[llvm][DebugInfo] Support for versioned DISourceLanguageName

Author: Michael137

llvm/include/llvm/IR/DebugInfoMetadata.h

@@ -88,6 +88,10 @@ class DISourceLanguageName {
 
   /// Returns a versioned or unversioned language name.
   uint16_t getName() const { return Name; }
+  uint32_t getVersion() const {
+    assert(hasVersionedName() && "Tried getting version on unversioned name");
+    return *Version;
+  }
 
   // Transitional API for cases where we do not yet support
   // versioned source language names. Use \ref getName instead.

llvm/lib/AsmParser/LLParser.cpp

@@ -5837,8 +5837,9 @@ bool LLParser::parseDICompileUnit(MDNode *&Result, bool IsDistinct) {
     return tokError("missing 'distinct', required for !DICompileUnit");
 
 #define VISIT_MD_FIELDS(OPTIONAL, REQUIRED)                                    \
-  REQUIRED(language, DwarfLangField, );                                        \
   REQUIRED(file, MDField, (/* AllowNull */ false));                            \
+  OPTIONAL(language, DwarfLangField, );                                        \
+  OPTIONAL(sourceLanguageName, MDUnsignedField, );                             \
   OPTIONAL(producer, MDStringField, );                                         \
   OPTIONAL(isOptimized, MDBoolField, );                                        \
   OPTIONAL(flags, MDStringField, );                                            \
@@ -5860,12 +5861,19 @@ bool LLParser::parseDICompileUnit(MDNode *&Result, bool IsDistinct) {
   PARSE_MD_FIELDS();
 #undef VISIT_MD_FIELDS
 
+  if (!language.Seen && !sourceLanguageName.Seen)
+    return tokError("missing one of 'language' or 'sourceLanguageName', "
+                    "required for !DICompileUnit");
+
   Result = DICompileUnit::getDistinct(
-      Context, DISourceLanguageName(language.Val), file.Val, producer.Val,
-      isOptimized.Val, flags.Val, runtimeVersion.Val, splitDebugFilename.Val,
-      emissionKind.Val, enums.Val, retainedTypes.Val, globals.Val, imports.Val,
-      macros.Val, dwoId.Val, splitDebugInlining.Val, debugInfoForProfiling.Val,
-      nameTableKind.Val, rangesBaseAddress.Val, sysroot.Val, sdk.Val);
+      Context,
+      language.Seen ? DISourceLanguageName(language.Val)
+                    : DISourceLanguageName(sourceLanguageName.Val, 0),
+      file.Val, producer.Val, isOptimized.Val, flags.Val, runtimeVersion.Val,
+      splitDebugFilename.Val, emissionKind.Val, enums.Val, retainedTypes.Val,
+      globals.Val, imports.Val, macros.Val, dwoId.Val, splitDebugInlining.Val,
+      debugInfoForProfiling.Val, nameTableKind.Val, rangesBaseAddress.Val,
+      sysroot.Val, sdk.Val);
   return false;
 }
 

llvm/lib/Bitcode/Reader/MetadataLoader.cpp

@@ -1867,12 +1867,18 @@ Error MetadataLoader::MetadataLoaderImpl::parseOneMetadata(
     // distinct.  It's always distinct.
     IsDistinct = true;
 
+    const auto LangVersionMask = (uint64_t(1) << 63);
+    const bool HasVersionedLanguage = Record[1] & LangVersionMask;
+
     auto *CU = DICompileUnit::getDistinct(
-        Context, DISourceLanguageName(Record[1]), getMDOrNull(Record[2]),
-        getMDString(Record[3]), Record[4], getMDString(Record[5]), Record[6],
-        getMDString(Record[7]), Record[8], getMDOrNull(Record[9]),
-        getMDOrNull(Record[10]), getMDOrNull(Record[12]),
-        getMDOrNull(Record[13]),
+        Context,
+        HasVersionedLanguage
+            ? DISourceLanguageName(Record[1] & ~LangVersionMask, 0)
+            : DISourceLanguageName(Record[1]),
+        getMDOrNull(Record[2]), getMDString(Record[3]), Record[4],
+        getMDString(Record[5]), Record[6], getMDString(Record[7]), Record[8],
+        getMDOrNull(Record[9]), getMDOrNull(Record[10]),
+        getMDOrNull(Record[12]), getMDOrNull(Record[13]),
         Record.size() <= 15 ? nullptr : getMDOrNull(Record[15]),
         Record.size() <= 14 ? 0 : Record[14],
         Record.size() <= 16 ? true : Record[16],

