libsanitizer/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cpp

   1 //===-- sanitizer_symbolizer_posix_libcdep.cpp ----------------------------===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8 //
   9 // This file is shared between AddressSanitizer and ThreadSanitizer
  10 // run-time libraries.
  11 // POSIX-specific implementation of symbolizer parts.
  12 //===----------------------------------------------------------------------===//
  13
  14 #include "sanitizer_platform.h"
  15 #if SANITIZER_POSIX
  16 #include "sanitizer_allocator_internal.h"
  17 #include "sanitizer_common.h"
  18 #include "sanitizer_file.h"
  19 #include "sanitizer_flags.h"
  20 #include "sanitizer_internal_defs.h"
  21 #include "sanitizer_linux.h"
  22 #include "sanitizer_placement_new.h"
  23 #include "sanitizer_posix.h"
  24 #include "sanitizer_procmaps.h"
  25 #include "sanitizer_symbolizer_internal.h"
  26 #include "sanitizer_symbolizer_libbacktrace.h"
  27 #include "sanitizer_symbolizer_mac.h"
  28
  29 #include <dlfcn.h>   // for dlsym()
  30 #include <errno.h>
  31 #include <stdint.h>
  32 #include <stdlib.h>
  33 #include <sys/wait.h>
  34 #include <unistd.h>
  35
  36 // C++ demangling function, as required by Itanium C++ ABI. This is weak,
  37 // because we do not require a C++ ABI library to be linked to a program
  38 // using sanitizers; if it's not present, we'll just use the mangled name.
  39 namespace __cxxabiv1 {
  40   extern "C" SANITIZER_WEAK_ATTRIBUTE
  41   char *__cxa_demangle(const char *mangled, char *buffer,
  42                                   size_t *length, int *status);
  43 }
  44
  45 namespace __sanitizer {
  46
  47 // Attempts to demangle the name via __cxa_demangle from __cxxabiv1.
  48 const char *DemangleCXXABI(const char *name) {
  49   // FIXME: __cxa_demangle aggressively insists on allocating memory.
  50   // There's not much we can do about that, short of providing our
  51   // own demangler (libc++abi's implementation could be adapted so that
  52   // it does not allocate). For now, we just call it anyway, and we leak
  53   // the returned value.
  54   if (&__cxxabiv1::__cxa_demangle)
  55     if (const char *demangled_name =
  56           __cxxabiv1::__cxa_demangle(name, 0, 0, 0))
  57       return demangled_name;
  58
  59   return name;
  60 }
  61
  62 // As of now, there are no headers for the Swift runtime. Once they are
  63 // present, we will weakly link since we do not require Swift runtime to be
  64 // linked.
  65 typedef char *(*swift_demangle_ft)(const char *mangledName,
  66                                    size_t mangledNameLength, char *outputBuffer,
  67                                    size_t *outputBufferSize, uint32_t flags);
  68 static swift_demangle_ft swift_demangle_f;
  69
  70 // This must not happen lazily at symbolication time, because dlsym uses
  71 // malloc and thread-local storage, which is not a good thing to do during
  72 // symbolication.
  73 static void InitializeSwiftDemangler() {
  74   swift_demangle_f = (swift_demangle_ft)dlsym(RTLD_DEFAULT, "swift_demangle");
  75   (void)dlerror(); // Cleanup error message in case of failure
  76 }
  77
  78 // Attempts to demangle a Swift name. The demangler will return nullptr if a
  79 // non-Swift name is passed in.
  80 const char *DemangleSwift(const char *name) {
  81   if (!name) return nullptr;
  82
  83   // Check if we are dealing with a Swift mangled name first.
  84   if (name[0] != '_' || name[1] != 'T') {
  85     return nullptr;
  86   }
  87
  88   if (swift_demangle_f)
  89     return swift_demangle_f(name, internal_strlen(name), 0, 0, 0);
  90
  91   return nullptr;
  92 }
  93
  94 const char *DemangleSwiftAndCXX(const char *name) {
  95   if (!name) return nullptr;
  96   if (const char *swift_demangled_name = DemangleSwift(name))
  97     return swift_demangled_name;
  98   return DemangleCXXABI(name);
  99 }
 100
 101 static bool CreateTwoHighNumberedPipes(int *infd_, int *outfd_) {
 102   int *infd = NULL;
 103   int *outfd = NULL;
 104   // The client program may close its stdin and/or stdout and/or stderr
 105   // thus allowing socketpair to reuse file descriptors 0, 1 or 2.
