[compiler-rt] [sanitizer_symbolizer] Symbolizer Markup for linux. (PR #65543)

[Florian Mayer via llvm-commits]

================
@@ -11,20 +11,140 @@
 // Implementation of offline markup symbolizer.
 //===----------------------------------------------------------------------===//
 
-#include "sanitizer_platform.h"
-#if SANITIZER_SYMBOLIZER_MARKUP
+#include "sanitizer_symbolizer_markup.h"
 
-#if SANITIZER_FUCHSIA
+#include "sanitizer_common.h"
+#include "sanitizer_libc.h"
+#include "sanitizer_platform.h"
+#include "sanitizer_stacktrace.h"
+#include "sanitizer_symbolizer.h"
 #include "sanitizer_symbolizer_fuchsia.h"
-#  endif
 
-#  include <limits.h>
-#  include <unwind.h>
+namespace __sanitizer {
 
-#  include "sanitizer_stacktrace.h"
-#  include "sanitizer_symbolizer.h"
+void RenderDataMarkup(InternalScopedString *buffer, const char *format,
+                      const DataInfo *DI, const char *strip_path_prefix) {
+  buffer->append(kFormatData, DI->start);
+}
 
-namespace __sanitizer {
+bool RenderNeedsSymbolizationMarkup(const char *format) { return false; }
+
+void RenderFrameMarkup(InternalScopedString *buffer, const char *format,
+                       int frame_no, uptr address, const AddressInfo *info,
+                       bool vs_style, const char *strip_path_prefix) {
+  CHECK(!RenderNeedsSymbolizationMarkup(format));
+  buffer->append(kFormatFrame, frame_no, address);
+}
+
+// Simplier view of a LoadedModule. It only holds information necessary to
+// identify unique modules.
+struct RenderedModule {
+  char *full_name;
+  u8 uuid[kModuleUUIDSize];  // BuildId
+  uptr base_address;
+};
+
+bool ModulesEq(const LoadedModule *module,
+               const RenderedModule *renderedModule) {
+  return module->base_address() == renderedModule->base_address &&
+         internal_memcmp(module->uuid(), renderedModule->uuid,
+                         module->uuid_size()) == 0 &&
+         internal_strcmp(module->full_name(), renderedModule->full_name) == 0;
+}
+
+bool ModuleHasBeenRendered(
+    const LoadedModule *module,
+    const InternalMmapVectorNoCtor<RenderedModule> *renderedModules) {
+  for (auto *it = renderedModules->begin(); it != renderedModules->end();
+       ++it) {
+    const auto &renderedModule = *it;
+    if (ModulesEq(module, &renderedModule)) {
+      return true;
+    }
+  }
+  return false;
+}
+
+void RenderModulesMarkup(InternalScopedString *buffer,
+                         const ListOfModules *modules) {
+  // Keeps track of the modules that have been rendered.
+  static bool initialized = false;
+  static InternalMmapVectorNoCtor<RenderedModule> renderedModules;
+  if (!initialized) {
+    renderedModules.Initialize(modules->size());
+    initialized = true;
+  }
+
+  if (!renderedModules.size()) {
+    buffer->append("{{{reset}}}\n");
+  }
+
+  for (auto *moduleIt = modules->begin(); moduleIt != modules->end();
+       ++moduleIt) {
+    const LoadedModule &module = *moduleIt;
+
+    if (ModuleHasBeenRendered(&module, &renderedModules)) {
+      continue;
+    }
+
+    buffer->append("{{{module:%d:%s:elf:", renderedModules.size(),
+                   module.full_name());
+    for (uptr i = 0; i < module.uuid_size(); i++) {
+      buffer->append("%02x", module.uuid()[i]);
+    }
+    buffer->append("}}}\n");
+
+    const auto ranges = module.ranges();
----------------
fmayer wrote:

Why do we need this copy? We don't seem to be touching `module` in the loop.

https://github.com/llvm/llvm-project/pull/65543