| author | |
| committer | |
| log | 5c1ec20c9a5c788af9f3402b9a065389eb818e76 |
| tree | d51908c173a8e6201fca9c0acf88aed5f9b15e7a |
| parent | 8f96553be81829f1290213d8657d9bc61394b1f7 |
...rather than trying to find the executable on the file system.
Also use a more robust PIE offset calculation based on the
available metadata.
And for the last function, use the data that tells the end
rather than assuming 4K.
Also they print in a consistent way with Linux stack traces.4 files changed, 605 insertions(+), 284 deletions(-)
CMakeLists.txt-1| ... | ... | @@ -485,7 +485,6 @@ set(ZIG_STD_FILES |
| 485 | 485 | "json.zig" |
| 486 | 486 | "lazy_init.zig" |
| 487 | 487 | "linked_list.zig" |
| 488 | "macho.zig" | |
| 489 | 488 | "math/acos.zig" |
| 490 | 489 | "math/acosh.zig" |
| 491 | 490 | "math/asin.zig" |
std/c/darwin.zig+354| ... | ... | @@ -1,5 +1,6 @@ |
| 1 | 1 | extern "c" fn __error() *c_int; |
| 2 | 2 | pub extern "c" fn _NSGetExecutablePath(buf: [*]u8, bufsize: *u32) c_int; |
| 3 | pub extern "c" fn _dyld_get_image_header(image_index: u32) ?*mach_header; | |
| 3 | 4 | |
| 4 | 5 | pub extern "c" fn __getdirentries64(fd: c_int, buf_ptr: [*]u8, buf_len: usize, basep: *i64) usize; |
| 5 | 6 | |
| ... | ... | @@ -33,6 +34,12 @@ pub extern "c" fn sysctlnametomib(name: [*]const u8, mibp: ?*c_int, sizep: ?*usi |
| 33 | 34 | pub extern "c" fn bind(socket: c_int, address: ?*const sockaddr, address_len: socklen_t) c_int; |
| 34 | 35 | pub extern "c" fn socket(domain: c_int, type: c_int, protocol: c_int) c_int; |
| 35 | 36 | |
| 37 | /// The value of the link editor defined symbol _MH_EXECUTE_SYM is the address | |
| 38 | /// of the mach header in a Mach-O executable file type. It does not appear in | |
| 39 | /// any file type other than a MH_EXECUTE file type. The type of the symbol is | |
| 40 | /// absolute as the header is not part of any section. | |
| 41 | pub extern "c" var _mh_execute_header: if (@sizeOf(usize) == 8) mach_header_64 else mach_header; | |
| 42 | ||
| 36 | 43 | pub use @import("../os/darwin/errno.zig"); |
| 37 | 44 | |
| 38 | 45 | pub const _errno = __error; |
| ... | ... | @@ -139,6 +146,353 @@ pub const Kevent = extern struct { |
| 139 | 146 | udata: usize, |
| 140 | 147 | }; |
| 141 | 148 | |
| 149 | pub const mach_header = extern struct { | |
| 150 | magic: u32, | |
| 151 | cputype: cpu_type_t, | |
| 152 | cpusubtype: cpu_subtype_t, | |
| 153 | filetype: u32, | |
| 154 | ncmds: u32, | |
| 155 | sizeofcmds: u32, | |
| 156 | flags: u32, | |
| 157 | }; | |
| 158 | ||
| 159 | pub const mach_header_64 = extern struct { | |
| 160 | magic: u32, | |
| 161 | cputype: cpu_type_t, | |
| 162 | cpusubtype: cpu_subtype_t, | |
| 163 | filetype: u32, | |
| 164 | ncmds: u32, | |
| 165 | sizeofcmds: u32, | |
| 166 | flags: u32, | |
| 167 | reserved: u32, | |
| 168 | }; | |
| 169 | ||
| 170 | pub const load_command = extern struct { | |
| 171 | cmd: u32, | |
| 172 | cmdsize: u32, | |
| 173 | }; | |
| 174 | ||
| 175 | ||
| 176 | /// The symtab_command contains the offsets and sizes of the link-edit 4.3BSD | |
| 177 | /// "stab" style symbol table information as described in the header files | |
| 178 | /// <nlist.h> and <stab.h>. | |
| 179 | pub const symtab_command = extern struct { | |
| 180 | cmd: u32, /// LC_SYMTAB | |
| 181 | cmdsize: u32, /// sizeof(struct symtab_command) | |
| 182 | symoff: u32, /// symbol table offset | |
| 183 | nsyms: u32, /// number of symbol table entries | |
| 184 | stroff: u32, /// string table offset | |
| 185 | strsize: u32, /// string table size in bytes | |
| 186 | }; | |
| 187 | ||
| 188 | /// The linkedit_data_command contains the offsets and sizes of a blob | |
| 189 | /// of data in the __LINKEDIT segment. | |
| 190 | const linkedit_data_command = extern struct { | |
| 191 | cmd: u32,/// LC_CODE_SIGNATURE, LC_SEGMENT_SPLIT_INFO, LC_FUNCTION_STARTS, LC_DATA_IN_CODE, LC_DYLIB_CODE_SIGN_DRS or LC_LINKER_OPTIMIZATION_HINT. | |
| 192 | cmdsize: u32, /// sizeof(struct linkedit_data_command) | |
| 193 | dataoff: u32 , /// file offset of data in __LINKEDIT segment | |
| 194 | datasize: u32 , /// file size of data in __LINKEDIT segment | |
| 195 | }; | |
| 196 | ||
| 197 | /// The segment load command indicates that a part of this file is to be | |
| 198 | /// mapped into the task's address space. The size of this segment in memory, | |
| 199 | /// vmsize, maybe equal to or larger than the amount to map from this file, | |
| 200 | /// filesize. The file is mapped starting at fileoff to the beginning of | |
| 201 | /// the segment in memory, vmaddr. The rest of the memory of the segment, | |
| 202 | /// if any, is allocated zero fill on demand. The segment's maximum virtual | |
| 203 | /// memory protection and initial virtual memory protection are specified | |
| 204 | /// by the maxprot and initprot fields. If the segment has sections then the | |
| 205 | /// section structures directly follow the segment command and their size is | |
| 206 | /// reflected in cmdsize. | |
| 207 | pub const segment_command = extern struct { | |
| 208 | cmd: u32,/// LC_SEGMENT | |
| 209 | cmdsize: u32,/// includes sizeof section structs | |
| 210 | segname: [16]u8,/// segment name | |
| 211 | vmaddr: u32,/// memory address of this segment | |
| 212 | vmsize: u32,/// memory size of this segment | |
| 213 | fileoff: u32,/// file offset of this segment | |
| 214 | filesize: u32,/// amount to map from the file | |
| 215 | maxprot: vm_prot_t,/// maximum VM protection | |
| 216 | initprot: vm_prot_t,/// initial VM protection | |
| 217 | nsects: u32,/// number of sections in segment | |
| 218 | flags: u32, | |
| 219 | }; | |
| 220 | ||
| 221 | /// The 64-bit segment load command indicates that a part of this file is to be | |
| 222 | /// mapped into a 64-bit task's address space. If the 64-bit segment has | |
| 223 | /// sections then section_64 structures directly follow the 64-bit segment | |
| 224 | /// command and their size is reflected in cmdsize. | |
| 225 | pub const segment_command_64 = extern struct { | |
| 226 | cmd: u32, /// LC_SEGMENT_64 | |
| 227 | cmdsize: u32, /// includes sizeof section_64 structs | |
| 228 | segname: [16]u8, /// segment name | |
| 229 | vmaddr: u64, /// memory address of this segment | |
| 230 | vmsize: u64, /// memory size of this segment | |
| 231 | fileoff: u64, /// file offset of this segment | |
| 232 | filesize: u64, /// amount to map from the file | |
| 233 | maxprot: vm_prot_t, /// maximum VM protection | |
| 234 | initprot: vm_prot_t, /// initial VM protection | |
| 235 | nsects: u32, /// number of sections in segment | |
| 236 | flags: u32, | |
| 237 | }; | |
| 238 | ||
| 239 | /// A segment is made up of zero or more sections. Non-MH_OBJECT files have | |