 106   // In this case the communication between the forked processes may be
 107   // broken if either the parent or the child tries to close or duplicate
 108   // these descriptors. The loop below produces two pairs of file
 109   // descriptors, each greater than 2 (stderr).
 110   int sock_pair[5][2];
 111   for (int i = 0; i < 5; i++) {
 112     if (pipe(sock_pair[i]) == -1) {
 113       for (int j = 0; j < i; j++) {
 114         internal_close(sock_pair[j][0]);
 115         internal_close(sock_pair[j][1]);
 116       }
 117       return false;
 118     } else if (sock_pair[i][0] > 2 && sock_pair[i][1] > 2) {
 119       if (infd == NULL) {
 120         infd = sock_pair[i];
 121       } else {
 122         outfd = sock_pair[i];
 123         for (int j = 0; j < i; j++) {
 124           if (sock_pair[j] == infd) continue;
 125           internal_close(sock_pair[j][0]);
 126           internal_close(sock_pair[j][1]);
 127         }
 128         break;
 129       }
 130     }
 131   }
 132   CHECK(infd);
 133   CHECK(outfd);
 134   infd_[0] = infd[0];
 135   infd_[1] = infd[1];
 136   outfd_[0] = outfd[0];
 137   outfd_[1] = outfd[1];
 138   return true;
 139 }
 140
 141 bool SymbolizerProcess::StartSymbolizerSubprocess() {
 142   if (!FileExists(path_)) {
 143     if (!reported_invalid_path_) {
 144       Report("WARNING: invalid path to external symbolizer!\n");
 145       reported_invalid_path_ = true;
 146     }
 147     return false;
 148   }
 149
 150   const char *argv[kArgVMax];
 151   GetArgV(path_, argv);
 152   pid_t pid;
 153
 154   // Report how symbolizer is being launched for debugging purposes.
 155   if (Verbosity() >= 3) {
 156     // Only use `Report` for first line so subsequent prints don't get prefixed
 157     // with current PID.
 158     Report("Launching Symbolizer process: ");
 159     for (unsigned index = 0; index < kArgVMax && argv[index]; ++index)
 160       Printf("%s ", argv[index]);
 161     Printf("\n");
 162   }
 163
 164   if (use_posix_spawn_) {
 165 #if SANITIZER_MAC
 166     fd_t fd = internal_spawn(argv, const_cast<const char **>(GetEnvP()), &pid);
 167     if (fd == kInvalidFd) {
 168       Report("WARNING: failed to spawn external symbolizer (errno: %d)\n",
 169              errno);
 170       return false;
 171     }
 172
 173     input_fd_ = fd;
 174     output_fd_ = fd;
 175 #else  // SANITIZER_MAC
 176     UNIMPLEMENTED();
 177 #endif  // SANITIZER_MAC
 178   } else {
 179     fd_t infd[2] = {}, outfd[2] = {};
 180     if (!CreateTwoHighNumberedPipes(infd, outfd)) {
 181       Report("WARNING: Can't create a socket pair to start "
 182              "external symbolizer (errno: %d)\n", errno);
 183       return false;
 184     }
 185
 186     pid = StartSubprocess(path_, argv, GetEnvP(), /* stdin */ outfd[0],
 187                           /* stdout */ infd[1]);
 188     if (pid < 0) {
 189       internal_close(infd[0]);
 190       internal_close(outfd[1]);
 191       return false;
 192     }
 193
 194     input_fd_ = infd[0];
 195     output_fd_ = outfd[1];
 196   }
 197
 198   CHECK_GT(pid, 0);
 199
 200   // Check that symbolizer subprocess started successfully.
 201   SleepForMillis(kSymbolizerStartupTimeMillis);
 202   if (!IsProcessRunning(pid)) {
 203     // Either waitpid failed, or child has already exited.
