Automerge: [lldb] Add filter option to AST dump command (#142164)

Depends on llvm/llvm-project#142163

This patch makes the `-ast-dump-filter` Clang option available to the
`target modules dump ast` command. This allows us to selectively dump
parts of the AST by name.

The AST can quickly grow way too large to skim on the console. This will
aid in debugging AST related issues.

Example:
```
(lldb) target modules dump ast --filter func
Dumping clang ast for 48 modules.
Dumping func:
FunctionDecl 0xc4b785008 <<invalid sloc>> <invalid sloc> func 'void (int)' extern
|-ParmVarDecl 0xc4b7853d8 <<invalid sloc>> <invalid sloc> x 'int'
`-AsmLabelAttr 0xc4b785358 <<invalid sloc>> Implicit "_Z4funcIiEvT_"

Dumping func<int>:
FunctionDecl 0xc4b7850b8 <<invalid sloc>> <invalid sloc> func<int> 'void (int)' implicit_instantiation extern
|-TemplateArgument type 'int'
| `-BuiltinType 0xc4b85b110 'int'
`-ParmVarDecl 0xc4b7853d8 <<invalid sloc>> <invalid sloc> x 'int'
```

The majority of this patch is adjust the `Dump` API. The main change in
behaviour is in `TypeSystemClang::Dump`, where we now use the
`ASTPrinter` for dumping the `TranslationUnitDecl`. This is where the
`-ast-dump-filter` functionality lives in Clang.

Author: Michael137

lldb/include/lldb/Symbol/SymbolFile.h

@@ -296,7 +296,7 @@ class SymbolFile : public PluginInterface {
                        lldb::SymbolContextItem resolve_scope,
                        SymbolContextList &sc_list);
 
-  virtual void DumpClangAST(Stream &s) {}
+  virtual void DumpClangAST(Stream &s, llvm::StringRef filter) {}
   virtual void FindGlobalVariables(ConstString name,
                                    const CompilerDeclContext &parent_decl_ctx,
                                    uint32_t max_matches,

lldb/include/lldb/Symbol/SymbolFileOnDemand.h

@@ -127,7 +127,7 @@ class SymbolFileOnDemand : public lldb_private::SymbolFile {
       lldb_private::SymbolContextList &sc_list) override;
 
   void Dump(lldb_private::Stream &s) override;
-  void DumpClangAST(lldb_private::Stream &s) override;
+  void DumpClangAST(lldb_private::Stream &s, llvm::StringRef filter) override;
 
   void
   FindGlobalVariables(lldb_private::ConstString name,

lldb/include/lldb/Symbol/TypeSystem.h

@@ -443,7 +443,11 @@ class TypeSystem : public PluginInterface,
   /// given stream.
   ///
   /// This should not modify the state of the TypeSystem if possible.
-  virtual void Dump(llvm::raw_ostream &output) = 0;
+  ///
+  /// \param[out] output Stream to dup the AST into.
+  /// \param[in] filter If empty, dump whole AST. If non-empty, will only
+  /// dump decls whose names contain \c filter.
+  virtual void Dump(llvm::raw_ostream &output, llvm::StringRef filter) = 0;
 
   /// This is used by swift.
   virtual bool IsRuntimeGeneratedType(lldb::opaque_compiler_type_t type) = 0;

lldb/source/Commands/CommandObjectTarget.cpp

@@ -2235,11 +2235,23 @@ class CommandObjectTargetModulesDumpClangAST
       : CommandObjectTargetModulesModuleAutoComplete(
             interpreter, "target modules dump ast",
             "Dump the clang ast for a given module's symbol file.",
-            //"target modules dump ast [<file1> ...]")
-            nullptr, eCommandRequiresTarget) {}
+            "target modules dump ast [--filter <name>] [<file1> ...]",
+            eCommandRequiresTarget),
+        m_filter(LLDB_OPT_SET_1, false, "filter", 'f', 0, eArgTypeName,
+                 "Dump only the decls whose names contain the specified filter "
+                 "string.",
+                 /*default_value=*/"") {
+    m_option_group.Append(&m_filter, LLDB_OPT_SET_ALL, LLDB_OPT_SET_1);
+    m_option_group.Finalize();
+  }
+
+  Options *GetOptions() override { return &m_option_group; }
 
