[llvm-debuginfo-analyzer] Add support for parsing DWARF / CodeView SourceLanguage (#137223)

This pull request adds support for parsing the source language in both
DWARF and CodeView. Specifically,

- The `LVSourceLanguage` class is introduced to represent any supported
language by any of the debug info representations.

- Update `LVDWARFReader.cpp` and `LVCodeViewVisitor.cpp` to parse the
source language where it applies. Added a new `=Language` attribute;
`getAttributeLanguage()` is internally used to control whether this
information is being printed.

Author: jalopezg-git

llvm/docs/CommandGuide/llvm-debuginfo-analyzer.rst

@@ -134,12 +134,13 @@ toolchain name, binary file format, etc.
  The following attributes describe the most common information for a
  logical element. They help to identify the lexical scope level; the
  element visibility across modules (global, local); the toolchain name
- that produced the binary file.
+ and source language that produced the binary file.
 
  .. code-block:: text
 
    =global: Element referenced across Compile Units.
    =format: Object file format name.
+   =language: Source language name.
    =level: Lexical scope level (File=0, Compile Unit=1).
    =local: Element referenced only in the Compile Unit.
    =producer: Toolchain identification name.
@@ -231,6 +232,7 @@ toolchain name, binary file format, etc.
    =filename
    =files
    =format
+   =language
    =level
    =producer
    =publics

llvm/include/llvm/DebugInfo/LogicalView/Core/LVElement.h

@@ -15,6 +15,7 @@
 #define LLVM_DEBUGINFO_LOGICALVIEW_CORE_LVELEMENT_H
 
 #include "llvm/DebugInfo/LogicalView/Core/LVObject.h"
+#include "llvm/DebugInfo/LogicalView/Core/LVSourceLanguage.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/Compiler.h"
 #include "llvm/Support/MathExtras.h"
@@ -221,6 +222,9 @@ class LLVM_ABI LVElement : public LVObject {
   virtual StringRef getProducer() const { return StringRef(); }
   virtual void setProducer(StringRef ProducerName) {}
 
+  virtual LVSourceLanguage getSourceLanguage() const { return {}; }
+  virtual void setSourceLanguage(LVSourceLanguage SL) {}
+
   virtual bool isCompileUnit() const { return false; }
   virtual bool isRoot() const { return false; }
 

llvm/include/llvm/DebugInfo/LogicalView/Core/LVOptions.h

@@ -107,6 +107,7 @@ enum class LVAttributeKind {
   Generated,     // --attribute=generated
   Global,        // --attribute=global
   Inserted,      // --attribute=inserted
+  Language,      // --attribute=language
   Level,         // --attribute=level
   Linkage,       // --attribute=linkage
   Local,         // --attribute=local
@@ -338,6 +339,7 @@ class LVOptions {
   ATTRIBUTE_OPTION(Generated);
   ATTRIBUTE_OPTION(Global);
   ATTRIBUTE_OPTION(Inserted);
+  ATTRIBUTE_OPTION(Language);
   ATTRIBUTE_OPTION(Level);
   ATTRIBUTE_OPTION(Linkage);
   ATTRIBUTE_OPTION(Location);

llvm/include/llvm/DebugInfo/LogicalView/Core/LVScope.h

@@ -419,6 +419,9 @@ class LLVM_ABI LVScopeCompileUnit final : public LVScope {
   // Compilation directory name.
   size_t CompilationDirectoryIndex = 0;
 
+  // Source language.
+  LVSourceLanguage SourceLanguage{};
+
   // Used by the CodeView Reader.
   codeview::CPUType CompilationCPUType = codeview::CPUType::X64;
 
@@ -549,6 +552,9 @@ class LLVM_ABI LVScopeCompileUnit final : public LVScope {
     ProducerIndex = getStringPool().getIndex(ProducerName);
   }
 