| 240 | /// all of their segments with the proper sections in each, and padded to the | |
| 241 | /// specified segment alignment when produced by the link editor. The first | |
| 242 | /// segment of a MH_EXECUTE and MH_FVMLIB format file contains the mach_header | |
| 243 | /// and load commands of the object file before its first section. The zero | |
| 244 | /// fill sections are always last in their segment (in all formats). This | |
| 245 | /// allows the zeroed segment padding to be mapped into memory where zero fill | |
| 246 | /// sections might be. The gigabyte zero fill sections, those with the section | |
| 247 | /// type S_GB_ZEROFILL, can only be in a segment with sections of this type. | |
| 248 | /// These segments are then placed after all other segments. | |
| 249 | /// | |
| 250 | /// The MH_OBJECT format has all of its sections in one segment for | |
| 251 | /// compactness. There is no padding to a specified segment boundary and the | |
| 252 | /// mach_header and load commands are not part of the segment. | |
| 253 | /// | |
| 254 | /// Sections with the same section name, sectname, going into the same segment, | |
| 255 | /// segname, are combined by the link editor. The resulting section is aligned | |
| 256 | /// to the maximum alignment of the combined sections and is the new section's | |
| 257 | /// alignment. The combined sections are aligned to their original alignment in | |
| 258 | /// the combined section. Any padded bytes to get the specified alignment are | |
| 259 | /// zeroed. | |
| 260 | /// | |
| 261 | /// The format of the relocation entries referenced by the reloff and nreloc | |
| 262 | /// fields of the section structure for mach object files is described in the | |
| 263 | /// header file <reloc.h>. | |
| 264 | pub const @"section" = extern struct { | |
| 265 | sectname: [16]u8, /// name of this section | |
| 266 | segname: [16]u8, /// segment this section goes in | |
| 267 | addr: u32, /// memory address of this section | |
| 268 | size: u32, /// size in bytes of this section | |
| 269 | offset: u32, /// file offset of this section | |
| 270 | @"align": u32, /// section alignment (power of 2) | |
| 271 | reloff: u32, /// file offset of relocation entries | |
| 272 | nreloc: u32, /// number of relocation entries | |
| 273 | flags: u32, /// flags (section type and attributes | |
| 274 | reserved1: u32, /// reserved (for offset or index) | |
| 275 | reserved2: u32, /// reserved (for count or sizeof) | |
| 276 | }; | |
| 277 | ||
| 278 | pub const section_64 = extern struct { | |
| 279 | sectname: [16]u8, /// name of this section | |
| 280 | segname: [16]u8, /// segment this section goes in | |
| 281 | addr: u64, /// memory address of this section | |
| 282 | size: u64, /// size in bytes of this section | |
| 283 | offset: u32, /// file offset of this section | |
| 284 | @"align": u32, /// section alignment (power of 2) | |
| 285 | reloff: u32, /// file offset of relocation entries | |
| 286 | nreloc: u32, /// number of relocation entries | |
| 287 | flags: u32, /// flags (section type and attributes | |
| 288 | reserved1: u32, /// reserved (for offset or index) | |
| 289 | reserved2: u32, /// reserved (for count or sizeof) | |
| 290 | reserved3: u32, /// reserved | |
| 291 | }; | |
| 292 | ||
| 293 | pub const nlist = extern struct { | |
| 294 | n_strx: u32, | |
| 295 | n_type: u8, | |
| 296 | n_sect: u8, | |
| 297 | n_desc: i16, | |
| 298 | n_value: u32, | |
| 299 | }; | |
| 300 | ||
| 301 | pub const nlist_64 = extern struct { | |
| 302 | n_strx: u32, | |
| 303 | n_type: u8, | |
| 304 | n_sect: u8, | |
| 305 | n_desc: u16, | |
| 306 | n_value: u64, | |
| 307 | }; | |
| 308 | ||
| 309 | /// After MacOS X 10.1 when a new load command is added that is required to be | |
| 310 | /// understood by the dynamic linker for the image to execute properly the | |
| 311 | /// LC_REQ_DYLD bit will be or'ed into the load command constant. If the dynamic | |
| 312 | /// linker sees such a load command it it does not understand will issue a | |
| 313 | /// "unknown load command required for execution" error and refuse to use the | |
| 314 | /// image. Other load commands without this bit that are not understood will | |
| 315 | /// simply be ignored. | |
| 316 | pub const LC_REQ_DYLD = 0x80000000; | |
| 317 | ||
| 318 | pub const LC_SEGMENT = 0x1; /// segment of this file to be mapped | |
| 319 | pub const LC_SYMTAB = 0x2; /// link-edit stab symbol table info | |
| 320 | pub const LC_SYMSEG = 0x3; /// link-edit gdb symbol table info (obsolete) | |
| 321 | pub const LC_THREAD = 0x4; /// thread | |
| 322 | pub const LC_UNIXTHREAD = 0x5; /// unix thread (includes a stack) | |
| 323 | pub const LC_LOADFVMLIB = 0x6; /// load a specified fixed VM shared library | |
| 324 | pub const LC_IDFVMLIB = 0x7; /// fixed VM shared library identification | |
| 325 | pub const LC_IDENT = 0x8; /// object identification info (obsolete) | |
| 326 | pub const LC_FVMFILE = 0x9; /// fixed VM file inclusion (internal use) | |
| 327 | pub const LC_PREPAGE = 0xa; /// prepage command (internal use) | |
| 328 | pub const LC_DYSYMTAB = 0xb; /// dynamic link-edit symbol table info | |
| 329 | pub const LC_LOAD_DYLIB = 0xc; /// load a dynamically linked shared library | |
| 330 | pub const LC_ID_DYLIB = 0xd; /// dynamically linked shared lib ident | |
| 331 | pub const LC_LOAD_DYLINKER = 0xe; /// load a dynamic linker | |
| 332 | pub const LC_ID_DYLINKER = 0xf; /// dynamic linker identification | |
| 333 | pub const LC_PREBOUND_DYLIB = 0x10; /// modules prebound for a dynamically | |
| 334 | pub const LC_ROUTINES = 0x11; /// image routines | |
| 335 | pub const LC_SUB_FRAMEWORK = 0x12; /// sub framework | |
| 336 | pub const LC_SUB_UMBRELLA = 0x13; /// sub umbrella | |
| 337 | pub const LC_SUB_CLIENT = 0x14; /// sub client | |
| 338 | pub const LC_SUB_LIBRARY = 0x15; /// sub library | |
| 339 | pub const LC_TWOLEVEL_HINTS = 0x16; /// two-level namespace lookup hints | |
| 340 | pub const LC_PREBIND_CKSUM = 0x17; /// prebind checksum | |
| 341 | ||
| 342 | /// load a dynamically linked shared library that is allowed to be missing | |
| 343 | /// (all symbols are weak imported). | |
| 344 | pub const LC_LOAD_WEAK_DYLIB = (0x18 | LC_REQ_DYLD); | |
| 345 | ||
| 346 | pub const LC_SEGMENT_64 = 0x19; /// 64-bit segment of this file to be mapped | |
| 347 | pub const LC_ROUTINES_64 = 0x1a; /// 64-bit image routines | |
| 348 | pub const LC_UUID = 0x1b; /// the uuid | |
| 349 | pub const LC_RPATH = (0x1c | LC_REQ_DYLD); /// runpath additions | |
| 350 | pub const LC_CODE_SIGNATURE = 0x1d; /// local of code signature | |
| 351 | pub const LC_SEGMENT_SPLIT_INFO = 0x1e; /// local of info to split segments | |
| 352 | pub const LC_REEXPORT_DYLIB = (0x1f | LC_REQ_DYLD); /// load and re-export dylib | |