llvm/lib/Bitcode/Writer/BitcodeWriter.cpp

@@ -2106,7 +2106,13 @@ void ModuleBitcodeWriter::writeDICompileUnit(const DICompileUnit *N,
   assert(N->isDistinct() && "Expected distinct compile units");
   Record.push_back(/* IsDistinct */ true);
 
-  Record.push_back(N->getSourceLanguage().getUnversionedName());
+  auto Lang = N->getSourceLanguage();
+  Record.push_back(Lang.getName());
+  // Set bit so the MetadataLoader can distniguish between versioned and
+  // unversioned names.
+  if (Lang.hasVersionedName())
+    Record.back() ^= (uint64_t(1) << 63);
+
   Record.push_back(VE.getMetadataOrNullID(N->getFile()));
   Record.push_back(VE.getMetadataOrNullID(N->getRawProducer()));
   Record.push_back(N->isOptimized());

llvm/lib/CodeGen/AsmPrinter/DwarfUnit.cpp

@@ -16,6 +16,7 @@
 #include "DwarfExpression.h"
 #include "llvm/ADT/APFloat.h"
 #include "llvm/ADT/APInt.h"
+#include "llvm/BinaryFormat/Dwarf.h"
 #include "llvm/CodeGen/TargetRegisterInfo.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DataLayout.h"
@@ -705,8 +706,15 @@ void DwarfUnit::addType(DIE &Entity, const DIType *Ty,
 }
 
 llvm::dwarf::SourceLanguage DwarfUnit::getSourceLanguage() const {
-  return static_cast<llvm::dwarf::SourceLanguage>(
-      getLanguage().getUnversionedName());
+  const auto &Lang = getLanguage();
+
+  if (!Lang.hasVersionedName())
+    return static_cast<llvm::dwarf::SourceLanguage>(Lang.getName());
+
+  return llvm::dwarf::toDW_LANG(
+             static_cast<llvm::dwarf::SourceLanguageName>(Lang.getName()),
+             Lang.getVersion())
+      .value_or(llvm::dwarf::DW_LANG_hi_user);
 }
 
 std::string DwarfUnit::getParentContextString(const DIScope *Context) const {

llvm/lib/IR/AsmWriter.cpp

@@ -2370,10 +2370,16 @@ static void writeDICompileUnit(raw_ostream &Out, const DICompileUnit *N,
   Out << "!DICompileUnit(";
   MDFieldPrinter Printer(Out, WriterCtx);
 
-  Printer.printDwarfEnum("language",
-                         N->getSourceLanguage().getUnversionedName(),
-                         dwarf::LanguageString,
-                         /* ShouldSkipZero */ false);
+  auto Lang = N->getSourceLanguage();
+  if (Lang.hasVersionedName())
+    Printer.printDwarfEnum(
+        "sourceLanguageName",
+        static_cast<llvm::dwarf::SourceLanguageName>(Lang.getName()),
+        dwarf::LanguageDescription,
+        /* ShouldSkipZero */ false);
+  else
+    Printer.printDwarfEnum("language", Lang.getName(), dwarf::LanguageString,
+                           /* ShouldSkipZero */ false);
 
   Printer.printMetadata("file", N->getRawFile(), /* ShouldSkipNull */ false);
   Printer.printString("producer", N->getProducer());