 204     Report("WARNING: external symbolizer didn't start up correctly!\n");
 205     return false;
 206   }
 207
 208   return true;
 209 }
 210
 211 class Addr2LineProcess : public SymbolizerProcess {
 212  public:
 213   Addr2LineProcess(const char *path, const char *module_name)
 214       : SymbolizerProcess(path), module_name_(internal_strdup(module_name)) {}
 215
 216   const char *module_name() const { return module_name_; }
 217
 218  private:
 219   void GetArgV(const char *path_to_binary,
 220                const char *(&argv)[kArgVMax]) const override {
 221     int i = 0;
 222     argv[i++] = path_to_binary;
 223     argv[i++] = "-iCfe";
 224     argv[i++] = module_name_;
 225     argv[i++] = nullptr;
 226   }
 227
 228   bool ReachedEndOfOutput(const char *buffer, uptr length) const override;
 229
 230   bool ReadFromSymbolizer(char *buffer, uptr max_length) override {
 231     if (!SymbolizerProcess::ReadFromSymbolizer(buffer, max_length))
 232       return false;
 233     // The returned buffer is empty when output is valid, but exceeds
 234     // max_length.
 235     if (*buffer == '\0')
 236       return true;
 237     // We should cut out output_terminator_ at the end of given buffer,
 238     // appended by addr2line to mark the end of its meaningful output.
 239     // We cannot scan buffer from it's beginning, because it is legal for it
 240     // to start with output_terminator_ in case given offset is invalid. So,
 241     // scanning from second character.
 242     char *garbage = internal_strstr(buffer + 1, output_terminator_);
 243     // This should never be NULL since buffer must end up with
 244     // output_terminator_.
 245     CHECK(garbage);
 246     // Trim the buffer.
 247     garbage[0] = '\0';
 248     return true;
 249   }
 250
 251   const char *module_name_;  // Owned, leaked.
 252   static const char output_terminator_[];
 253 };
 254
 255 const char Addr2LineProcess::output_terminator_[] = "??\n??:0\n";
 256
 257 bool Addr2LineProcess::ReachedEndOfOutput(const char *buffer,
 258                                           uptr length) const {
 259   const size_t kTerminatorLen = sizeof(output_terminator_) - 1;
 260   // Skip, if we read just kTerminatorLen bytes, because Addr2Line output
 261   // should consist at least of two pairs of lines:
 262   // 1. First one, corresponding to given offset to be symbolized
 263   // (may be equal to output_terminator_, if offset is not valid).
 264   // 2. Second one for output_terminator_, itself to mark the end of output.
 265   if (length <= kTerminatorLen) return false;
 266   // Addr2Line output should end up with output_terminator_.
 267   return !internal_memcmp(buffer + length - kTerminatorLen,
 268                           output_terminator_, kTerminatorLen);
 269 }
 270
 271 class Addr2LinePool : public SymbolizerTool {
 272  public:
 273   explicit Addr2LinePool(const char *addr2line_path,
 274                          LowLevelAllocator *allocator)
 275       : addr2line_path_(addr2line_path), allocator_(allocator) {
 276     addr2line_pool_.reserve(16);
 277   }
 278
 279   bool SymbolizePC(uptr addr, SymbolizedStack *stack) override {
 280     if (const char *buf =
 281             SendCommand(stack->info.module, stack->info.module_offset)) {
 282       ParseSymbolizePCOutput(buf, stack);
 283       return true;
 284     }
 285     return false;
 286   }
 287
 288   bool SymbolizeData(uptr addr, DataInfo *info) override {
 289     return false;
 290   }
 291
 292  private:
 293   const char *SendCommand(const char *module_name, uptr module_offset) {
 294     Addr2LineProcess *addr2line = 0;
 295     for (uptr i = 0; i < addr2line_pool_.size(); ++i) {
 296       if (0 ==
 297           internal_strcmp(module_name, addr2line_pool_[i]->module_name())) {
 298         addr2line = addr2line_pool_[i];
 299         break;