| 353 | pub const LC_LAZY_LOAD_DYLIB = 0x20; /// delay load of dylib until first use | |
| 354 | pub const LC_ENCRYPTION_INFO = 0x21; /// encrypted segment information | |
| 355 | pub const LC_DYLD_INFO = 0x22; /// compressed dyld information | |
| 356 | pub const LC_DYLD_INFO_ONLY = (0x22|LC_REQ_DYLD); /// compressed dyld information only | |
| 357 | pub const LC_LOAD_UPWARD_DYLIB = (0x23 | LC_REQ_DYLD); /// load upward dylib | |
| 358 | pub const LC_VERSION_MIN_MACOSX = 0x24; /// build for MacOSX min OS version | |
| 359 | pub const LC_VERSION_MIN_IPHONEOS = 0x25; /// build for iPhoneOS min OS version | |
| 360 | pub const LC_FUNCTION_STARTS = 0x26; /// compressed table of function start addresses | |
| 361 | pub const LC_DYLD_ENVIRONMENT = 0x27; /// string for dyld to treat like environment variable | |
| 362 | pub const LC_MAIN = (0x28|LC_REQ_DYLD); /// replacement for LC_UNIXTHREAD | |
| 363 | pub const LC_DATA_IN_CODE = 0x29; /// table of non-instructions in __text | |
| 364 | pub const LC_SOURCE_VERSION = 0x2A; /// source version used to build binary | |
| 365 | pub const LC_DYLIB_CODE_SIGN_DRS = 0x2B; /// Code signing DRs copied from linked dylibs | |
| 366 | pub const LC_ENCRYPTION_INFO_64 = 0x2C; /// 64-bit encrypted segment information | |
| 367 | pub const LC_LINKER_OPTION = 0x2D; /// linker options in MH_OBJECT files | |
| 368 | pub const LC_LINKER_OPTIMIZATION_HINT = 0x2E; /// optimization hints in MH_OBJECT files | |
| 369 | pub const LC_VERSION_MIN_TVOS = 0x2F; /// build for AppleTV min OS version | |
| 370 | pub const LC_VERSION_MIN_WATCHOS = 0x30; /// build for Watch min OS version | |
| 371 | pub const LC_NOTE = 0x31; /// arbitrary data included within a Mach-O file | |
| 372 | pub const LC_BUILD_VERSION = 0x32; /// build for platform min OS version | |
| 373 | ||
| 374 | pub const MH_MAGIC = 0xfeedface; /// the mach magic number | |
| 375 | pub const MH_CIGAM = 0xcefaedfe; /// NXSwapInt(MH_MAGIC) | |
| 376 | ||
| 377 | pub const MH_MAGIC_64 = 0xfeedfacf; /// the 64-bit mach magic number | |
| 378 | pub const MH_CIGAM_64 = 0xcffaedfe; /// NXSwapInt(MH_MAGIC_64) | |
| 379 | ||
| 380 | pub const MH_OBJECT = 0x1; /// relocatable object file | |
| 381 | pub const MH_EXECUTE = 0x2; /// demand paged executable file | |
| 382 | pub const MH_FVMLIB = 0x3; /// fixed VM shared library file | |
| 383 | pub const MH_CORE = 0x4; /// core file | |
| 384 | pub const MH_PRELOAD = 0x5; /// preloaded executable file | |
| 385 | pub const MH_DYLIB = 0x6; /// dynamically bound shared library | |
| 386 | pub const MH_DYLINKER = 0x7; /// dynamic link editor | |
| 387 | pub const MH_BUNDLE = 0x8; /// dynamically bound bundle file | |
| 388 | pub const MH_DYLIB_STUB = 0x9; /// shared library stub for static linking only, no section contents | |
| 389 | pub const MH_DSYM = 0xa; /// companion file with only debug sections | |
| 390 | pub const MH_KEXT_BUNDLE = 0xb; /// x86_64 kexts | |
| 391 | ||
| 392 | // Constants for the flags field of the mach_header | |
| 393 | ||
| 394 | pub const MH_NOUNDEFS = 0x1; /// the object file has no undefined references | |
| 395 | pub const MH_INCRLINK = 0x2; /// the object file is the output of an incremental link against a base file and can't be link edited again | |
| 396 | pub const MH_DYLDLINK = 0x4; /// the object file is input for the dynamic linker and can't be staticly link edited again | |
| 397 | pub const MH_BINDATLOAD = 0x8; /// the object file's undefined references are bound by the dynamic linker when loaded. | |
| 398 | pub const MH_PREBOUND = 0x10; /// the file has its dynamic undefined references prebound. | |
| 399 | pub const MH_SPLIT_SEGS = 0x20; /// the file has its read-only and read-write segments split | |
| 400 | pub const MH_LAZY_INIT = 0x40; /// the shared library init routine is to be run lazily via catching memory faults to its writeable segments (obsolete) | |
| 401 | pub const MH_TWOLEVEL = 0x80; /// the image is using two-level name space bindings | |
| 402 | pub const MH_FORCE_FLAT = 0x100; /// the executable is forcing all images to use flat name space bindings | |
| 403 | pub const MH_NOMULTIDEFS = 0x200; /// this umbrella guarantees no multiple defintions of symbols in its sub-images so the two-level namespace hints can always be used. | |
| 404 | pub const MH_NOFIXPREBINDING = 0x400; /// do not have dyld notify the prebinding agent about this executable | |
| 405 | pub const MH_PREBINDABLE = 0x800; /// the binary is not prebound but can have its prebinding redone. only used when MH_PREBOUND is not set. | |
| 406 | pub const MH_ALLMODSBOUND = 0x1000; /// indicates that this binary binds to all two-level namespace modules of its dependent libraries. only used when MH_PREBINDABLE and MH_TWOLEVEL are both set. | |
| 407 | pub const MH_SUBSECTIONS_VIA_SYMBOLS = 0x2000;/// safe to divide up the sections into sub-sections via symbols for dead code stripping | |
| 408 | pub const MH_CANONICAL = 0x4000; /// the binary has been canonicalized via the unprebind operation | |
| 409 | pub const MH_WEAK_DEFINES = 0x8000; /// the final linked image contains external weak symbols | |
| 410 | pub const MH_BINDS_TO_WEAK = 0x10000; /// the final linked image uses weak symbols | |
| 411 | ||
| 412 | pub const MH_ALLOW_STACK_EXECUTION = 0x20000;/// When this bit is set, all stacks in the task will be given stack execution privilege. Only used in MH_EXECUTE filetypes. | |
| 413 | pub const MH_ROOT_SAFE = 0x40000; /// When this bit is set, the binary declares it is safe for use in processes with uid zero | |
| 414 | ||
| 415 | pub const MH_SETUID_SAFE = 0x80000; /// When this bit is set, the binary declares it is safe for use in processes when issetugid() is true | |
| 416 | ||
| 417 | pub const MH_NO_REEXPORTED_DYLIBS = 0x100000; /// When this bit is set on a dylib, the static linker does not need to examine dependent dylibs to see if any are re-exported | |
| 418 | pub const MH_PIE = 0x200000; /// When this bit is set, the OS will load the main executable at a random address. Only used in MH_EXECUTE filetypes. | |
| 419 | pub const MH_DEAD_STRIPPABLE_DYLIB = 0x400000; /// Only for use on dylibs. When linking against a dylib that has this bit set, the static linker will automatically not create a LC_LOAD_DYLIB load command to the dylib if no symbols are being referenced from the dylib. | |
| 420 | pub const MH_HAS_TLV_DESCRIPTORS = 0x800000; /// Contains a section of type S_THREAD_LOCAL_VARIABLES | |
| 421 | ||
| 422 | pub const MH_NO_HEAP_EXECUTION = 0x1000000; /// When this bit is set, the OS will run the main executable with a non-executable heap even on platforms (e.g. i386) that don't require it. Only used in MH_EXECUTE filetypes. | |