   ~CommandObjectTargetModulesDumpClangAST() override = default;
 
+  OptionGroupOptions m_option_group;
+  OptionGroupString m_filter;
+
 protected:
   void DoExecute(Args &command, CommandReturnObject &result) override {
     Target &target = GetTarget();
@@ -2251,6 +2263,8 @@ class CommandObjectTargetModulesDumpClangAST
       return;
     }
 
+    llvm::StringRef filter = m_filter.GetOptionValue().GetCurrentValueAsRef();
+
     if (command.GetArgumentCount() == 0) {
       // Dump all ASTs for all modules images
       result.GetOutputStream().Format("Dumping clang ast for {0} modules.\n",
@@ -2259,7 +2273,7 @@ class CommandObjectTargetModulesDumpClangAST
         if (INTERRUPT_REQUESTED(GetDebugger(), "Interrupted dumping clang ast"))
           break;
         if (SymbolFile *sf = module_sp->GetSymbolFile())
-          sf->DumpClangAST(result.GetOutputStream());
+          sf->DumpClangAST(result.GetOutputStream(), filter);
       }
       result.SetStatus(eReturnStatusSuccessFinishResult);
       return;
@@ -2288,7 +2302,7 @@ class CommandObjectTargetModulesDumpClangAST
 
         Module *m = module_list.GetModulePointerAtIndex(i);
         if (SymbolFile *sf = m->GetSymbolFile())
-          sf->DumpClangAST(result.GetOutputStream());
+          sf->DumpClangAST(result.GetOutputStream(), filter);
       }
     }
     result.SetStatus(eReturnStatusSuccessFinishResult);
@@ -5272,7 +5286,7 @@ class CommandObjectTargetDumpTypesystem : public CommandObjectParsed {
     // Go over every scratch TypeSystem and dump to the command output.
     for (lldb::TypeSystemSP ts : GetTarget().GetScratchTypeSystems())
       if (ts)
-        ts->Dump(result.GetOutputStream().AsRawOstream());
+        ts->Dump(result.GetOutputStream().AsRawOstream(), "");
 
     result.SetStatus(eReturnStatusSuccessFinishResult);
   }

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

@@ -4127,15 +4127,15 @@ void SymbolFileDWARF::Dump(lldb_private::Stream &s) {
   m_index->Dump(s);
 }
 
-void SymbolFileDWARF::DumpClangAST(Stream &s) {
+void SymbolFileDWARF::DumpClangAST(Stream &s, llvm::StringRef filter) {
   auto ts_or_err = GetTypeSystemForLanguage(eLanguageTypeC_plus_plus);
   if (!ts_or_err)
     return;
   auto ts = *ts_or_err;
   TypeSystemClang *clang = llvm::dyn_cast_or_null<TypeSystemClang>(ts.get());
   if (!clang)
     return;
-  clang->Dump(s.AsRawOstream());
+  clang->Dump(s.AsRawOstream(), filter);
 }
 
 bool SymbolFileDWARF::GetSeparateDebugInfo(StructuredData::Dictionary &d,

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

@@ -276,7 +276,7 @@ class SymbolFileDWARF : public SymbolFileCommon {
 
   void Dump(Stream &s) override;
 
-  void DumpClangAST(Stream &s) override;
+  void DumpClangAST(Stream &s, llvm::StringRef filter) override;
 
   /// List separate dwo files.
   bool GetSeparateDebugInfo(StructuredData::Dictionary &d,

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

@@ -1267,9 +1267,9 @@ CompilerDeclContext SymbolFileDWARFDebugMap::FindNamespace(
   return matching_namespace;
 }
 
