Reland"[lldb] Add count for errors of DWO files in statistics and combine DWO file count functions" (#156980)

This relands changes in #155023
for adding a count of dwo errors and combine all the dwo related stats
into one struct.

The previous PR was reverted in
#156777 as the newly added unit
test `test_dwo_id_mismatch_error_stats` sometimes fails due to
inappropriate use of `glob.glob`.
This change modified the tests created in the former PR to collect and
modify the dwo files by there names instead of using index after
`glob.glob`. This will avoid the possible failure in these tests if the
order of dwo files changes.

[Original PR:
https://github.com/llvm/llvm-project/pull/155023](https://github.com/llvm/llvm-project/pull/155023)

## Testing
Ran unit tests
```
$ ./bin/llvm-lit /data/users/ziyiww/llvm-project/lldb/test/API/commands/statistics/basic/TestStats.py
./bin/llvm-lit /data/users/ziyiww/llvm-project/lldb/test/API/commands/statistics/basic/TestStats.py -v
-- Testing: 1 tests, 1 workers --
PASS: lldb-api :: commands/statistics/basic/TestStats.py (1 of 1)

Testing Time: 388.52s

Total Discovered Tests: 1
  Passed: 1 (100.00%)

$ bin/lldb-dotest -p TestStats.py /data/users/ziyiww/llvm-project/lldb/test/API/commands/statistics/basic/
----------------------------------------------------------------------
Ran 27 tests in 386.302s

OK (skipped=3)
```

Author: zw3917

lldb/include/lldb/Symbol/SymbolFile.h

@@ -488,13 +488,16 @@ class SymbolFile : public PluginInterface {
     return false;
   };
 
-  /// Get number of loaded/parsed DWO files. This is emitted in "statistics
-  /// dump"
+  /// Retrieves statistics about DWO files associated with this symbol file.
+  /// This function returns a DWOStats struct containing:
+  ///   - The number of successfully loaded/parsed DWO files.
+  ///   - The total number of DWO files encountered.
+  ///   - The number of DWO CUs that failed to load due to errors.
+  /// If this symbol file does not support DWO files, all counts will be zero.
   ///
   /// \returns
-  ///     A pair containing (loaded_dwo_count, total_dwo_count). If this
-  ///     symbol file doesn't support DWO files, both counts will be 0.
-  virtual std::pair<uint32_t, uint32_t> GetDwoFileCounts() { return {0, 0}; }
+  ///   A DWOStats struct with loaded, total, and error counts for DWO files.
+  virtual DWOStats GetDwoStats() { return {}; }
 
   virtual lldb::TypeSP
   MakeType(lldb::user_id_t uid, ConstString name,

lldb/include/lldb/Target/Statistics.h

@@ -123,6 +123,25 @@ struct StatsSuccessFail {
   uint32_t failures = 0;
 };
 
+/// Holds statistics about DWO (Debug With Object) files.
+struct DWOStats {
+  uint32_t loaded_dwo_file_count = 0;
+  uint32_t dwo_file_count = 0;
+  uint32_t dwo_error_count = 0;
+
+  DWOStats &operator+=(const DWOStats &rhs) {
+    loaded_dwo_file_count += rhs.loaded_dwo_file_count;
+    dwo_file_count += rhs.dwo_file_count;
+    dwo_error_count += rhs.dwo_error_count;
+    return *this;
+  }
+
+  friend DWOStats operator+(DWOStats lhs, const DWOStats &rhs) {
+    lhs += rhs;
+    return lhs;
+  }
+};
+
 /// A class that represents statistics for a since lldb_private::Module.
 struct ModuleStats {
   llvm::json::Value ToJSON() const;
@@ -153,8 +172,7 @@ struct ModuleStats {
   bool symtab_stripped = false;
   bool debug_info_had_variable_errors = false;
   bool debug_info_had_incomplete_types = false;
-  uint32_t dwo_file_count = 0;
-  uint32_t loaded_dwo_file_count = 0;
+  DWOStats dwo_stats;
 };
 
 struct ConstStringStats {

lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp

@@ -4623,9 +4623,8 @@ void SymbolFileDWARF::GetCompileOptions(
   }
 }
 
-std::pair<uint32_t, uint32_t> SymbolFileDWARF::GetDwoFileCounts() {
-  uint32_t total_dwo_count = 0;
-  uint32_t loaded_dwo_count = 0;
+DWOStats SymbolFileDWARF::GetDwoStats() {
+  DWOStats stats;
 