| 423 | ||
| 424 | pub const MH_APP_EXTENSION_SAFE = 0x02000000; /// The code was linked for use in an application extension. | |
| 425 | ||
| 426 | pub const MH_NLIST_OUTOFSYNC_WITH_DYLDINFO = 0x04000000; /// The external symbols listed in the nlist symbol table do not include all the symbols listed in the dyld info. | |
| 427 | ||
| 428 | ||
| 429 | /// The flags field of a section structure is separated into two parts a section | |
| 430 | /// type and section attributes. The section types are mutually exclusive (it | |
| 431 | /// can only have one type) but the section attributes are not (it may have more | |
| 432 | /// than one attribute). | |
| 433 | /// 256 section types | |
| 434 | pub const SECTION_TYPE = 0x000000ff; | |
| 435 | pub const SECTION_ATTRIBUTES = 0xffffff00; /// 24 section attributes | |
| 436 | ||
| 437 | pub const S_REGULAR = 0x0; /// regular section | |
| 438 | pub const S_ZEROFILL = 0x1; /// zero fill on demand section | |
| 439 | pub const S_CSTRING_LITERALS = 0x2; /// section with only literal C string | |
| 440 | pub const S_4BYTE_LITERALS = 0x3; /// section with only 4 byte literals | |
| 441 | pub const S_8BYTE_LITERALS = 0x4; /// section with only 8 byte literals | |
| 442 | pub const S_LITERAL_POINTERS = 0x5; /// section with only pointers to | |
| 443 | ||
| 444 | ||
| 445 | pub const N_STAB = 0xe0; /// if any of these bits set, a symbolic debugging entry | |
| 446 | pub const N_PEXT = 0x10; /// private external symbol bit | |
| 447 | pub const N_TYPE = 0x0e; /// mask for the type bits | |
| 448 | pub const N_EXT = 0x01; /// external symbol bit, set for external symbols | |
| 449 | ||
| 450 | ||
| 451 | pub const N_GSYM = 0x20; /// global symbol: name,,NO_SECT,type,0 | |
| 452 | pub const N_FNAME = 0x22; /// procedure name (f77 kludge): name,,NO_SECT,0,0 | |
| 453 | pub const N_FUN = 0x24; /// procedure: name,,n_sect,linenumber,address | |
| 454 | pub const N_STSYM = 0x26; /// static symbol: name,,n_sect,type,address | |
| 455 | pub const N_LCSYM = 0x28; /// .lcomm symbol: name,,n_sect,type,address | |
| 456 | pub const N_BNSYM = 0x2e; /// begin nsect sym: 0,,n_sect,0,address | |
| 457 | pub const N_AST = 0x32; /// AST file path: name,,NO_SECT,0,0 | |
| 458 | pub const N_OPT = 0x3c; /// emitted with gcc2_compiled and in gcc source | |
| 459 | pub const N_RSYM = 0x40; /// register sym: name,,NO_SECT,type,register | |
| 460 | pub const N_SLINE = 0x44; /// src line: 0,,n_sect,linenumber,address | |
| 461 | pub const N_ENSYM = 0x4e; /// end nsect sym: 0,,n_sect,0,address | |
| 462 | pub const N_SSYM = 0x60; /// structure elt: name,,NO_SECT,type,struct_offset | |
| 463 | pub const N_SO = 0x64; /// source file name: name,,n_sect,0,address | |
| 464 | pub const N_OSO = 0x66; /// object file name: name,,0,0,st_mtime | |
| 465 | pub const N_LSYM = 0x80; /// local sym: name,,NO_SECT,type,offset | |
| 466 | pub const N_BINCL = 0x82; /// include file beginning: name,,NO_SECT,0,sum | |
| 467 | pub const N_SOL = 0x84; /// #included file name: name,,n_sect,0,address | |
| 468 | pub const N_PARAMS = 0x86; /// compiler parameters: name,,NO_SECT,0,0 | |
| 469 | pub const N_VERSION = 0x88; /// compiler version: name,,NO_SECT,0,0 | |
| 470 | pub const N_OLEVEL = 0x8A; /// compiler -O level: name,,NO_SECT,0,0 | |
| 471 | pub const N_PSYM = 0xa0; /// parameter: name,,NO_SECT,type,offset | |
| 472 | pub const N_EINCL = 0xa2; /// include file end: name,,NO_SECT,0,0 | |
| 473 | pub const N_ENTRY = 0xa4; /// alternate entry: name,,n_sect,linenumber,address | |
| 474 | pub const N_LBRAC = 0xc0; /// left bracket: 0,,NO_SECT,nesting level,address | |
| 475 | pub const N_EXCL = 0xc2; /// deleted include file: name,,NO_SECT,0,sum | |
| 476 | pub const N_RBRAC = 0xe0; /// right bracket: 0,,NO_SECT,nesting level,address | |
| 477 | pub const N_BCOMM = 0xe2; /// begin common: name,,NO_SECT,0,0 | |
| 478 | pub const N_ECOMM = 0xe4; /// end common: name,,n_sect,0,0 | |
| 479 | pub const N_ECOML = 0xe8; /// end common (local name): 0,,n_sect,0,address | |
| 480 | pub const N_LENG = 0xfe; /// second stab entry with length information | |
| 481 | ||
| 482 | /// If a segment contains any sections marked with S_ATTR_DEBUG then all | |
| 483 | /// sections in that segment must have this attribute. No section other than | |
| 484 | /// a section marked with this attribute may reference the contents of this | |
| 485 | /// section. A section with this attribute may contain no symbols and must have | |
| 486 | /// a section type S_REGULAR. The static linker will not copy section contents | |
| 487 | /// from sections with this attribute into its output file. These sections | |
| 488 | /// generally contain DWARF debugging info. | |
| 489 | pub const S_ATTR_DEBUG = 0x02000000; /// a debug section | |
| 490 | ||
| 491 | pub const cpu_type_t = integer_t; | |
| 492 | pub const cpu_subtype_t = integer_t; | |
| 493 | pub const integer_t = c_int; | |
| 494 | pub const vm_prot_t = c_int; | |
| 495 | ||
| 142 | 496 | // sys/types.h on macos uses #pragma pack(4) so these checks are |
| 143 | 497 | // to make sure the struct is laid out the same. These values were |
| 144 | 498 | // produced from C code using the offsetof macro. |
std/debug/index.zig+251-111| ... | ... | @@ -5,7 +5,6 @@ const io = std.io; |
| 5 | 5 | const os = std.os; |
| 6 | 6 | const elf = std.elf; |
| 7 | 7 | const DW = std.dwarf; |
| 8 | const macho = std.macho; | |
| 9 | 8 | const ArrayList = std.ArrayList; |
| 10 | 9 | const builtin = @import("builtin"); |
| 11 | 10 | |
| ... | ... | @@ -19,9 +18,10 @@ pub const runtime_safety = switch (builtin.mode) { |
| 19 | 18 | |
| 20 | 19 | /// Tries to write to stderr, unbuffered, and ignores any error returned. |
| 21 | 20 | /// Does not append a newline. |
| 22 | /// TODO atomic/multithread support | |
| 23 | 21 | var stderr_file: os.File = undefined; |
| 24 | 22 | var stderr_file_out_stream: io.FileOutStream = undefined; |
| 23 | ||
| 24 | /// TODO multithreaded awareness | |
| 25 | 25 | var stderr_stream: ?*io.OutStream(io.FileOutStream.Error) = null; |
| 26 | 26 | var stderr_mutex = std.Mutex.init(); |
| 27 | 27 | pub fn warn(comptime fmt: []const u8, args: ...) void { |
| ... | ... | @@ -30,6 +30,7 @@ pub fn warn(comptime fmt: []const u8, args: ...) void { |
| 30 | 30 | const stderr = getStderrStream() catch return; |
| 31 | 31 | stderr.print(fmt, args) catch return; |
| 32 | 32 | } |
| 33 | ||
| 33 | 34 | pub fn getStderrStream() !*io.OutStream(io.FileOutStream.Error) { |
| 34 | 35 | if (stderr_stream) |st| { |
| 35 | 36 | return st; |
| ... | ... | @@ -42,14 +43,15 @@ pub fn getStderrStream() !*io.OutStream(io.FileOutStream.Error) { |
| 42 | 43 | } |
| 43 | 44 | } |
| 44 | 45 | |
| 45 | var self_debug_info: ?*DebugInfo = null; | |