+  LVSourceLanguage getSourceLanguage() const override { return SourceLanguage; }
+  void setSourceLanguage(LVSourceLanguage SL) override { SourceLanguage = SL; }
+
   void setCPUType(codeview::CPUType Type) { CompilationCPUType = Type; }
   codeview::CPUType getCPUType() { return CompilationCPUType; }
 

llvm/include/llvm/DebugInfo/LogicalView/Core/LVSourceLanguage.h

@@ -0,0 +1,69 @@
+//===-- LVSourceLanguage.h --------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines the LVSourceLanguage struct, a unified representation of
+// the source language used in a compile unit.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_DEBUGINFO_LOGICALVIEW_CORE_LVSOURCELANGUAGE_H
+#define LLVM_DEBUGINFO_LOGICALVIEW_CORE_LVSOURCELANGUAGE_H
+
+#include "llvm/ADT/StringRef.h"
+#include "llvm/BinaryFormat/Dwarf.h"
+#include "llvm/DebugInfo/CodeView/CodeView.h"
+
+namespace llvm {
+namespace logicalview {
+
+/// A source language supported by any of the debug info representations.
+struct LVSourceLanguage {
+  static constexpr unsigned TagDwarf = 0x00;
+  static constexpr unsigned TagCodeView = 0x01;
+
+  enum TaggedLanguage : uint32_t {
+    Invalid = -1U,
+
+  // DWARF
+#define HANDLE_DW_LANG(ID, NAME, LOWER_BOUND, VERSION, VENDOR)                 \
+  DW_LANG_##NAME = (TagDwarf << 16) | ID,
+#include "llvm/BinaryFormat/Dwarf.def"
+  // CodeView
+#define CV_LANGUAGE(NAME, ID) CV_LANG_##NAME = (TagCodeView << 16) | ID,
+#include "llvm/DebugInfo/CodeView/CodeViewLanguages.def"
+  };
+
+  LVSourceLanguage() = default;
+  LVSourceLanguage(llvm::dwarf::SourceLanguage SL)
+      : LVSourceLanguage(TagDwarf, SL) {}
+  LVSourceLanguage(llvm::codeview::SourceLanguage SL)
+      : LVSourceLanguage(TagCodeView, SL) {}
+  bool operator==(const LVSourceLanguage &SL) const {
+    return get() == SL.get();
+  }
+  bool operator==(const LVSourceLanguage::TaggedLanguage &TL) const {
+    return get() == TL;
+  }
+
+  bool isValid() const { return Language != Invalid; }
+  TaggedLanguage get() const { return Language; }
+  StringRef getName() const;
+
+private:
+  TaggedLanguage Language = Invalid;
+
+  LVSourceLanguage(unsigned Tag, unsigned Lang)
+      : Language(static_cast<TaggedLanguage>((Tag << 16) | Lang)) {}
+  unsigned getTag() const { return Language >> 16; }
+  unsigned getLang() const { return Language & 0xffff; }
+};
+
+} // end namespace logicalview
+} // end namespace llvm
+
+#endif // LLVM_DEBUGINFO_LOGICALVIEW_CORE_LVSOURCELANGUAGE_H

llvm/lib/DebugInfo/LogicalView/CMakeLists.txt

@@ -14,6 +14,7 @@ add_lv_impl_folder(Core
   Core/LVReader.cpp
   Core/LVScope.cpp
   Core/LVSort.cpp
+  Core/LVSourceLanguage.cpp
   Core/LVSupport.cpp
   Core/LVSymbol.cpp
   Core/LVType.cpp

llvm/lib/DebugInfo/LogicalView/Core/LVOptions.cpp

@@ -39,6 +39,7 @@ void LVOptions::resolveDependencies() {
     setAttributeFilename();
     setAttributeFiles();
     setAttributeFormat();
+    setAttributeLanguage();
     setAttributeLevel();
     setAttributeProducer();
     setAttributePublics();

llvm/lib/DebugInfo/LogicalView/Core/LVScope.cpp

@@ -1717,11 +1717,19 @@ void LVScopeCompileUnit::print(raw_ostream &OS, bool Full) const {
 
