[sanitizers] include build ids in stacks on linux.

Reviewed By: eugenis

Differential Revision: https://reviews.llvm.org/D114294

compiler-rt/lib/sanitizer_common/sanitizer_common.cpp

@@ -138,9 +138,17 @@ void LoadedModule::set(const char *module_name, uptr base_address,
   set(module_name, base_address);
   arch_ = arch;
   internal_memcpy(uuid_, uuid, sizeof(uuid_));
+  uuid_size_ = kModuleUUIDSize;
   instrumented_ = instrumented;
 }
 
+void LoadedModule::setUuid(const char *uuid, uptr size) {
+  if (size > kModuleUUIDSize)
+    size = kModuleUUIDSize;
+  internal_memcpy(uuid_, uuid, size);
+  uuid_size_ = size;
+}
+
 void LoadedModule::clear() {
   InternalFree(full_name_);
   base_address_ = 0;

compiler-rt/lib/sanitizer_common/sanitizer_common.h

@@ -770,7 +770,7 @@ inline const char *ModuleArchToString(ModuleArch arch) {
   return "";
 }
 
-const uptr kModuleUUIDSize = 16;
+const uptr kModuleUUIDSize = 32;
 const uptr kMaxSegName = 16;
 
 // Represents a binary loaded into virtual memory (e.g. this can be an
@@ -782,6 +782,7 @@ class LoadedModule {
         base_address_(0),
         max_executable_address_(0),
         arch_(kModuleArchUnknown),
+        uuid_size_(0),
         instrumented_(false) {
     internal_memset(uuid_, 0, kModuleUUIDSize);
     ranges_.clear();
@@ -789,6 +790,7 @@ class LoadedModule {
   void set(const char *module_name, uptr base_address);
   void set(const char *module_name, uptr base_address, ModuleArch arch,
            u8 uuid[kModuleUUIDSize], bool instrumented);
+  void setUuid(const char *uuid, uptr size);
   void clear();
   void addAddressRange(uptr beg, uptr end, bool executable, bool writable,
                        const char *name = nullptr);
@@ -799,6 +801,7 @@ class LoadedModule {
   uptr max_executable_address() const { return max_executable_address_; }
   ModuleArch arch() const { return arch_; }
   const u8 *uuid() const { return uuid_; }
+  uptr uuid_size() const { return uuid_size_; }
   bool instrumented() const { return instrumented_; }
 
   struct AddressRange {
@@ -827,6 +830,7 @@ class LoadedModule {
   uptr base_address_;
   uptr max_executable_address_;
   ModuleArch arch_;
+  uptr uuid_size_;
   u8 uuid_[kModuleUUIDSize];
   bool instrumented_;
   IntrusiveList<AddressRange> ranges_;

compiler-rt/lib/sanitizer_common/sanitizer_linux_libcdep.cpp

@@ -603,6 +603,32 @@ static int AddModuleSegments(const char *module_name, dl_phdr_info *info,
       bool writable = phdr->p_flags & PF_W;
       cur_module.addAddressRange(cur_beg, cur_end, executable,
                                  writable);
+    } else if (phdr->p_type == PT_NOTE) {
+      uptr off = 0;
+      while (off < phdr->p_memsz - sizeof(ElfW(Nhdr))) {
+        auto *nhdr = reinterpret_cast<const ElfW(Nhdr) *>(info->dlpi_addr +
+                                                          phdr->p_vaddr + off);
+        constexpr auto kGnuNamesz = 4;  // "GNU" with NUL-byte.
+        static_assert(kGnuNamesz % 4 == 0, "kGnuNameSize is aligned to 4.");
+        if (nhdr->n_type == NT_GNU_BUILD_ID && nhdr->n_namesz == kGnuNamesz) {
+          if (off + sizeof(ElfW(Nhdr)) + nhdr->n_namesz + nhdr->n_descsz >
+              phdr->p_memsz) {
+            // Something is very wrong, bail out instead of reading potentially
+            // arbitrary memory.
+            break;
+          }
+          const char *name =
+              reinterpret_cast<const char *>(nhdr) + sizeof(*nhdr);
+          if (internal_memcmp(name, "GNU", 3) == 0) {
+            const char *value = reinterpret_cast<const char *>(nhdr) +
+                                sizeof(*nhdr) + kGnuNamesz;
+            cur_module.setUuid(value, nhdr->n_descsz);
+            break;
+          }
+        }
+        off += sizeof(*nhdr) + RoundUpTo(nhdr->n_namesz, 4) +
+               RoundUpTo(nhdr->n_descsz, 4);
+      }
     }
   }
   modules->push_back(cur_module);

compiler-rt/lib/sanitizer_common/sanitizer_stacktrace_printer.cpp

@@ -104,6 +104,19 @@ static const char *DemangleFunctionName(const char *function) {
   return function;
 }
 