| 46 | /// TODO multithreaded awareness | |
| 47 | var self_debug_info: ?DebugInfo = null; | |
| 48 | ||
| 46 | 49 | pub fn getSelfDebugInfo() !*DebugInfo { |
| 47 | if (self_debug_info) |info| { | |
| 50 | if (self_debug_info) |*info| { | |
| 48 | 51 | return info; |
| 49 | 52 | } else { |
| 50 | const info = try openSelfDebugInfo(getDebugInfoAllocator()); | |
| 51 | self_debug_info = info; | |
| 52 | return info; | |
| 53 | self_debug_info = try openSelfDebugInfo(getDebugInfoAllocator()); | |
| 54 | return &self_debug_info.?; | |
| 53 | 55 | } |
| 54 | 56 | } |
| 55 | 57 | |
| ... | ... | @@ -127,6 +129,7 @@ pub fn panic(comptime format: []const u8, args: ...) noreturn { |
| 127 | 129 | panicExtra(null, first_trace_addr, format, args); |
| 128 | 130 | } |
| 129 | 131 | |
| 132 | /// TODO multithreaded awareness | |
| 130 | 133 | var panicking: u8 = 0; // TODO make this a bool |
| 131 | 134 | |
| 132 | 135 | pub fn panicExtra(trace: ?*const builtin.StackTrace, first_trace_addr: ?usize, comptime format: []const u8, args: ...) noreturn { |
| ... | ... | @@ -220,80 +223,147 @@ pub fn writeCurrentStackTrace(out_stream: var, allocator: *mem.Allocator, debug_ |
| 220 | 223 | |
| 221 | 224 | pub fn printSourceAtAddress(debug_info: *DebugInfo, out_stream: var, address: usize, tty_color: bool) !void { |
| 222 | 225 | switch (builtin.os) { |
| 223 | builtin.Os.windows => return error.UnsupportedDebugInfo, | |
| 224 | builtin.Os.macosx => { | |
| 225 | // TODO(bnoordhuis) It's theoretically possible to obtain the | |
| 226 | // compilation unit from the symbtab but it's not that useful | |
| 227 | // in practice because the compiler dumps everything in a single | |
| 228 | // object file. Future improvement: use external dSYM data when | |
| 229 | // available. | |
| 230 | const unknown = macho.Symbol{ | |
| 231 | .name = "???", | |
| 232 | .address = address, | |
| 233 | }; | |
| 234 | const symbol = debug_info.symbol_table.search(address) orelse &unknown; | |
| 235 | try out_stream.print(WHITE ++ "{}" ++ RESET ++ ": " ++ DIM ++ "0x{x}" ++ " in ??? (???)" ++ RESET ++ "\n", symbol.name, address); | |
| 226 | builtin.Os.macosx => return printSourceAtAddressMacOs(debug_info, out_stream, address, tty_color), | |
| 227 | builtin.Os.linux => return printSourceAtAddressLinux(debug_info, out_stream, address, tty_color), | |
| 228 | builtin.Os.windows => { | |
| 229 | // TODO https://github.com/ziglang/zig/issues/721 | |
| 230 | return error.UnsupportedOperatingSystem; | |
| 236 | 231 | }, |
| 237 | else => { | |
| 238 | const compile_unit = findCompileUnit(debug_info, address) catch { | |
| 239 | if (tty_color) { | |
| 240 | try out_stream.print("???:?:?: " ++ DIM ++ "0x{x} in ??? (???)" ++ RESET ++ "\n ???\n\n", address); | |
| 241 | } else { | |
| 242 | try out_stream.print("???:?:?: 0x{x} in ??? (???)\n ???\n\n", address); | |
| 243 | } | |
| 244 | return; | |
| 245 | }; | |
| 246 | const compile_unit_name = try compile_unit.die.getAttrString(debug_info, DW.AT_name); | |
| 247 | if (getLineNumberInfo(debug_info, compile_unit, address - 1)) |line_info| { | |
| 248 | defer line_info.deinit(); | |
| 249 | if (tty_color) { | |
| 250 | try out_stream.print( | |
| 251 | WHITE ++ "{}:{}:{}" ++ RESET ++ ": " ++ DIM ++ "0x{x} in ??? ({})" ++ RESET ++ "\n", | |
| 252 | line_info.file_name, | |
| 253 | line_info.line, | |
| 254 | line_info.column, | |
| 255 | address, | |
| 256 | compile_unit_name, | |
| 257 | ); | |
| 258 | if (printLineFromFile(out_stream, line_info)) { | |
| 259 | if (line_info.column == 0) { | |
| 260 | try out_stream.write("\n"); | |
| 261 | } else { | |
| 262 | { | |
| 263 | var col_i: usize = 1; | |
| 264 | while (col_i < line_info.column) : (col_i += 1) { | |
| 265 | try out_stream.writeByte(' '); | |
| 266 | } | |
| 267 | } | |
| 268 | try out_stream.write(GREEN ++ "^" ++ RESET ++ "\n"); | |
| 269 | } | |
| 270 | } else |err| switch (err) { | |
| 271 | error.EndOfFile => {}, | |
| 272 | else => return err, | |
| 232 | else => return error.UnsupportedOperatingSystem, | |
| 233 | } | |
| 234 | } | |
| 235 | ||
| 236 | fn machoSearchSymbols(symbols: []const MachoSymbol, address: usize) ?*const MachoSymbol { | |
| 237 | var min: usize = 0; | |
| 238 | var max: usize = symbols.len - 1; // Exclude sentinel. | |
| 239 | while (min < max) { | |
| 240 | const mid = min + (max - min) / 2; | |
| 241 | const curr = &symbols[mid]; | |
| 242 | const next = &symbols[mid + 1]; | |
| 243 | if (address >= next.address()) { | |
| 244 | min = mid + 1; | |
| 245 | } else if (address < curr.address()) { | |
| 246 | max = mid; | |
| 247 | } else { | |
| 248 | return curr; | |
| 249 | } | |
| 250 | } | |
| 251 | return null; | |
| 252 | } | |
| 253 | ||
| 254 | fn printSourceAtAddressMacOs(debug_info: *DebugInfo, out_stream: var, address: usize, tty_color: bool) !void { | |
| 255 | const base_addr = @ptrToInt(&std.c._mh_execute_header); | |
| 256 | const adjusted_addr = 0x100000000 + (address - base_addr); | |
| 257 | ||
| 258 | const symbol = machoSearchSymbols(debug_info.symbols, adjusted_addr) orelse { | |
| 259 | if (tty_color) { | |
| 260 | try out_stream.print("???:?:?: " ++ DIM ++ "0x{x} in ??? (???)" ++ RESET ++ "\n\n\n", address); | |
| 261 | } else { | |
| 262 | try out_stream.print("???:?:?: 0x{x} in ??? (???)\n\n\n", address); | |
| 263 | } | |
| 264 | return; | |
| 265 | }; | |
| 266 | ||
| 267 | const symbol_name = mem.toSliceConst(u8, debug_info.strings.ptr + symbol.nlist.n_strx); | |
| 268 | if (getLineNumberInfoMacOs(debug_info, symbol.*, address)) |line_info| { | |
| 269 | const compile_unit_name = "???"; | |
| 270 | try printLineInfo(debug_info, out_stream, line_info, address, symbol_name, compile_unit_name, tty_color); | |
| 271 | } else |err| switch (err) { | |
| 272 | error.MissingDebugInfo => { | |
| 273 | if (tty_color) { | |
| 274 | try out_stream.print("???:?:?: " ++ DIM ++ "0x{x} in {} (???)" ++ RESET ++ "\n\n\n", address, symbol_name); | |
| 275 | } else { | |
| 276 | try out_stream.print("???:?:?: 0x{x} in {} (???)\n\n\n", address, symbol_name); | |
| 277 | } | |
| 278 | }, | |
| 279 | else => return err, | |
| 280 | } | |
| 281 | } | |
| 282 | ||
| 283 | pub fn printSourceAtAddressLinux(debug_info: *DebugInfo, out_stream: var, address: usize, tty_color: bool) !void { | |
| 284 | const compile_unit = findCompileUnit(debug_info, address) catch { | |
| 285 | if (tty_color) { | |
| 286 | try out_stream.print("???:?:?: " ++ DIM ++ "0x{x} in ??? (???)" ++ RESET ++ "\n\n\n", address); | |
| 287 | } else { | |
| 288 | try out_stream.print("???:?:?: 0x{x} in ??? (???)\n\n\n", address); | |
| 289 | } | |
| 290 | return; | |
| 291 | }; | |
| 292 | const compile_unit_name = try compile_unit.die.getAttrString(debug_info, DW.AT_name); | |