 void LVScopeCompileUnit::printExtra(raw_ostream &OS, bool Full) const {
   OS << formattedKind(kind()) << " '" << getName() << "'\n";
-  if (options().getPrintFormatting() && options().getAttributeProducer())
-    printAttributes(OS, Full, "{Producer} ",
-                    const_cast<LVScopeCompileUnit *>(this), getProducer(),
-                    /*UseQuotes=*/true,
-                    /*PrintRef=*/false);
+  if (options().getPrintFormatting()) {
+    if (options().getAttributeProducer())
+      printAttributes(OS, Full, "{Producer} ",
+                      const_cast<LVScopeCompileUnit *>(this), getProducer(),
+                      /*UseQuotes=*/true,
+                      /*PrintRef=*/false);
+    if (options().getAttributeLanguage())
+      if (auto SL = getSourceLanguage(); SL.isValid())
+        printAttributes(OS, Full, "{Language} ",
+                        const_cast<LVScopeCompileUnit *>(this), SL.getName(),
+                        /*UseQuotes=*/true,
+                        /*PrintRef=*/false);
+  }
 
   // Reset file index, to allow its children to print the correct filename.
   options().resetFilenameIndex();

llvm/lib/DebugInfo/LogicalView/Core/LVSourceLanguage.cpp

@@ -0,0 +1,33 @@
+//===-- LVSourceLanguage.cpp ----------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file implements LVSourceLanguage.
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/DebugInfo/LogicalView/Core/LVSourceLanguage.h"
+#include "llvm/DebugInfo/CodeView/EnumTables.h"
+#include "llvm/Support/ScopedPrinter.h"
+
+using namespace llvm;
+using namespace llvm::logicalview;
+
+StringRef LVSourceLanguage::getName() const {
+  if (!isValid())
+    return {};
+  switch (getTag()) {
+  case LVSourceLanguage::TagDwarf:
+    return llvm::dwarf::LanguageString(getLang());
+  case LVSourceLanguage::TagCodeView: {
+    static auto LangNames = llvm::codeview::getSourceLanguageNames();
+    return LangNames[getLang()].Name;
+  }
+  default:
+    llvm_unreachable("Unsupported language");
+  }
+}

llvm/lib/DebugInfo/LogicalView/Readers/LVCodeViewVisitor.cpp

@@ -949,6 +949,9 @@ Error LVSymbolVisitor::visitKnownRecord(CVSymbol &Record,
     Scope->setName(CurrentObjectName);
     if (options().getAttributeProducer())
       Scope->setProducer(Compile2.Version);
+    if (options().getAttributeLanguage())
+      Scope->setSourceLanguage(LVSourceLanguage{
+          static_cast<llvm::codeview::SourceLanguage>(Compile2.getLanguage())});
     getReader().isSystemEntry(Scope, CurrentObjectName);
 
     // The line records in CodeView are recorded per Module ID. Update
@@ -994,6 +997,9 @@ Error LVSymbolVisitor::visitKnownRecord(CVSymbol &Record,
     Scope->setName(CurrentObjectName);
     if (options().getAttributeProducer())
       Scope->setProducer(Compile3.Version);
+    if (options().getAttributeLanguage())
+      Scope->setSourceLanguage(LVSourceLanguage{
+          static_cast<llvm::codeview::SourceLanguage>(Compile3.getLanguage())});
     getReader().isSystemEntry(Scope, CurrentObjectName);
 
     // The line records in CodeView are recorded per Module ID. Update

llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp

@@ -172,6 +172,11 @@ void LVDWARFReader::processOneAttribute(const DWARFDie &Die,
     if (options().getAttributeProducer())
       CurrentElement->setProducer(dwarf::toStringRef(FormValue));
     break;
+  case dwarf::DW_AT_language:
+    if (options().getAttributeLanguage())
+      CurrentElement->setSourceLanguage(LVSourceLanguage{
+          static_cast<llvm::dwarf::SourceLanguage>(GetAsUnsignedConstant())});
+    break;
   case dwarf::DW_AT_upper_bound:
     CurrentElement->setUpperBound(GetBoundValue(FormValue));
     break;

llvm/test/tools/llvm-debuginfo-analyzer/COFF/02-coff-logical-lines.test

@@ -15,7 +15,7 @@
 ; The logical views shows the intermixed lines and assembler instructions,
 ; allowing to compare the code generated by the different toolchains.
 