-void SymbolFileDWARFDebugMap::DumpClangAST(Stream &s) {
-  ForEachSymbolFile("Dumping clang AST", [&s](SymbolFileDWARF &oso_dwarf) {
-    oso_dwarf.DumpClangAST(s);
+void SymbolFileDWARFDebugMap::DumpClangAST(Stream &s, llvm::StringRef filter) {
+  ForEachSymbolFile("Dumping clang AST", [&](SymbolFileDWARF &oso_dwarf) {
+    oso_dwarf.DumpClangAST(s, filter);
     // The underlying assumption is that DumpClangAST(...) will obtain the
     // AST from the underlying TypeSystem and therefore we only need to do
     // this once and can stop after the first iteration hence we return true.

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

@@ -129,7 +129,7 @@ class SymbolFileDWARFDebugMap : public SymbolFileCommon {
   std::vector<std::unique_ptr<CallEdge>>
   ParseCallEdgesInFunction(UserID func_id) override;
 
-  void DumpClangAST(Stream &s) override;
+  void DumpClangAST(Stream &s, llvm::StringRef filter) override;
 
   /// List separate oso files.
   bool GetSeparateDebugInfo(StructuredData::Dictionary &d,

lldb/source/Plugins/SymbolFile/NativePDB/PdbAstBuilder.cpp

@@ -1449,6 +1449,6 @@ PdbAstBuilder::FromCompilerDeclContext(CompilerDeclContext context) {
   return static_cast<clang::DeclContext *>(context.GetOpaqueDeclContext());
 }
 
-void PdbAstBuilder::Dump(Stream &stream) {
-  m_clang.Dump(stream.AsRawOstream());
+void PdbAstBuilder::Dump(Stream &stream, llvm::StringRef filter) {
+  m_clang.Dump(stream.AsRawOstream(), filter);
 }

lldb/source/Plugins/SymbolFile/NativePDB/PdbAstBuilder.h

@@ -87,7 +87,7 @@ class PdbAstBuilder {
   TypeSystemClang &clang() { return m_clang; }
   ClangASTImporter &GetClangASTImporter() { return m_importer; }
 
-  void Dump(Stream &stream);
+  void Dump(Stream &stream, llvm::StringRef filter);
 
 private:
   clang::Decl *TryGetDecl(PdbSymUid uid) const;

lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp

@@ -1630,15 +1630,15 @@ size_t SymbolFileNativePDB::ParseSymbolArrayInScope(
   return count;
 }
 
-void SymbolFileNativePDB::DumpClangAST(Stream &s) {
+void SymbolFileNativePDB::DumpClangAST(Stream &s, llvm::StringRef filter) {
   auto ts_or_err = GetTypeSystemForLanguage(eLanguageTypeC_plus_plus);
   if (!ts_or_err)
     return;
   auto ts = *ts_or_err;
   TypeSystemClang *clang = llvm::dyn_cast_or_null<TypeSystemClang>(ts.get());
   if (!clang)
     return;
-  clang->GetNativePDBParser()->Dump(s);
+  clang->GetNativePDBParser()->Dump(s, filter);
 }
 
 void SymbolFileNativePDB::FindGlobalVariables(

lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h

@@ -157,7 +157,7 @@ class SymbolFileNativePDB : public SymbolFileCommon {
 
   PdbIndex &GetIndex() { return *m_index; };
 
-  void DumpClangAST(Stream &s) override;
+  void DumpClangAST(Stream &s, llvm::StringRef filter) override;
 
   std::optional<llvm::codeview::TypeIndex>
   GetParentType(llvm::codeview::TypeIndex ti);

lldb/source/Plugins/SymbolFile/PDB/SymbolFilePDB.cpp

@@ -1431,7 +1431,7 @@ void SymbolFilePDB::AddSymbols(lldb_private::Symtab &symtab) {
   symtab.Finalize();
 }
 