 300       }
 301     }
 302     if (!addr2line) {
 303       addr2line =
 304           new(*allocator_) Addr2LineProcess(addr2line_path_, module_name);
 305       addr2line_pool_.push_back(addr2line);
 306     }
 307     CHECK_EQ(0, internal_strcmp(module_name, addr2line->module_name()));
 308     char buffer[kBufferSize];
 309     internal_snprintf(buffer, kBufferSize, "0x%zx\n0x%zx\n",
 310                       module_offset, dummy_address_);
 311     return addr2line->SendCommand(buffer);
 312   }
 313
 314   static const uptr kBufferSize = 64;
 315   const char *addr2line_path_;
 316   LowLevelAllocator *allocator_;
 317   InternalMmapVector<Addr2LineProcess*> addr2line_pool_;
 318   static const uptr dummy_address_ =
 319       FIRST_32_SECOND_64(UINT32_MAX, UINT64_MAX);
 320 };
 321
 322 #if SANITIZER_SUPPORTS_WEAK_HOOKS
 323 extern "C" {
 324 SANITIZER_INTERFACE_ATTRIBUTE SANITIZER_WEAK_ATTRIBUTE
 325 bool __sanitizer_symbolize_code(const char *ModuleName, u64 ModuleOffset,
 326                                 char *Buffer, int MaxLength);
 327 SANITIZER_INTERFACE_ATTRIBUTE SANITIZER_WEAK_ATTRIBUTE
 328 bool __sanitizer_symbolize_data(const char *ModuleName, u64 ModuleOffset,
 329                                 char *Buffer, int MaxLength);
 330 SANITIZER_INTERFACE_ATTRIBUTE SANITIZER_WEAK_ATTRIBUTE
 331 void __sanitizer_symbolize_flush();
 332 SANITIZER_INTERFACE_ATTRIBUTE SANITIZER_WEAK_ATTRIBUTE
 333 int __sanitizer_symbolize_demangle(const char *Name, char *Buffer,
 334                                    int MaxLength);
 335 }  // extern "C"
 336
 337 class InternalSymbolizer : public SymbolizerTool {
 338  public:
 339   static InternalSymbolizer *get(LowLevelAllocator *alloc) {
 340     if (__sanitizer_symbolize_code != 0 &&
 341         __sanitizer_symbolize_data != 0) {
 342       return new(*alloc) InternalSymbolizer();
 343     }
 344     return 0;
 345   }
 346
 347   bool SymbolizePC(uptr addr, SymbolizedStack *stack) override {
 348     bool result = __sanitizer_symbolize_code(
 349         stack->info.module, stack->info.module_offset, buffer_, kBufferSize);
 350     if (result) ParseSymbolizePCOutput(buffer_, stack);
 351     return result;
 352   }
 353
 354   bool SymbolizeData(uptr addr, DataInfo *info) override {
 355     bool result = __sanitizer_symbolize_data(info->module, info->module_offset,
 356                                              buffer_, kBufferSize);
 357     if (result) {
 358       ParseSymbolizeDataOutput(buffer_, info);
 359       info->start += (addr - info->module_offset);  // Add the base address.
 360     }
 361     return result;
 362   }
 363
 364   void Flush() override {
 365     if (__sanitizer_symbolize_flush)
 366       __sanitizer_symbolize_flush();
 367   }
 368
 369   const char *Demangle(const char *name) override {
 370     if (__sanitizer_symbolize_demangle) {
 371       for (uptr res_length = 1024;
 372            res_length <= InternalSizeClassMap::kMaxSize;) {
 373         char *res_buff = static_cast<char*>(InternalAlloc(res_length));
 374         uptr req_length =
 375             __sanitizer_symbolize_demangle(name, res_buff, res_length);
 376         if (req_length > res_length) {
 377           res_length = req_length + 1;
 378           InternalFree(res_buff);
 379           continue;
 380         }
 381         return res_buff;
 382       }
 383     }
 384     return name;
 385   }
 386
 387  private:
 388   InternalSymbolizer() { }
 389
 390   static const int kBufferSize = 16 * 1024;
 391   char buffer_[kBufferSize];
 392 };
 393 #else  // SANITIZER_SUPPORTS_WEAK_HOOKS
 394
 395 class InternalSymbolizer : public SymbolizerTool {
 396  public:
 397   static InternalSymbolizer *get(LowLevelAllocator *alloc) { return 0; }
 398 };
 399
 400 #endif  // SANITIZER_SUPPORTS_WEAK_HOOKS
 401
 402 const char *Symbolizer::PlatformDemangle(const char *name) {