-; RUN: llvm-debuginfo-analyzer --attribute=level,format,producer \
+; RUN: llvm-debuginfo-analyzer --attribute=language,level,format,producer \
 ; RUN:                         --print=lines,instructions \
 ; RUN:                         %p/Inputs/hello-world-codeview-clang.o \
 ; RUN:                         %p/Inputs/hello-world-codeview-msvc.o 2>&1 | \
@@ -26,6 +26,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'hello-world.cpp'
 ; ONE-NEXT: [002]               {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Function} extern not_inlined 'main' -> 'int'
 ; ONE-NEXT: [003]     4           {Line}
 ; ONE-NEXT: [003]                 {Code} 'subq	$0x28, %rsp'
@@ -43,6 +44,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'hello-world.cpp'
 ; ONE-NEXT: [002]               {Producer} 'Microsoft (R) Optimizing Compiler'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Function} extern not_inlined 'main' -> 'int'
 ; ONE-NEXT: [003]     4           {Line}
 ; ONE-NEXT: [003]                 {Code} 'subq	$0x28, %rsp'

llvm/test/tools/llvm-debuginfo-analyzer/COFF/03-coff-incorrect-lexical-scope-typedef.test

@@ -30,7 +30,7 @@
 ; emits both typedefs at the same lexical scope (3), which is wrong.
 ; GCC and MSVC emit correct lexical scope for both typedefs.
 
-; RUN: llvm-debuginfo-analyzer --attribute=level,format,producer \
+; RUN: llvm-debuginfo-analyzer --attribute=language,level,format,producer \
 ; RUN:                         --output-sort=kind \
 ; RUN:                         --print=symbols,types,lines \
 ; RUN:                         %p/Inputs/pr-44884-codeview-clang.o \
@@ -42,6 +42,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-44884.cpp'
 ; ONE-NEXT: [002]               {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Function} extern not_inlined 'bar' -> 'int'
 ; ONE-NEXT: [003]                 {Parameter} 'Input' -> 'float'
 ; ONE-NEXT: [003]     1           {Line}
@@ -63,6 +64,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-44884.cpp'
 ; ONE-NEXT: [002]               {Producer} 'Microsoft (R) Optimizing Compiler'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Function} extern not_inlined 'bar' -> 'int'
 ; ONE-NEXT: [003]                 {Variable} 'Input' -> 'float'
 ; ONE-NEXT: [003]     1           {Line}

llvm/test/tools/llvm-debuginfo-analyzer/COFF/04-coff-missing-nested-enumerators.test

@@ -25,7 +25,7 @@
 ; references to the enumerators 'RED' and 'BLUE'. The CodeView generated
 ; by GCC and MSVC, does include such references.
 
-; RUN: llvm-debuginfo-analyzer --attribute=level,format,producer,size \
+; RUN: llvm-debuginfo-analyzer --attribute=language,level,format,producer,size \
 ; RUN:                         --output-sort=name \
 ; RUN:                         --print=symbols,types \
 ; RUN:                         %p/Inputs/pr-46466-codeview-clang.o \
@@ -37,6 +37,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-46466.cpp'
 ; ONE-NEXT: [002]               {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Variable} extern 'S' -> 'Struct'
 ; ONE-NEXT: [002]     1         {Struct} 'Struct' [Size = 1]
 ; ONE-NEXT: [003]                 {Member} public 'U' -> 'Union'
@@ -50,6 +51,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-46466.cpp'
 ; ONE-NEXT: [002]               {Producer} 'Microsoft (R) Optimizing Compiler'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Variable} extern 'S' -> 'Struct'
 ; ONE-NEXT: [002]     1         {Struct} 'Struct' [Size = 1]
 ; ONE-NEXT: [003]                 {Member} public 'U' -> 'Union'

llvm/test/tools/llvm-debuginfo-analyzer/COFF/05-coff-incorrect-lexical-scope-variable.test

@@ -31,7 +31,7 @@
 ; The CodeView generated by MSVC, show those variables at the correct
 ; lexical scope: '3' and '4' respectively.
 