-void SymbolFilePDB::DumpClangAST(Stream &s) {
+void SymbolFilePDB::DumpClangAST(Stream &s, llvm::StringRef filter) {
   auto type_system_or_err =
       GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
   if (auto err = type_system_or_err.takeError()) {
@@ -1445,7 +1445,7 @@ void SymbolFilePDB::DumpClangAST(Stream &s) {
       llvm::dyn_cast_or_null<TypeSystemClang>(ts.get());
   if (!clang_type_system)
     return;
-  clang_type_system->Dump(s.AsRawOstream());
+  clang_type_system->Dump(s.AsRawOstream(), filter);
 }
 
 void SymbolFilePDB::FindTypesByRegex(

lldb/source/Plugins/SymbolFile/PDB/SymbolFilePDB.h

@@ -157,7 +157,7 @@ class SymbolFilePDB : public lldb_private::SymbolFileCommon {
 
   const llvm::pdb::IPDBSession &GetPDBSession() const;
 
-  void DumpClangAST(lldb_private::Stream &s) override;
+  void DumpClangAST(lldb_private::Stream &s, llvm::StringRef filter) override;
 
 private:
   struct SecContribInfo {

lldb/source/Plugins/TypeSystem/Clang/TypeSystemClang.cpp

@@ -10,6 +10,7 @@
 
 #include "clang/AST/DeclBase.h"
 #include "clang/AST/ExprCXX.h"
+#include "clang/Frontend/ASTConsumers.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/FormatAdapters.h"
 #include "llvm/Support/FormatVariadic.h"
@@ -8499,8 +8500,16 @@ TypeSystemClang::dump(lldb::opaque_compiler_type_t type) const {
 }
 #endif
 
-void TypeSystemClang::Dump(llvm::raw_ostream &output) {
-  GetTranslationUnitDecl()->dump(output);
+void TypeSystemClang::Dump(llvm::raw_ostream &output, llvm::StringRef filter) {
+  auto consumer =
+      clang::CreateASTDumper(output, filter,
+                             /*DumpDecls=*/true,
+                             /*Deserialize=*/false,
+                             /*DumpLookups=*/false,
+                             /*DumpDeclTypes=*/false, clang::ADOF_Default);
+  assert(consumer);
+  assert(m_ast_up);
+  consumer->HandleTranslationUnit(*m_ast_up);
 }
 
 void TypeSystemClang::DumpFromSymbolFile(Stream &s,
@@ -9625,10 +9634,11 @@ GetNameForIsolatedASTKind(ScratchTypeSystemClang::IsolatedASTKind kind) {
   llvm_unreachable("Unimplemented IsolatedASTKind?");
 }
 
-void ScratchTypeSystemClang::Dump(llvm::raw_ostream &output) {
+void ScratchTypeSystemClang::Dump(llvm::raw_ostream &output,
+                                  llvm::StringRef filter) {
   // First dump the main scratch AST.
   output << "State of scratch Clang type system:\n";
-  TypeSystemClang::Dump(output);
+  TypeSystemClang::Dump(output, filter);
 
   // Now sort the isolated sub-ASTs.
   typedef std::pair<IsolatedASTKey, TypeSystem *> KeyAndTS;
@@ -9643,7 +9653,7 @@ void ScratchTypeSystemClang::Dump(llvm::raw_ostream &output) {
         static_cast<ScratchTypeSystemClang::IsolatedASTKind>(a.first);
     output << "State of scratch Clang type subsystem "
            << GetNameForIsolatedASTKind(kind) << ":\n";
-    a.second->Dump(output);
+    a.second->Dump(output, filter);
   }
 }
 

lldb/source/Plugins/TypeSystem/Clang/TypeSystemClang.h

@@ -1074,7 +1074,7 @@ class TypeSystemClang : public TypeSystem {
 #endif
 
   /// \see lldb_private::TypeSystem::Dump
-  void Dump(llvm::raw_ostream &output) override;
+  void Dump(llvm::raw_ostream &output, llvm::StringRef filter) override;
 
