| 1611 | '''Quit command''' |
| 1612 | return True |
| 1613 | |
| 1614 | def do_quit(self, line): |
| 1615 | '''Quit command''' |
| 1616 | return True |
| 1617 | |
| 1618 | def do_header(self, line): |
| 1619 | '''Dump mach-o file headers''' |
| 1620 | self.mach.dump_header(True, self.options) |
| 1621 | return False |
| 1622 | |
| 1623 | def do_load(self, line): |
| 1624 | '''Dump all mach-o load commands''' |
| 1625 | self.mach.dump_load_commands(True, self.options) |
| 1626 | return False |
| 1627 | |
| 1628 | def do_sections(self, line): |
| 1629 | '''Dump all mach-o sections''' |
| 1630 | self.mach.dump_sections(True, self.options) |
| 1631 | return False |
| 1632 | |
| 1633 | def do_symtab(self, line): |
| 1634 | '''Dump all mach-o symbols in the symbol table''' |
| 1635 | self.mach.dump_symtab(True, self.options) |
| 1636 | return False |
| 1637 | |
| 1638 | if __name__ == '__main__': |
| 1639 | parser = optparse.OptionParser(description='A script that parses skinny and universal mach-o files.') |
| 1640 | parser.add_option('--arch', '-a', type='string', metavar='arch', dest='archs', action='append', help='specify one or more architectures by name') |
| 1641 | parser.add_option('-v', '--verbose', action='store_true', dest='verbose', help='display verbose debug info', default=False) |
| 1642 | parser.add_option('-H', '--header', action='store_true', dest='dump_header', help='dump the mach-o file header', default=False) |
| 1643 | parser.add_option('-l', '--load-commands', action='store_true', dest='dump_load_commands', help='dump the mach-o load commands', default=False) |
| 1644 | parser.add_option('-s', '--symtab', action='store_true', dest='dump_symtab', help='dump the mach-o symbol table', default=False) |
| 1645 | parser.add_option('-S', '--sections', action='store_true', dest='dump_sections', help='dump the mach-o sections', default=False) |
| 1646 | parser.add_option('--section', type='string', metavar='sectname', dest='section_names', action='append', help='Specify one or more section names to dump', default=[]) |
| 1647 | parser.add_option('-o', '--out', type='string', dest='outfile', help='Used in conjunction with the --section=NAME option to save a single section\'s data to disk.', default=False) |
| 1648 | parser.add_option('-i', '--interactive', action='store_true', dest='interactive', help='enable interactive mode', default=False) |
| 1649 | parser.add_option('-m', '--mangled', action='store_true', dest='find_mangled', help='dump all mangled names in a mach file', default=False) |
| 1650 | parser.add_option('-c', '--compare', action='store_true', dest='compare', help='compare two mach files', default=False) |
| 1651 | parser.add_option('-M', '--extract-modules', action='store_true', dest='extract_modules', help='Extract modules from file', default=False) |
| 1652 | parser.add_option('-C', '--count', type='int', dest='max_count', help='Sets the max byte count when dumping section data', default=-1) |
| 1653 | |
| 1654 | (options, mach_files) = parser.parse_args() |
| 1655 | if options.extract_modules: |
| 1656 | if options.section_names: |
| 1657 | print "error: can't use --section option with the --extract-modules option" |
| 1658 | exit(1) |
| 1659 | if not options.outfile: |
| 1660 | print "error: the --output=FILE option must be specified with the --extract-modules option" |
| 1661 | exit(1) |
| 1662 | options.section_names.append("__apple_ast") |
| 1663 | if options.compare: |
| 1664 | if len(mach_files) == 2: |
| 1665 | mach_a = Mach() |
| 1666 | mach_b = Mach() |
| 1667 | mach_a.parse(mach_files[0]) |
| 1668 | mach_b.parse(mach_files[1]) |
| 1669 | mach_a.compare(mach_b) |
| 1670 | else: |
| 1671 | print 'error: --compare takes two mach files as arguments' |
| 1672 | else: |
| 1673 | if not (options.dump_header or options.dump_load_commands or options.dump_symtab or options.dump_sections or options.find_mangled or options.section_names): |
| 1674 | options.dump_header = True |
| 1675 | options.dump_load_commands = True |
| 1676 | if options.verbose: |
| 1677 | print 'options', options |
| 1678 | print 'mach_files', mach_files |
| 1679 | for path in mach_files: |
| 1680 | mach = Mach() |
| 1681 | mach.parse(path) |
| 1682 | if options.interactive: |
| 1683 | interpreter = Mach.Interactive(mach, options) |
| 1684 | interpreter.cmdloop() |
| 1685 | else: |
| 1686 | mach.dump(options) |
| 1687 | |