-; RUN: llvm-debuginfo-analyzer --attribute=level,format,producer \
+; RUN: llvm-debuginfo-analyzer --attribute=language,level,format,producer \
 ; RUN:                         --output-sort=name \
 ; RUN:                         --print=symbols \
 ; RUN:                         %p/Inputs/pr-43860-codeview-clang.o \
@@ -43,6 +43,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-43860.cpp'
 ; ONE-NEXT: [002]               {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]     2         {Function} inlined 'InlineFunction' -> 'int'
 ; ONE-NEXT: [003]                 {Parameter} '' -> 'int'
 ; ONE-NEXT: [002]               {Function} extern not_inlined 'test' -> 'int'
@@ -59,6 +60,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'pr-43860.cpp'
 ; ONE-NEXT: [002]               {Producer} 'Microsoft (R) Optimizing Compiler'
+; ONE-NEXT: [002]               {Language} 'Cpp'
 ; ONE-NEXT: [002]               {Function} extern declared_inlined 'InlineFunction' -> 'int'
 ; ONE-NEXT: [003]                 {Block}
 ; ONE-NEXT: [004]                   {Variable} 'Var_2' -> 'int'

llvm/test/tools/llvm-debuginfo-analyzer/COFF/06-coff-full-logical-view.test

@@ -28,6 +28,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [0x0000000000][001]              {CompileUnit} 'test.cpp'
 ; ONE-NEXT: [0x0000000000][002]                {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [0x0000000000][002]                {Language} 'Cpp'
 ; ONE-NEXT:                                    {Directory} 'test.cpp'
 ; ONE-NEXT:                                    {Directory} 'x:/tests/input'
 ; ONE-NEXT:                                    {File} 'general'

llvm/test/tools/llvm-debuginfo-analyzer/DWARF/02-dwarf-logical-lines.test

@@ -15,7 +15,7 @@
 ; The logical views shows the intermixed lines and assembler instructions,
 ; allowing to compare the code generated by the different toolchains.
 
-; RUN: llvm-debuginfo-analyzer --attribute=level,format,producer \
+; RUN: llvm-debuginfo-analyzer --attribute=language,level,format,producer \
 ; RUN:                         --print=lines,instructions \
 ; RUN:                         %p/Inputs/hello-world-dwarf-clang.o \
 ; RUN:                         %p/Inputs/hello-world-dwarf-gcc.o 2>&1 | \
@@ -26,6 +26,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'hello-world.cpp'
 ; ONE-NEXT: [002]               {Producer} 'clang version 15.0.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'DW_LANG_C_plus_plus_14'
 ; ONE-NEXT: [002]     3         {Function} extern not_inlined 'main' -> 'int'
 ; ONE-NEXT: [003]     4           {Line}
 ; ONE-NEXT: [003]                 {Code} 'pushq	%rbp'
@@ -48,6 +49,7 @@
 ; ONE-EMPTY:
 ; ONE-NEXT: [001]             {CompileUnit} 'hello-world.cpp'
 ; ONE-NEXT: [002]               {Producer} 'GNU C++14 10.3.0 {{.*}}'
+; ONE-NEXT: [002]               {Language} 'DW_LANG_C_plus_plus'
 ; ONE-NEXT: [002]     3         {Function} extern not_inlined 'main' -> 'int'
 ; ONE-NEXT: [003]     4           {Line}
 ; ONE-NEXT: [003]                 {Code} 'endbr64'