   DWARFDebugInfo &info = DebugInfo();
   const size_t num_cus = info.GetNumUnits();
@@ -4638,16 +4637,21 @@ std::pair<uint32_t, uint32_t> SymbolFileDWARF::GetDwoFileCounts() {
     if (!dwarf_cu->GetDWOId().has_value())
       continue;
 
-    total_dwo_count++;
+    stats.dwo_file_count++;
 
     // If we have a DWO symbol file, that means we were able to successfully
     // load it.
     SymbolFile *dwo_symfile =
         dwarf_cu->GetDwoSymbolFile(/*load_all_debug_info=*/false);
     if (dwo_symfile) {
-      loaded_dwo_count++;
+      stats.loaded_dwo_file_count++;
     }
+
+    // Check if this unit has a DWO load error, false by default.
+    const Status &dwo_error = dwarf_cu->GetDwoError();
+    if (dwo_error.Fail())
+      stats.dwo_error_count++;
   }
 
-  return {loaded_dwo_count, total_dwo_count};
+  return stats;
 }

lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.h

@@ -283,10 +283,11 @@ class SymbolFileDWARF : public SymbolFileCommon {
   bool GetSeparateDebugInfo(StructuredData::Dictionary &d, bool errors_only,
                             bool load_all_debug_info = false) override;
 
-  // Gets a pair of loaded and total dwo file counts.
-  // For split-dwarf files, this reports the counts for successfully loaded DWO
-  // CUs and total DWO CUs. For non-split-dwarf files, this reports 0 for both.
-  std::pair<uint32_t, uint32_t> GetDwoFileCounts() override;
+  /// Gets statistics about dwo files associated with this symbol file.
+  /// For split-dwarf files, this reports the counts for successfully loaded DWO
+  /// CUs, total DWO CUs, and the number of DWO CUs with loading errors.
+  /// For non-split-dwarf files, this reports 0 for all.
+  DWOStats GetDwoStats() override;
 
   DWARFContext &GetDWARFContext() { return m_context; }
 

lldb/source/Target/Statistics.cpp

@@ -73,8 +73,9 @@ json::Value ModuleStats::ToJSON() const {
                      debug_info_had_incomplete_types);
   module.try_emplace("symbolTableStripped", symtab_stripped);
   module.try_emplace("symbolTableSymbolCount", symtab_symbol_count);
-  module.try_emplace("dwoFileCount", dwo_file_count);
-  module.try_emplace("loadedDwoFileCount", loaded_dwo_file_count);
+  module.try_emplace("dwoFileCount", dwo_stats.dwo_file_count);
+  module.try_emplace("loadedDwoFileCount", dwo_stats.loaded_dwo_file_count);
+  module.try_emplace("dwoErrorCount", dwo_stats.dwo_error_count);
 
   if (!symbol_locator_time.map.empty()) {
     json::Object obj;
@@ -324,8 +325,7 @@ llvm::json::Value DebuggerStats::ReportStatistics(
   uint32_t num_modules_with_incomplete_types = 0;
   uint32_t num_stripped_modules = 0;
   uint32_t symtab_symbol_count = 0;
-  uint32_t total_loaded_dwo_file_count = 0;
-  uint32_t total_dwo_file_count = 0;
+  DWOStats total_dwo_stats;
   for (size_t image_idx = 0; image_idx < num_modules; ++image_idx) {
     Module *module = target != nullptr
                          ? target->GetImages().GetModuleAtIndex(image_idx).get()
@@ -357,10 +357,9 @@ llvm::json::Value DebuggerStats::ReportStatistics(
         for (const auto &symbol_module : symbol_modules.Modules())
           module_stat.symfile_modules.push_back((intptr_t)symbol_module.get());
       }
-      std::tie(module_stat.loaded_dwo_file_count, module_stat.dwo_file_count) =
-          sym_file->GetDwoFileCounts();
-      total_dwo_file_count += module_stat.dwo_file_count;
-      total_loaded_dwo_file_count += module_stat.loaded_dwo_file_count;
+      DWOStats current_dwo_stats = sym_file->GetDwoStats();
+      module_stat.dwo_stats += current_dwo_stats;
+      total_dwo_stats += current_dwo_stats;
       module_stat.debug_info_index_loaded_from_cache =
           sym_file->GetDebugInfoIndexWasLoadedFromCache();
       if (module_stat.debug_info_index_loaded_from_cache)
@@ -435,8 +434,9 @@ llvm::json::Value DebuggerStats::ReportStatistics(
       {"totalDebugInfoEnabled", num_debug_info_enabled_modules},
       {"totalSymbolTableStripped", num_stripped_modules},
       {"totalSymbolTableSymbolCount", symtab_symbol_count},
-      {"totalLoadedDwoFileCount", total_loaded_dwo_file_count},
-      {"totalDwoFileCount", total_dwo_file_count},
+      {"totalLoadedDwoFileCount", total_dwo_stats.loaded_dwo_file_count},
+      {"totalDwoFileCount", total_dwo_stats.dwo_file_count},
+      {"totalDwoErrorCount", total_dwo_stats.dwo_error_count},
   };
 