 403   return DemangleSwiftAndCXX(name);
 404 }
 405
 406 static SymbolizerTool *ChooseExternalSymbolizer(LowLevelAllocator *allocator) {
 407   const char *path = common_flags()->external_symbolizer_path;
 408   const char *binary_name = path ? StripModuleName(path) : "";
 409   if (path && path[0] == '\0') {
 410     VReport(2, "External symbolizer is explicitly disabled.\n");
 411     return nullptr;
 412   } else if (!internal_strcmp(binary_name, "llvm-symbolizer")) {
 413     VReport(2, "Using llvm-symbolizer at user-specified path: %s\n", path);
 414     return new(*allocator) LLVMSymbolizer(path, allocator);
 415   } else if (!internal_strcmp(binary_name, "atos")) {
 416 #if SANITIZER_MAC
 417     VReport(2, "Using atos at user-specified path: %s\n", path);
 418     return new(*allocator) AtosSymbolizer(path, allocator);
 419 #else  // SANITIZER_MAC
 420     Report("ERROR: Using `atos` is only supported on Darwin.\n");
 421     Die();
 422 #endif  // SANITIZER_MAC
 423   } else if (!internal_strcmp(binary_name, "addr2line")) {
 424     VReport(2, "Using addr2line at user-specified path: %s\n", path);
 425     return new(*allocator) Addr2LinePool(path, allocator);
 426   } else if (path) {
 427     Report("ERROR: External symbolizer path is set to '%s' which isn't "
 428            "a known symbolizer. Please set the path to the llvm-symbolizer "
 429            "binary or other known tool.\n", path);
 430     Die();
 431   }
 432
 433   // Otherwise symbolizer program is unknown, let's search $PATH
 434   CHECK(path == nullptr);
 435 #if SANITIZER_MAC
 436   if (const char *found_path = FindPathToBinary("atos")) {
 437     VReport(2, "Using atos found at: %s\n", found_path);
 438     return new(*allocator) AtosSymbolizer(found_path, allocator);
 439   }
 440 #endif  // SANITIZER_MAC
 441   if (const char *found_path = FindPathToBinary("llvm-symbolizer")) {
 442     VReport(2, "Using llvm-symbolizer found at: %s\n", found_path);
 443     return new(*allocator) LLVMSymbolizer(found_path, allocator);
 444   }
 445   if (common_flags()->allow_addr2line) {
 446     if (const char *found_path = FindPathToBinary("addr2line")) {
 447       VReport(2, "Using addr2line found at: %s\n", found_path);
 448       return new(*allocator) Addr2LinePool(found_path, allocator);
 449     }
 450   }
 451   return nullptr;
 452 }
 453
 454 static void ChooseSymbolizerTools(IntrusiveList<SymbolizerTool> *list,
 455                                   LowLevelAllocator *allocator) {
 456   if (!common_flags()->symbolize) {
 457     VReport(2, "Symbolizer is disabled.\n");
 458     return;
 459   }
 460   if (IsAllocatorOutOfMemory()) {
 461     VReport(2, "Cannot use internal symbolizer: out of memory\n");
 462   } else if (SymbolizerTool *tool = InternalSymbolizer::get(allocator)) {
 463     VReport(2, "Using internal symbolizer.\n");
 464     list->push_back(tool);
 465     return;
 466   }
 467   if (SymbolizerTool *tool = LibbacktraceSymbolizer::get(allocator)) {
 468     VReport(2, "Using libbacktrace symbolizer.\n");
 469     list->push_back(tool);
 470     return;
 471   }
 472
 473   if (SymbolizerTool *tool = ChooseExternalSymbolizer(allocator)) {
 474     list->push_back(tool);
 475   }
 476
 477 #if SANITIZER_MAC
 478   VReport(2, "Using dladdr symbolizer.\n");
 479   list->push_back(new(*allocator) DlAddrSymbolizer());
 480 #endif  // SANITIZER_MAC
 481 }
 482
 483 Symbolizer *Symbolizer::PlatformInit() {
 484   IntrusiveList<SymbolizerTool> list;
 485   list.clear();
 486   ChooseSymbolizerTools(&list, &symbolizer_allocator_);
 487   return new(symbolizer_allocator_) Symbolizer(list);
 488 }
 489
 490 void Symbolizer::LateInitialize() {
 491   Symbolizer::GetOrInit()->LateInitializeTools();
 492   InitializeSwiftDemangler();
 493 }
 494
 495 }  // namespace __sanitizer
 496
 497 #endif  // SANITIZER_POSIX