+static void MaybeBuildIdToBuffer(const AddressInfo &info, bool PrefixSpace,
+                                 InternalScopedString *buffer) {
+  if (info.uuid_size) {
+    if (PrefixSpace)
+      buffer->append(" ");
+    buffer->append("(BuildId: ");
+    for (uptr i = 0; i < info.uuid_size; ++i) {
+      buffer->append("%02x", info.uuid[i]);
+    }
+    buffer->append(")");
+  }
+}
+
 static const char kDefaultFormat[] = "    #%n %p %F %L";
 
 void RenderFrame(InternalScopedString *buffer, const char *format, int frame_no,
@@ -140,6 +153,9 @@ void RenderFrame(InternalScopedString *buffer, const char *format, int frame_no,
     case 'o':
       buffer->append("0x%zx", info->module_offset);
       break;
+    case 'b':
+      MaybeBuildIdToBuffer(*info, /*PrefixSpace=*/false, buffer);
+      break;
     case 'f':
       buffer->append("%s", DemangleFunctionName(StripFunctionName(
                                info->function, strip_func_prefix)));
@@ -181,6 +197,8 @@ void RenderFrame(InternalScopedString *buffer, const char *format, int frame_no,
       } else if (info->module) {
         RenderModuleLocation(buffer, info->module, info->module_offset,
                              info->module_arch, strip_path_prefix);
+
+        MaybeBuildIdToBuffer(*info, /*PrefixSpace=*/true, buffer);
       } else {
         buffer->append("(<unknown module>)");
       }
@@ -193,6 +211,7 @@ void RenderFrame(InternalScopedString *buffer, const char *format, int frame_no,
         // Always strip the module name for %M.
         RenderModuleLocation(buffer, StripModuleName(info->module),
                              info->module_offset, info->module_arch, "");
+        MaybeBuildIdToBuffer(*info, /*PrefixSpace=*/true, buffer);
       } else {
         buffer->append("(%p)", (void *)address);
       }

compiler-rt/lib/sanitizer_common/sanitizer_symbolizer.cpp

@@ -11,10 +11,11 @@
 //===----------------------------------------------------------------------===//
 
 #include "sanitizer_allocator_internal.h"
-#include "sanitizer_platform.h"
+#include "sanitizer_common.h"
 #include "sanitizer_internal_defs.h"
 #include "sanitizer_libc.h"
 #include "sanitizer_placement_new.h"
+#include "sanitizer_platform.h"
 #include "sanitizer_symbolizer_internal.h"
 
 namespace __sanitizer {
@@ -30,6 +31,7 @@ void AddressInfo::Clear() {
   InternalFree(file);
   internal_memset(this, 0, sizeof(AddressInfo));
   function_offset = kUnknown;
+  uuid_size = 0;
 }
 
 void AddressInfo::FillModuleInfo(const char *mod_name, uptr mod_offset,
@@ -37,6 +39,16 @@ void AddressInfo::FillModuleInfo(const char *mod_name, uptr mod_offset,
   module = internal_strdup(mod_name);
   module_offset = mod_offset;
   module_arch = mod_arch;
+  uuid_size = 0;
+}
+
+void AddressInfo::FillModuleInfo(const LoadedModule &mod) {
+  module = internal_strdup(mod.full_name());
+  module_offset = address - mod.base_address();
+  module_arch = mod.arch();
+  if (mod.uuid_size())
+    internal_memcpy(uuid, mod.uuid(), mod.uuid_size());
+  uuid_size = mod.uuid_size();
 }
 