   if (include_targets) {

lldb/test/API/commands/statistics/basic/TestStats.py

@@ -1,6 +1,7 @@
 import json
 import os
 import re
+import shutil
 
 import lldb
 from lldbsuite.test.decorators import *
@@ -179,6 +180,7 @@ def test_default_no_run(self):
             "totalDebugInfoParseTime",
             "totalDwoFileCount",
             "totalLoadedDwoFileCount",
+            "totalDwoErrorCount",
         ]
         self.verify_keys(debug_stats, '"debug_stats"', debug_stat_keys, None)
         if self.getPlatform() != "windows":
@@ -291,6 +293,7 @@ def test_default_with_run(self):
             "totalDebugInfoParseTime",
             "totalDwoFileCount",
             "totalLoadedDwoFileCount",
+            "totalDwoErrorCount",
         ]
         self.verify_keys(debug_stats, '"debug_stats"', debug_stat_keys, None)
         stats = debug_stats["targets"][0]
@@ -331,6 +334,7 @@ def test_memory(self):
             "totalDebugInfoByteSize",
             "totalDwoFileCount",
             "totalLoadedDwoFileCount",
+            "totalDwoErrorCount",
         ]
         self.verify_keys(debug_stats, '"debug_stats"', debug_stat_keys, None)
 
@@ -385,6 +389,7 @@ def test_modules(self):
             "totalDebugInfoByteSize",
             "totalDwoFileCount",
             "totalLoadedDwoFileCount",
+            "totalDwoErrorCount",
         ]
         self.verify_keys(debug_stats, '"debug_stats"', debug_stat_keys, None)
         stats = debug_stats["targets"][0]
@@ -407,6 +412,7 @@ def test_modules(self):
             "symbolTableSavedToCache",
             "dwoFileCount",
             "loadedDwoFileCount",
+            "dwoErrorCount",
             "triple",
             "uuid",
         ]
@@ -497,6 +503,7 @@ def test_breakpoints(self):
             "totalDebugInfoByteSize",
             "totalDwoFileCount",
             "totalLoadedDwoFileCount",
+            "totalDwoErrorCount",
         ]
         self.verify_keys(debug_stats, '"debug_stats"', debug_stat_keys, None)
         target_stats = debug_stats["targets"][0]
@@ -655,6 +662,99 @@ def test_dwp_dwo_file_count(self):
         self.assertEqual(debug_stats["totalDwoFileCount"], 2)
         self.assertEqual(debug_stats["totalLoadedDwoFileCount"], 2)
 
