Allow for mixing source/no-source DIFiles in one CU

The DWARF proposal that the DW_LNCT_LLVM_source extension is based on
(https://dwarfstd.org/issues/180201.1.html) allows to mix source and
non-source files in the same CU by storing an empty string as a
sentinel value.

This patch implements this feature.

Review in #73877

Author: adrian-prantl

llvm/include/llvm/MC/MCDwarf.h

@@ -265,7 +265,8 @@ struct MCDwarfLineTableHeader {
   StringMap<unsigned> SourceIdMap;
   std::string CompilationDir;
   MCDwarfFile RootFile;
-  bool HasSource = false;
+  bool HasAnySource = false;
+
 private:
   bool HasAllMD5 = true;
   bool HasAnyMD5 = false;
@@ -305,15 +306,15 @@ struct MCDwarfLineTableHeader {
     RootFile.Checksum = Checksum;
     RootFile.Source = Source;
     trackMD5Usage(Checksum.has_value());
-    HasSource = Source.has_value();
+    HasAnySource |= Source.has_value();
   }
 
   void resetFileTable() {
     MCDwarfDirs.clear();
     MCDwarfFiles.clear();
     RootFile.Name.clear();
     resetMD5Usage();
-    HasSource = false;
+    HasAnySource = false;
   }
 
 private:
@@ -385,7 +386,7 @@ class MCDwarfLineTable {
     Header.RootFile.Checksum = Checksum;
     Header.RootFile.Source = Source;
     Header.trackMD5Usage(Checksum.has_value());
-    Header.HasSource = Source.has_value();
+    Header.HasAnySource |= Source.has_value();
   }
 
   void resetFileTable() { Header.resetFileTable(); }

llvm/lib/DebugInfo/DWARF/DWARFDebugLine.cpp

@@ -170,9 +170,14 @@ void DWARFDebugLine::Prologue::dump(raw_ostream &OS,
       if (ContentTypes.HasLength)
         OS << format("         length: 0x%8.8" PRIx64 "\n", FileEntry.Length);
       if (ContentTypes.HasSource) {
-        OS <<        "         source: ";
-        FileEntry.Source.dump(OS, DumpOptions);
-        OS << '\n';
+        auto Source = FileEntry.Source.getAsCString();
+        if (!Source)
+          consumeError(Source.takeError());
+        else if ((*Source)[0]) {
+          OS << "         source: ";
+          FileEntry.Source.dump(OS, DumpOptions);
+          OS << '\n';
+        }
       }
     }
   }

llvm/lib/MC/MCDwarf.cpp

@@ -386,7 +386,7 @@ void MCDwarfLineTableHeader::emitV2FileDirTables(MCStreamer *MCOS) const {
 }
 
 static void emitOneV5FileEntry(MCStreamer *MCOS, const MCDwarfFile &DwarfFile,
-                               bool EmitMD5, bool HasSource,
+                               bool EmitMD5, bool HasAnySource,
                                std::optional<MCDwarfLineStr> &LineStr) {
   assert(!DwarfFile.Name.empty());
   if (LineStr)
@@ -401,7 +401,7 @@ static void emitOneV5FileEntry(MCStreamer *MCOS, const MCDwarfFile &DwarfFile,
     MCOS->emitBinaryData(
         StringRef(reinterpret_cast<const char *>(Cksum.data()), Cksum.size()));
   }
-  if (HasSource) {
+  if (HasAnySource) {
     if (LineStr)
       LineStr->emitRef(MCOS, DwarfFile.Source.value_or(StringRef()));
     else {
@@ -452,7 +452,7 @@ void MCDwarfLineTableHeader::emitV5FileDirTables(
   uint64_t Entries = 2;
   if (HasAllMD5)
     Entries += 1;
-  if (HasSource)
+  if (HasAnySource)
     Entries += 1;
   MCOS->emitInt8(Entries);
   MCOS->emitULEB128IntValue(dwarf::DW_LNCT_path);
@@ -464,7 +464,7 @@ void MCDwarfLineTableHeader::emitV5FileDirTables(
     MCOS->emitULEB128IntValue(dwarf::DW_LNCT_MD5);
     MCOS->emitULEB128IntValue(dwarf::DW_FORM_data16);
   }
-  if (HasSource) {
+  if (HasAnySource) {
     MCOS->emitULEB128IntValue(dwarf::DW_LNCT_LLVM_source);
     MCOS->emitULEB128IntValue(LineStr ? dwarf::DW_FORM_line_strp
                                       : dwarf::DW_FORM_string);
@@ -479,9 +479,9 @@ void MCDwarfLineTableHeader::emitV5FileDirTables(
   assert((!RootFile.Name.empty() || MCDwarfFiles.size() >= 1) &&
          "No root file and no .file directives");
   emitOneV5FileEntry(MCOS, RootFile.Name.empty() ? MCDwarfFiles[1] : RootFile,
-                     HasAllMD5, HasSource, LineStr);
+                     HasAllMD5, HasAnySource, LineStr);
   for (unsigned i = 1; i < MCDwarfFiles.size(); ++i)
-    emitOneV5FileEntry(MCOS, MCDwarfFiles[i], HasAllMD5, HasSource, LineStr);
+    emitOneV5FileEntry(MCOS, MCDwarfFiles[i], HasAllMD5, HasAnySource, LineStr);
 }
 