   /// Dump clang AST types from the symbol file.
   ///
@@ -1318,7 +1318,7 @@ class ScratchTypeSystemClang : public TypeSystemClang {
   }
 
   /// \see lldb_private::TypeSystem::Dump
-  void Dump(llvm::raw_ostream &output) override;
+  void Dump(llvm::raw_ostream &output, llvm::StringRef filter) override;
 
   UserExpression *GetUserExpression(llvm::StringRef expr,
                                     llvm::StringRef prefix,

lldb/source/Symbol/SymbolFileOnDemand.cpp

@@ -305,13 +305,14 @@ void SymbolFileOnDemand::Dump(lldb_private::Stream &s) {
   return m_sym_file_impl->Dump(s);
 }
 
-void SymbolFileOnDemand::DumpClangAST(lldb_private::Stream &s) {
+void SymbolFileOnDemand::DumpClangAST(lldb_private::Stream &s,
+                                      llvm::StringRef filter) {
   if (!m_debug_info_enabled) {
     LLDB_LOG(GetLog(), "[{0}] {1} is skipped", GetSymbolFileName(),
              __FUNCTION__);
     return;
   }
-  return m_sym_file_impl->DumpClangAST(s);
+  return m_sym_file_impl->DumpClangAST(s, filter);
 }
 
 void SymbolFileOnDemand::FindGlobalVariables(const RegularExpression &regex,

lldb/test/Shell/Commands/command-image-dump-ast.test

@@ -0,0 +1,70 @@
+# Test `image dump ast` command.
+
+# RUN: split-file %s %t
+# RUN: %clang_host -g -gdwarf %t/main.cpp -o %t.out
+# RUN: %lldb -x -b -s %t/commands.input %t.out -o exit 2>&1 \
+# RUN:       | FileCheck %s
+
+#--- main.cpp
+
+void A() {}
+void A1() {}
+void BA1() {}
+void AB() {}
+
+int main() {
+  A();
+  A1();
+  BA1();
+  AB();
+}
+
+#--- commands.input
+
+break set -n main
+run
+expr A(); A1(); BA1(); AB()
+
+image dump ast
+
+# CHECK: image dump ast
+# CHECK-DAG: FunctionDecl {{.*}} main
+# CHECK-DAG: FunctionDecl {{.*}} A
+# CHECK-DAG: FunctionDecl {{.*}} A1
+# CHECK-DAG: FunctionDecl {{.*}} BA1
+# CHECK-DAG: FunctionDecl {{.*}} AB
+
+image dump ast --filter A
+
+# CHECK: image dump ast --filter A
+# CHECK: Dumping A
+# CHECK-NOT: FunctionDecl {{.*}} main
+# CHECK-DAG: FunctionDecl {{.*}} A1
+# CHECK-DAG: FunctionDecl {{.*}} BA1
+# CHECK-DAG: FunctionDecl {{.*}} AB
+
+image dump ast --filter A1
+
+# CHECK: image dump ast --filter A1
+# CHECK: Dumping A
+# CHECK-NOT: FunctionDecl {{.*}} main
+# CHECK-NOT: FunctionDecl {{.*}} AB
+# CHECK-DAG: FunctionDecl {{.*}} A1
+# CHECK-DAG: FunctionDecl {{.*}} BA1
+
+image dump ast --filter ""
+
+# CHECK: image dump ast --filter ""
+# CHECK-DAG: FunctionDecl {{.*}} main
+# CHECK-DAG: FunctionDecl {{.*}} AB
+# CHECK-DAG: FunctionDecl {{.*}} A1
+# CHECK-DAG: FunctionDecl {{.*}} BA1
+
+image dump ast -f AB
+
+# CHECK: image dump ast -f AB
+# CHECK: Dumping AB
+# CHECK-NOT: FunctionDecl {{.*}} main
+# CHECK-NOT: FunctionDecl {{.*}} A1
+# CHECK-NOT: FunctionDecl {{.*}} BA1
+# CHECK: FunctionDecl {{.*}} AB