| 293 | if (getLineNumberInfoLinux(debug_info, compile_unit, address - 1)) |line_info| { | |
| 294 | defer line_info.deinit(); | |
| 295 | const symbol_name = "???"; | |
| 296 | try printLineInfo(debug_info, out_stream, line_info, address, symbol_name, compile_unit_name, tty_color); | |
| 297 | } else |err| switch (err) { | |
| 298 | error.MissingDebugInfo, error.InvalidDebugInfo => { | |
| 299 | if (tty_color) { | |
| 300 | try out_stream.print("???:?:?: " ++ DIM ++ "0x{x} in ??? ({})" ++ RESET ++ "\n\n\n", address, compile_unit_name); | |
| 301 | } else { | |
| 302 | try out_stream.print("???:?:?: 0x{x} in ??? ({})\n\n\n", address, compile_unit_name); | |
| 303 | } | |
| 304 | }, | |
| 305 | else => return err, | |
| 306 | } | |
| 307 | } | |
| 308 | ||
| 309 | fn printLineInfo( | |
| 310 | debug_info: *DebugInfo, | |
| 311 | out_stream: var, | |
| 312 | line_info: LineInfo, | |
| 313 | address: usize, | |
| 314 | symbol_name: []const u8, | |
| 315 | compile_unit_name: []const u8, | |
| 316 | tty_color: bool, | |
| 317 | ) !void { | |
| 318 | if (tty_color) { | |
| 319 | try out_stream.print( | |
| 320 | WHITE ++ "{}:{}:{}" ++ RESET ++ ": " ++ DIM ++ "0x{x} in {} ({})" ++ RESET ++ "\n", | |
| 321 | line_info.file_name, | |
| 322 | line_info.line, | |
| 323 | line_info.column, | |
| 324 | address, | |
| 325 | symbol_name, | |
| 326 | compile_unit_name, | |
| 327 | ); | |
| 328 | if (printLineFromFile(out_stream, line_info)) { | |
| 329 | if (line_info.column == 0) { | |
| 330 | try out_stream.write("\n"); | |
| 331 | } else { | |
| 332 | { | |
| 333 | var col_i: usize = 1; | |
| 334 | while (col_i < line_info.column) : (col_i += 1) { | |
| 335 | try out_stream.writeByte(' '); | |
| 273 | 336 | } |
| 274 | } else { | |
| 275 | try out_stream.print( | |
| 276 | "{}:{}:{}: 0x{x} in ??? ({})\n", | |
| 277 | line_info.file_name, | |
| 278 | line_info.line, | |
| 279 | line_info.column, | |
| 280 | address, | |
| 281 | compile_unit_name, | |
| 282 | ); | |
| 283 | 337 | } |
| 284 | } else |err| switch (err) { | |
| 285 | error.MissingDebugInfo, error.InvalidDebugInfo => { | |
| 286 | try out_stream.print("0x{x} in ??? ({})\n", address, compile_unit_name); | |
| 287 | }, | |
| 288 | else => return err, | |
| 338 | try out_stream.write(GREEN ++ "^" ++ RESET ++ "\n"); | |
| 289 | 339 | } |
| 290 | }, | |
| 340 | } else |err| switch (err) { | |
| 341 | error.EndOfFile => {}, | |
| 342 | else => return err, | |
| 343 | } | |
| 344 | } else { | |
| 345 | try out_stream.print( | |
| 346 | "{}:{}:{}: 0x{x} in {} ({})\n", | |
| 347 | line_info.file_name, | |
| 348 | line_info.line, | |
| 349 | line_info.column, | |
| 350 | address, | |
| 351 | symbol_name, | |
| 352 | compile_unit_name, | |
| 353 | ); | |
| 291 | 354 | } |
| 292 | 355 | } |
| 293 | 356 | |
| 294 | pub fn openSelfDebugInfo(allocator: *mem.Allocator) !*DebugInfo { | |
| 295 | switch (builtin.object_format) { | |
| 296 | builtin.ObjectFormat.elf => { | |
| 357 | // TODO use this | |
| 358 | pub const OpenSelfDebugInfoError = error{ | |
| 359 | MissingDebugInfo, | |
| 360 | OutOfMemory, | |
| 361 | UnsupportedOperatingSystem, | |
| 362 | }; | |
| 363 | ||
| 364 | pub fn openSelfDebugInfo(allocator: *mem.Allocator) !DebugInfo { | |
| 365 | switch (builtin.os) { | |
| 366 | builtin.Os.linux => { | |
| 297 | 367 | const st = try allocator.create(DebugInfo{ |
| 298 | 368 | .self_exe_file = undefined, |
| 299 | 369 | .elf = undefined, |
| ... | ... | @@ -320,24 +390,79 @@ pub fn openSelfDebugInfo(allocator: *mem.Allocator) !*DebugInfo { |
| 320 | 390 | try scanAllCompileUnits(st); |
| 321 | 391 | return st; |
| 322 | 392 | }, |
| 323 | builtin.ObjectFormat.macho => { | |
| 324 | var exe_file = try os.openSelfExe(); | |
| 325 | defer exe_file.close(); | |
| 326 | ||
| 327 | const st = try allocator.create(DebugInfo{ .symbol_table = try macho.loadSymbols(allocator, &io.FileInStream.init(&exe_file)) }); | |
| 328 | errdefer allocator.destroy(st); | |
| 329 | return st; | |
| 330 | }, | |
| 331 | builtin.ObjectFormat.coff => { | |
| 332 | return error.TodoSupportCoffDebugInfo; | |
| 333 | }, | |
| 334 | builtin.ObjectFormat.wasm => { | |
| 335 | return error.TodoSupportCOFFDebugInfo; | |
| 336 | }, | |
| 337 | builtin.ObjectFormat.unknown => { | |
| 338 | return error.UnknownObjectFormat; | |
| 393 | builtin.Os.macosx, builtin.Os.ios => return openSelfDebugInfoMacOs(allocator), | |
| 394 | builtin.Os.windows => { | |
| 395 | // TODO: https://github.com/ziglang/zig/issues/721 | |
| 396 | return error.UnsupportedOperatingSystem; | |
| 339 | 397 | }, |
| 398 | else => return error.UnsupportedOperatingSystem, | |
| 399 | } | |
| 400 | } | |
| 401 | ||
| 402 | fn openSelfDebugInfoMacOs(allocator: *mem.Allocator) !DebugInfo { | |
| 403 | const hdr = &std.c._mh_execute_header; | |
| 404 | assert(hdr.magic == std.c.MH_MAGIC_64); | |
| 405 | ||
| 406 | const hdr_base = @ptrCast([*]u8, hdr); | |
| 407 | var ptr = hdr_base + @sizeOf(std.c.mach_header_64); | |
| 408 | var ncmd: u32 = hdr.ncmds; | |
| 409 | const symtab = while (ncmd != 0) : (ncmd -= 1) { | |
| 410 | const lc = @ptrCast(*std.c.load_command, ptr); | |
| 411 | switch (lc.cmd) { | |
| 412 | std.c.LC_SYMTAB => break @ptrCast(*std.c.symtab_command, ptr), | |
| 413 | else => {}, | |
| 414 | } | |
| 415 | ptr += lc.cmdsize; // TODO https://github.com/ziglang/zig/issues/1403 | |
| 416 | } else { | |
| 417 | return error.MissingDebugInfo; | |
| 418 | }; | |
| 419 | const syms = @ptrCast([*]std.c.nlist_64, hdr_base + symtab.symoff)[0..symtab.nsyms]; | |
| 420 | const strings = @ptrCast([*]u8, hdr_base + symtab.stroff)[0..symtab.strsize]; | |
| 421 | ||
| 422 | const symbols_buf = try allocator.alloc(MachoSymbol, syms.len); | |
| 423 | ||
| 424 | var ofile: ?*std.c.nlist_64 = null; | |
| 425 | var symbol_index: usize = 0; | |
| 426 | var last_len: u64 = 0; | |
| 427 | for (syms) |*sym| { | |
| 428 | if (sym.n_type & std.c.N_STAB != 0) { | |
| 429 | switch (sym.n_type) { | |
| 430 | std.c.N_OSO => ofile = sym, | |
| 431 | std.c.N_FUN => { | |
| 432 | if (sym.n_sect == 0) { | |
| 433 | last_len = sym.n_value; | |
| 434 | } else { | |
| 435 | symbols_buf[symbol_index] = MachoSymbol{ | |
| 436 | .nlist = sym, | |
| 437 | .ofile = ofile, | |
| 438 | }; | |
| 439 | symbol_index += 1; | |
| 440 | } | |
| 441 | }, | |
| 442 | else => continue, | |
| 443 | } | |
| 444 | } | |
| 340 | 445 | } |
| 446 | const sentinel = try allocator.createOne(std.c.nlist_64); | |
| 447 | sentinel.* = std.c.nlist_64{ | |
| 448 | .n_strx = 0, | |
| 449 | .n_type = 36, | |
| 450 | .n_sect = 0, | |
| 451 | .n_desc = 0, | |
| 452 | .n_value = symbols_buf[symbol_index - 1].nlist.n_value + last_len, | |
| 453 | }; | |
| 454 | ||