 std::pair<MCSymbol *, MCSymbol *>
@@ -598,7 +598,7 @@ MCDwarfLineTableHeader::tryGetFile(StringRef &Directory, StringRef &FileName,
   // If any files have embedded source, they all must.
   if (MCDwarfFiles.empty()) {
     trackMD5Usage(Checksum.has_value());
-    HasSource = (Source != std::nullopt);
+    HasAnySource |= Source.has_value();
   }
   if (DwarfVersion >= 5 && isRootFile(RootFile, Directory, FileName, Checksum))
     return 0;
@@ -625,11 +625,6 @@ MCDwarfLineTableHeader::tryGetFile(StringRef &Directory, StringRef &FileName,
     return make_error<StringError>("file number already allocated",
                                    inconvertibleErrorCode());
 
-  // If any files have embedded source, they all must.
-  if (HasSource != (Source != std::nullopt))
-    return make_error<StringError>("inconsistent use of embedded source",
-                                   inconvertibleErrorCode());
-
   if (Directory.empty()) {
     // Separate the directory part from the basename of the FileName.
     StringRef tFileName = sys::path::filename(FileName);
@@ -662,8 +657,8 @@ MCDwarfLineTableHeader::tryGetFile(StringRef &Directory, StringRef &FileName,
   File.Checksum = Checksum;
   trackMD5Usage(Checksum.has_value());
   File.Source = Source;
-  if (Source)
-    HasSource = true;
+  if (Source.has_value())
+    HasAnySource = true;
 
   // return the allocated FileNumber.
   return FileNumber;

llvm/test/DebugInfo/Generic/mixed-source.ll

@@ -0,0 +1,37 @@
+; RUN: %llc_dwarf -O0 -filetype=obj -o - < %s | llvm-dwarfdump -debug-line - | FileCheck %s
+
+; CHECK: include_directories[  0] = "dir"
+; CHECK-NEXT: file_names[  0]:
+; CHECK-NEXT:            name: "foo.c"
+; CHECK-NEXT:       dir_index: 0
+; CHECK-NEXT:          source: "void foo() { }\n"
+; CHECK-NEXT: file_names[  1]:
+; CHECK-NEXT:            name: "bar.h"
+; CHECK-NEXT:       dir_index: 0
+; CHECK-NOT:           source:
+
+; Test that DIFiles mixing source and no-source within a DICompileUnit works.
+
+define dso_local void @foo() !dbg !5 {
+  ret void, !dbg !7
+}
+
+define dso_local void @bar() !dbg !6 {
+  ret void, !dbg !8
+}
+
+!llvm.dbg.cu = !{!4}
+!llvm.module.flags = !{!0, !1}
+
+!0 = !{i32 2, !"Dwarf Version", i32 5}
+!1 = !{i32 2, !"Debug Info Version", i32 3}
+
+!2 = !DIFile(filename: "foo.c", directory: "dir", source: "void foo() { }\0A")
+!3 = !DIFile(filename: "bar.h", directory: "dir")
+
+!4 = distinct !DICompileUnit(language: DW_LANG_C99, emissionKind: FullDebug, file: !2)
+!5 = distinct !DISubprogram(name: "foo", file: !2, line: 1, type: !9, scopeLine: 1, spFlags: DISPFlagDefinition, unit: !4)
+!6 = distinct !DISubprogram(name: "bar", file: !3, line: 1, type: !9, scopeLine: 1, spFlags: DISPFlagDefinition, unit: !4)
+!7 = !DILocation(line: 1, scope: !5)
+!8 = !DILocation(line: 1, scope: !6)
+!9 = !DISubroutineType(types: !{})