+    @add_test_categories(["dwo"])
+    def test_dwo_missing_error_stats(self):
+        """
+        Test that DWO missing errors are reported correctly in statistics.
+        This test:
+        1) Builds a program with split DWARF (.dwo files)
+        2) Delete one of the two .dwo files
+        3) Verify that 1 DWO error is reported in statistics
+        """
+        da = {
+            "CXX_SOURCES": "dwo_error_main.cpp dwo_error_foo.cpp",
+            "EXE": self.getBuildArtifact("a.out"),
+        }
+        # -gsplit-dwarf creates separate .dwo files,
+        # Expected output: dwo_error_main.dwo (contains main) and dwo_error_foo.dwo (contains foo struct/function)
+        self.build(dictionary=da, debug_info="dwo")
+        exe = self.getBuildArtifact("a.out")
+
+        expected_dwo_files = [
+            self.getBuildArtifact("dwo_error_main.dwo"),
+            self.getBuildArtifact("dwo_error_foo.dwo"),
+        ]
+
+        # Verify expected files exist
+        for dwo_file in expected_dwo_files:
+            self.assertTrue(
+                os.path.exists(dwo_file),
+                f"Expected .dwo file does not exist: {dwo_file}",
+            )
+
+        # Remove one of .dwo files to trigger DWO load error
+        dwo_main_file = self.getBuildArtifact("dwo_error_main.dwo")
+        os.remove(dwo_main_file)
+
+        target = self.createTestTarget(file_path=exe)
+        debug_stats = self.get_stats()
+
+        # Check DWO load error statistics are reported
+        self.assertIn("totalDwoErrorCount", debug_stats)
+        self.assertEqual(debug_stats["totalDwoErrorCount"], 1)
+
+        # Since there's only one module, module stats should have the same count as total count
+        self.assertIn("dwoErrorCount", debug_stats["modules"][0])
+        self.assertEqual(debug_stats["modules"][0]["dwoErrorCount"], 1)
+
+    @add_test_categories(["dwo"])
+    def test_dwo_id_mismatch_error_stats(self):
+        """
+        Test that DWO ID mismatch errors are reported correctly in statistics.
+        This test:
+        1) Builds a program with split DWARF (.dwo files)
+        2) Replace one of the .dwo files with a mismatched one to cause a DWO ID mismatch error
+        3) Verifies that a DWO error is reported in statistics
+        """
+        da = {
+            "CXX_SOURCES": "dwo_error_main.cpp dwo_error_foo.cpp",
+            "EXE": self.getBuildArtifact("a.out"),
+        }
+        # -gsplit-dwarf creates separate .dwo files,
+        # Expected output: dwo_error_main.dwo (contains main) and dwo_error_foo.dwo (contains foo struct/function)
+        self.build(dictionary=da, debug_info="dwo")
+        exe = self.getBuildArtifact("a.out")
+
+        expected_dwo_files = [
+            self.getBuildArtifact("dwo_error_main.dwo"),
+            self.getBuildArtifact("dwo_error_foo.dwo"),
+        ]
+
+        # Verify expected files exist
+        for dwo_file in expected_dwo_files:
+            self.assertTrue(
+                os.path.exists(dwo_file),
+                f"Expected .dwo file does not exist: {dwo_file}",
+            )
+
+        # Replace one of the original .dwo file content with another one to trigger DWO ID mismatch error
+        dwo_foo_file = self.getBuildArtifact("dwo_error_foo.dwo")
+        dwo_main_file = self.getBuildArtifact("dwo_error_main.dwo")
+
+        shutil.copy(dwo_main_file, dwo_foo_file)
+
+        # Create a new target and get stats
+        target = self.createTestTarget(file_path=exe)
+        debug_stats = self.get_stats()
+
+        # Check that DWO load error statistics are reported
+        self.assertIn("totalDwoErrorCount", debug_stats)
+        self.assertEqual(debug_stats["totalDwoErrorCount"], 1)
+
+        # Since there's only one module, module stats should have the same count as total count
+        self.assertIn("dwoErrorCount", debug_stats["modules"][0])
+        self.assertEqual(debug_stats["modules"][0]["dwoErrorCount"], 1)
+
     @skipUnlessDarwin
     @no_debug_info_test
     def test_dsym_binary_has_symfile_in_stats(self):

lldb/test/API/commands/statistics/basic/dwo_error_foo.cpp

@@ -0,0 +1,10 @@
+struct foo {
+  int x;
+  bool y;
+};
+
+void dwo_error_foo() {
+  foo f;
+  f.x = 1;
+  f.y = true;
+}

lldb/test/API/commands/statistics/basic/dwo_error_main.cpp

@@ -0,0 +1,5 @@
+void dwo_error_foo();
+int main() {
+  dwo_error_foo();
+  return 0;
+}