| 455 | const symbols = allocator.shrink(MachoSymbol, symbols_buf, symbol_index); | |
| 456 | ||
| 457 | // Even though lld emits symbols in ascending order, this debug code | |
| 458 | // should work for programs linked in any valid way. | |
| 459 | // This sort is so that we can binary search later. | |
| 460 | std.sort.sort(MachoSymbol, symbols, MachoSymbol.addressLessThan); | |
| 461 | ||
| 462 | return DebugInfo{ | |
| 463 | .symbols = symbols, | |
| 464 | .strings = strings, | |
| 465 | }; | |
| 341 | 466 | } |
| 342 | 467 | |
| 343 | 468 | fn printLineFromFile(out_stream: var, line_info: *const LineInfo) !void { |
| ... | ... | @@ -372,13 +497,24 @@ fn printLineFromFile(out_stream: var, line_info: *const LineInfo) !void { |
| 372 | 497 | } |
| 373 | 498 | } |
| 374 | 499 | |
| 500 | const MachoSymbol = struct { | |
| 501 | nlist: *std.c.nlist_64, | |
| 502 | ofile: ?*std.c.nlist_64, | |
| 503 | ||
| 504 | /// Returns the address from the macho file | |
| 505 | fn address(self: MachoSymbol) u64 { | |
| 506 | return self.nlist.n_value; | |
| 507 | } | |
| 508 | ||
| 509 | fn addressLessThan(lhs: MachoSymbol, rhs: MachoSymbol) bool { | |
| 510 | return lhs.address() < rhs.address(); | |
| 511 | } | |
| 512 | }; | |
| 513 | ||
| 375 | 514 | pub const DebugInfo = switch (builtin.os) { |
| 376 | 515 | builtin.Os.macosx => struct { |
| 377 | symbol_table: macho.SymbolTable, | |
| 378 | ||
| 379 | pub fn close(self: *DebugInfo) void { | |
| 380 | self.symbol_table.deinit(); | |
| 381 | } | |
| 516 | symbols: []const MachoSymbol, | |
| 517 | strings: []const u8, | |
| 382 | 518 | }, |
| 383 | 519 | else => struct { |
| 384 | 520 | self_exe_file: os.File, |
| ... | ... | @@ -803,12 +939,16 @@ fn parseDie(st: *DebugInfo, abbrev_table: *const AbbrevTable, is_64: bool) !Die |
| 803 | 939 | return result; |
| 804 | 940 | } |
| 805 | 941 | |
| 806 | fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_address: usize) !LineInfo { | |
| 807 | const compile_unit_cwd = try compile_unit.die.getAttrString(st, DW.AT_comp_dir); | |
| 942 | fn getLineNumberInfoMacOs(di: *DebugInfo, symbol: MachoSymbol, target_address: usize) !LineInfo { | |
| 943 | return error.MissingDebugInfo; | |
| 944 | } | |
| 808 | 945 | |
| 809 | const in_file = &st.self_exe_file; | |
| 810 | const debug_line_end = st.debug_line.offset + st.debug_line.size; | |
| 811 | var this_offset = st.debug_line.offset; | |
| 946 | fn getLineNumberInfoLinux(di: *DebugInfo, compile_unit: *const CompileUnit, target_address: usize) !LineInfo { | |
| 947 | const compile_unit_cwd = try compile_unit.die.getAttrString(di, DW.AT_comp_dir); | |
| 948 | ||
| 949 | const in_file = &di.self_exe_file; | |
| 950 | const debug_line_end = di.debug_line.offset + di.debug_line.size; | |
| 951 | var this_offset = di.debug_line.offset; | |
| 812 | 952 | var this_index: usize = 0; |
| 813 | 953 | |
| 814 | 954 | var in_file_stream = io.FileInStream.init(in_file); |
| ... | ... | @@ -827,11 +967,11 @@ fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_ad |
| 827 | 967 | continue; |
| 828 | 968 | } |
| 829 | 969 | |
| 830 | const version = try in_stream.readInt(st.elf.endian, u16); | |
| 970 | const version = try in_stream.readInt(di.elf.endian, u16); | |
| 831 | 971 | // TODO support 3 and 5 |
| 832 | 972 | if (version != 2 and version != 4) return error.InvalidDebugInfo; |
| 833 | 973 | |
| 834 | const prologue_length = if (is_64) try in_stream.readInt(st.elf.endian, u64) else try in_stream.readInt(st.elf.endian, u32); | |
| 974 | const prologue_length = if (is_64) try in_stream.readInt(di.elf.endian, u64) else try in_stream.readInt(di.elf.endian, u32); | |
| 835 | 975 | const prog_start_offset = (try in_file.getPos()) + prologue_length; |
| 836 | 976 | |
| 837 | 977 | const minimum_instruction_length = try in_stream.readByte(); |
| ... | ... | @@ -850,7 +990,7 @@ fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_ad |
| 850 | 990 | |
| 851 | 991 | const opcode_base = try in_stream.readByte(); |
| 852 | 992 | |
| 853 | const standard_opcode_lengths = try st.allocator().alloc(u8, opcode_base - 1); | |
| 993 | const standard_opcode_lengths = try di.allocator().alloc(u8, opcode_base - 1); | |
| 854 | 994 | |
| 855 | 995 | { |
| 856 | 996 | var i: usize = 0; |
| ... | ... | @@ -859,19 +999,19 @@ fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_ad |
| 859 | 999 | } |
| 860 | 1000 | } |
| 861 | 1001 | |
| 862 | var include_directories = ArrayList([]u8).init(st.allocator()); | |
| 1002 | var include_directories = ArrayList([]u8).init(di.allocator()); | |
| 863 | 1003 | try include_directories.append(compile_unit_cwd); |
| 864 | 1004 | while (true) { |
| 865 | const dir = try st.readString(); | |
| 1005 | const dir = try di.readString(); | |
| 866 | 1006 | if (dir.len == 0) break; |
| 867 | 1007 | try include_directories.append(dir); |
| 868 | 1008 | } |
| 869 | 1009 | |
| 870 | var file_entries = ArrayList(FileEntry).init(st.allocator()); | |
| 1010 | var file_entries = ArrayList(FileEntry).init(di.allocator()); | |
| 871 | 1011 | var prog = LineNumberProgram.init(default_is_stmt, include_directories.toSliceConst(), &file_entries, target_address); |
| 872 | 1012 | |
| 873 | 1013 | while (true) { |
| 874 | const file_name = try st.readString(); | |
| 1014 | const file_name = try di.readString(); | |
| 875 | 1015 | if (file_name.len == 0) break; |
| 876 | 1016 | const dir_index = try readULeb128(in_stream); |
| 877 | 1017 | const mtime = try readULeb128(in_stream); |
| ... | ... | @@ -901,11 +1041,11 @@ fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_ad |
| 901 | 1041 | return error.MissingDebugInfo; |
| 902 | 1042 | }, |
| 903 | 1043 | DW.LNE_set_address => { |
| 904 | const addr = try in_stream.readInt(st.elf.endian, usize); | |
| 1044 | const addr = try in_stream.readInt(di.elf.endian, usize); | |
| 905 | 1045 | prog.address = addr; |
| 906 | 1046 | }, |
| 907 | 1047 | DW.LNE_define_file => { |
| 908 | const file_name = try st.readString(); | |
| 1048 | const file_name = try di.readString(); | |
| 909 | 1049 | const dir_index = try readULeb128(in_stream); |
| 910 | 1050 | const mtime = try readULeb128(in_stream); |
| 911 | 1051 | const len_bytes = try readULeb128(in_stream); |
| ... | ... | @@ -963,7 +1103,7 @@ fn getLineNumberInfo(st: *DebugInfo, compile_unit: *const CompileUnit, target_ad |
| 963 | 1103 | prog.address += inc_addr; |
| 964 | 1104 | }, |
| 965 | 1105 | DW.LNS_fixed_advance_pc => { |
| 966 | const arg = try in_stream.readInt(st.elf.endian, u16); | |
| 1106 | const arg = try in_stream.readInt(di.elf.endian, u16); | |
| 967 | 1107 | prog.address += arg; |
| 968 | 1108 | }, |
| 969 | 1109 | DW.LNS_set_prologue_end => {}, |
| ... | ... | @@ -1142,7 +1282,7 @@ pub const global_allocator = &global_fixed_allocator.allocator; |