 SymbolizedStack::SymbolizedStack() : next(nullptr), info() {}

compiler-rt/lib/sanitizer_common/sanitizer_symbolizer.h

@@ -32,6 +32,8 @@ struct AddressInfo {
   char *module;
   uptr module_offset;
   ModuleArch module_arch;
+  u8 uuid[kModuleUUIDSize];
+  uptr uuid_size;
 
   static const uptr kUnknown = ~(uptr)0;
   char *function;
@@ -45,6 +47,8 @@ struct AddressInfo {
   // Deletes all strings and resets all fields.
   void Clear();
   void FillModuleInfo(const char *mod_name, uptr mod_offset, ModuleArch arch);
+  void FillModuleInfo(const LoadedModule &mod);
+  uptr module_base() const { return address - module_offset; }
 };
 
 // Linked list of symbolized frames (each frame is described by AddressInfo).

compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cpp

@@ -84,15 +84,12 @@ const char *ExtractTokenUpToDelimiter(const char *str, const char *delimiter,
 
 SymbolizedStack *Symbolizer::SymbolizePC(uptr addr) {
   Lock l(&mu_);
-  const char *module_name = nullptr;
-  uptr module_offset;
-  ModuleArch arch;
   SymbolizedStack *res = SymbolizedStack::New(addr);
-  if (!FindModuleNameAndOffsetForAddress(addr, &module_name, &module_offset,
+  auto *mod = FindModuleForAddress(addr);
-                                         &arch))
+  if (!mod)
     return res;
   // Always fill data about module name and offset.
-  res->info.FillModuleInfo(module_name, module_offset, arch);
+  res->info.FillModuleInfo(*mod);
   for (auto &tool : tools_) {
     SymbolizerScope sym_scope(this);
     if (tool.SymbolizePC(addr, res)) {

compiler-rt/lib/sanitizer_common/tests/sanitizer_stacktrace_printer_test.cpp

@@ -112,6 +112,28 @@ TEST(SanitizerStacktracePrinter, RenderFrame) {
   EXPECT_STREQ("(/path/to/module+0x200)", str.data());
   str.clear();
 
+  RenderFrame(&str, "%b", frame_no, info.address, &info, false);
+  EXPECT_STREQ("", str.data());
+  str.clear();
+
+  info.uuid_size = 2;
+  info.uuid[0] = 0x55;
+  info.uuid[1] = 0x66;
+
+  RenderFrame(&str, "%M", frame_no, info.address, &info, false);
+  EXPECT_NE(nullptr, internal_strstr(str.data(), "(module+0x"));
+  EXPECT_NE(nullptr, internal_strstr(str.data(), "200"));
+  EXPECT_NE(nullptr, internal_strstr(str.data(), "BuildId: 5566"));
+  str.clear();
+
+  RenderFrame(&str, "%L", frame_no, info.address, &info, false);
+  EXPECT_STREQ("(/path/to/module+0x200) (BuildId: 5566)", str.data());
+  str.clear();
+
+  RenderFrame(&str, "%b", frame_no, info.address, &info, false);
+  EXPECT_STREQ("(BuildId: 5566)", str.data());
+  str.clear();
+
   info.function = internal_strdup("my_function");
   RenderFrame(&str, "%F", frame_no, info.address, &info, false);
   EXPECT_STREQ("in my_function", str.data());

compiler-rt/test/hwasan/TestCases/build-ids.c

@@ -0,0 +1,17 @@
+// RUN: %clang_hwasan -Wl,--build-id=0xaba493998257fbdd %s -o %t
+// RUN: %env_hwasan_opts=symbolize=0 not %run %t 2>&1 | FileCheck %s --check-prefixes=CHECK,NOSYM
+// RUN: not %run %t 2>&1 | FileCheck %s --check-prefixes=CHECK,SYM
+
+#include <stdlib.h>
+
+#include <sanitizer/hwasan_interface.h>
+
+int main(int argc, char **argv) {
+  __hwasan_enable_allocator_tagging();
+  char *buf = (char *)malloc(1);
+  buf[32] = 'x';
+  // CHECK: ERROR: HWAddressSanitizer: tag-mismatch
+  // NOSYM:  #0 0x{{.*}}  {{.*}}build-ids.c{{.*}} (BuildId: aba493998257fbdd)
+  // SYM:  #0 0x{{.*}} in main {{.*}}build-ids.c:[[@LINE-3]]:{{[0-9]+}}
+  return 0;
+}