| 1142 | 1282 | var global_fixed_allocator = std.heap.ThreadSafeFixedBufferAllocator.init(global_allocator_mem[0..]); |
| 1143 | 1283 | var global_allocator_mem: [100 * 1024]u8 = undefined; |
| 1144 | 1284 | |
| 1145 | // TODO make thread safe | |
| 1285 | /// TODO multithreaded awareness | |
| 1146 | 1286 | var debug_info_allocator: ?*mem.Allocator = null; |
| 1147 | 1287 | var debug_info_direct_allocator: std.heap.DirectAllocator = undefined; |
| 1148 | 1288 | var debug_info_arena_allocator: std.heap.ArenaAllocator = undefined; |
std/macho.zig deleted-172| ... | ... | @@ -1,172 +0,0 @@ |
| 1 | const builtin = @import("builtin"); | |
| 2 | const std = @import("index.zig"); | |
| 3 | const io = std.io; | |
| 4 | const mem = std.mem; | |
| 5 | ||
| 6 | const MH_MAGIC_64 = 0xFEEDFACF; | |
| 7 | const MH_PIE = 0x200000; | |
| 8 | const LC_SYMTAB = 2; | |
| 9 | ||
| 10 | const MachHeader64 = packed struct { | |
| 11 | magic: u32, | |
| 12 | cputype: u32, | |
| 13 | cpusubtype: u32, | |
| 14 | filetype: u32, | |
| 15 | ncmds: u32, | |
| 16 | sizeofcmds: u32, | |
| 17 | flags: u32, | |
| 18 | reserved: u32, | |
| 19 | }; | |
| 20 | ||
| 21 | const LoadCommand = packed struct { | |
| 22 | cmd: u32, | |
| 23 | cmdsize: u32, | |
| 24 | }; | |
| 25 | ||
| 26 | const SymtabCommand = packed struct { | |
| 27 | symoff: u32, | |
| 28 | nsyms: u32, | |
| 29 | stroff: u32, | |
| 30 | strsize: u32, | |
| 31 | }; | |
| 32 | ||
| 33 | const Nlist64 = packed struct { | |
| 34 | n_strx: u32, | |
| 35 | n_type: u8, | |
| 36 | n_sect: u8, | |
| 37 | n_desc: u16, | |
| 38 | n_value: u64, | |
| 39 | }; | |
| 40 | ||
| 41 | pub const Symbol = struct { | |
| 42 | name: []const u8, | |
| 43 | address: u64, | |
| 44 | ||
| 45 | fn addressLessThan(lhs: Symbol, rhs: Symbol) bool { | |
| 46 | return lhs.address < rhs.address; | |
| 47 | } | |
| 48 | }; | |
| 49 | ||
| 50 | pub const SymbolTable = struct { | |
| 51 | allocator: *mem.Allocator, | |
| 52 | symbols: []const Symbol, | |
| 53 | strings: []const u8, | |
| 54 | ||
| 55 | // Doubles as an eyecatcher to calculate the PIE slide, see loadSymbols(). | |
| 56 | // Ideally we'd use _mh_execute_header because it's always at 0x100000000 | |
| 57 | // in the image but as it's located in a different section than executable | |
| 58 | // code, its displacement is different. | |
| 59 | pub fn deinit(self: *SymbolTable) void { | |
| 60 | self.allocator.free(self.symbols); | |
| 61 | self.symbols = []const Symbol{}; | |
| 62 | ||
| 63 | self.allocator.free(self.strings); | |
| 64 | self.strings = []const u8{}; | |
| 65 | } | |
| 66 | ||
| 67 | pub fn search(self: *const SymbolTable, address: usize) ?*const Symbol { | |
| 68 | var min: usize = 0; | |
| 69 | var max: usize = self.symbols.len - 1; // Exclude sentinel. | |
| 70 | while (min < max) { | |
| 71 | const mid = min + (max - min) / 2; | |
| 72 | const curr = &self.symbols[mid]; | |
| 73 | const next = &self.symbols[mid + 1]; | |
| 74 | if (address >= next.address) { | |
| 75 | min = mid + 1; | |
| 76 | } else if (address < curr.address) { | |
| 77 | max = mid; | |
| 78 | } else { | |
| 79 | return curr; | |
| 80 | } | |
| 81 | } | |
| 82 | return null; | |
| 83 | } | |
| 84 | }; | |
| 85 | ||
| 86 | pub fn loadSymbols(allocator: *mem.Allocator, in: *io.FileInStream) !SymbolTable { | |
| 87 | var file = in.file; | |
| 88 | try file.seekTo(0); | |
| 89 | ||
| 90 | var hdr: MachHeader64 = undefined; | |
| 91 | try readOneNoEof(in, MachHeader64, &hdr); | |
| 92 | if (hdr.magic != MH_MAGIC_64) return error.MissingDebugInfo; | |
| 93 | const is_pie = MH_PIE == (hdr.flags & MH_PIE); | |
| 94 | ||
| 95 | var pos: usize = @sizeOf(@typeOf(hdr)); | |
| 96 | var ncmd: u32 = hdr.ncmds; | |
| 97 | while (ncmd != 0) : (ncmd -= 1) { | |
| 98 | try file.seekTo(pos); | |
| 99 | var lc: LoadCommand = undefined; | |
| 100 | try readOneNoEof(in, LoadCommand, &lc); | |
| 101 | if (lc.cmd == LC_SYMTAB) break; | |
| 102 | pos += lc.cmdsize; | |
| 103 | } else { | |
| 104 | return error.MissingDebugInfo; | |
| 105 | } | |
| 106 | ||
| 107 | var cmd: SymtabCommand = undefined; | |
| 108 | try readOneNoEof(in, SymtabCommand, &cmd); | |
| 109 | ||
| 110 | try file.seekTo(cmd.symoff); | |
| 111 | var syms = try allocator.alloc(Nlist64, cmd.nsyms); | |
| 112 | defer allocator.free(syms); | |
| 113 | try readNoEof(in, Nlist64, syms); | |
| 114 | ||
| 115 | try file.seekTo(cmd.stroff); | |
| 116 | var strings = try allocator.alloc(u8, cmd.strsize); | |
| 117 | errdefer allocator.free(strings); | |
| 118 | try in.stream.readNoEof(strings); | |
| 119 | ||
| 120 | var nsyms: usize = 0; | |
| 121 | for (syms) |sym| | |
| 122 | if (isSymbol(sym)) nsyms += 1; | |
| 123 | if (nsyms == 0) return error.MissingDebugInfo; | |
| 124 | ||
| 125 | var symbols = try allocator.alloc(Symbol, nsyms + 1); // Room for sentinel. | |
| 126 | errdefer allocator.free(symbols); | |
| 127 | ||
| 128 | var pie_slide: usize = 0; | |
| 129 | var nsym: usize = 0; | |
| 130 | for (syms) |sym| { | |
| 131 | if (!isSymbol(sym)) continue; | |
| 132 | const start = sym.n_strx; | |
| 133 | const end = mem.indexOfScalarPos(u8, strings, start, 0).?; | |
| 134 | const name = strings[start..end]; | |
| 135 | const address = sym.n_value; | |
| 136 | symbols[nsym] = Symbol{ .name = name, .address = address }; | |
| 137 | nsym += 1; | |
| 138 | if (is_pie and mem.eql(u8, name, "_SymbolTable_deinit")) { | |
| 139 | pie_slide = @ptrToInt(SymbolTable.deinit) - address; | |
| 140 | } | |
| 141 | } | |
| 142 | ||
| 143 | // Effectively a no-op, lld emits symbols in ascending order. | |
| 144 | std.sort.sort(Symbol, symbols[0..nsyms], Symbol.addressLessThan); | |
| 145 | ||
| 146 | // Insert the sentinel. Since we don't know where the last function ends, | |
| 147 | // we arbitrarily limit it to the start address + 4 KB. | |
| 148 | const top = symbols[nsyms - 1].address + 4096; | |
| 149 | symbols[nsyms] = Symbol{ .name = "", .address = top }; | |
| 150 | ||
| 151 | if (pie_slide != 0) { | |
| 152 | for (symbols) |*symbol| | |
| 153 | symbol.address += pie_slide; | |
| 154 | } | |
| 155 | ||
| 156 | return SymbolTable{ | |
| 157 | .allocator = allocator, | |
| 158 | .symbols = symbols, | |
| 159 | .strings = strings, | |
| 160 | }; | |
| 161 | } | |
| 162 | ||
| 163 | fn readNoEof(in: *io.FileInStream, comptime T: type, result: []T) !void { | |
| 164 | return in.stream.readNoEof(@sliceToBytes(result)); | |
| 165 | } | |
| 166 | fn readOneNoEof(in: *io.FileInStream, comptime T: type, result: *T) !void { | |
| 167 | return readNoEof(in, T, (*[1]T)(result)[0..]); | |
| 168 | } | |
| 169 | ||
| 170 | fn isSymbol(sym: *const Nlist64) bool { | |
| 171 | return sym.n_value != 0 and sym.n_desc == 0; | |
| 172 | } |