[BOLT][DWARF] Update DW_AT_comp_dir/DW_AT_dwo_name for DWO TUs (#91486)

Type unit DIE generated by clang contains DW_AT_comp_dir/DW_AT_dwo_name.
This was added to clang to help LLDB to figure out where type unit come
from when accessing an entry in a .debug_names accelerator table and
type units in .dwp file.

When BOLT writes out .dwo files it changes the name of them. User can
also specify directory of where they can be written out. Added support
to BOLT to update those attributes.

Author: ayermolo

bolt/include/bolt/Core/DIEBuilder.h

@@ -129,6 +129,9 @@ class DIEBuilder {
   uint64_t UnitSize{0};
   llvm::DenseSet<uint64_t> AllProcessed;
   DWARF5AcceleratorTable &DebugNamesTable;
+  // Unordered map to handle name collision if output DWO directory is
+  // specified.
+  std::unordered_map<std::string, uint32_t> NameToIndexMap;
 
   /// Returns current state of the DIEBuilder
   State &getState() { return *BuilderState.get(); }
@@ -384,6 +387,17 @@ class DIEBuilder {
   bool deleteValue(DIEValueList *Die, dwarf::Attribute Attribute) {
     return Die->deleteValue(Attribute);
   }
+  /// Updates DWO Name and Compilation directory for Skeleton CU \p Unit.
+  std::string updateDWONameCompDir(DebugStrOffsetsWriter &StrOffstsWriter,
+                                   DebugStrWriter &StrWriter,
+                                   DWARFUnit &SkeletonCU,
+                                   std::optional<StringRef> DwarfOutputPath,
+                                   std::optional<StringRef> DWONameToUse);
+  /// Updates DWO Name and Compilation directory for Type Units.
+  void updateDWONameCompDirForTypes(DebugStrOffsetsWriter &StrOffstsWriter,
+                                    DebugStrWriter &StrWriter, DWARFUnit &Unit,
+                                    std::optional<StringRef> DwarfOutputPath,
+                                    const StringRef DWOName);
 };
 } // namespace bolt
 } // namespace llvm

bolt/include/bolt/Core/DebugData.h

@@ -430,7 +430,7 @@ class DebugAddrWriterDwarf5 : public DebugAddrWriter {
 using DebugStrOffsetsBufferVector = SmallVector<char, 16>;
 class DebugStrOffsetsWriter {
 public:
-  DebugStrOffsetsWriter() {
+  DebugStrOffsetsWriter(BinaryContext &BC) : BC(BC) {
     StrOffsetsBuffer = std::make_unique<DebugStrOffsetsBufferVector>();
     StrOffsetsStream = std::make_unique<raw_svector_ostream>(*StrOffsetsBuffer);
   }
@@ -460,20 +460,27 @@ class DebugStrOffsetsWriter {
     StrOffsets.clear();
   }
 
+  bool isStrOffsetsSectionModified() const {
+    return StrOffsetSectionWasModified;
+  }
+
 private:
   std::unique_ptr<DebugStrOffsetsBufferVector> StrOffsetsBuffer;
   std::unique_ptr<raw_svector_ostream> StrOffsetsStream;
   std::map<uint32_t, uint32_t> IndexToAddressMap;
   SmallVector<uint32_t, 5> StrOffsets;
   std::unordered_map<uint64_t, uint64_t> ProcessedBaseOffsets;
   bool StrOffsetSectionWasModified = false;
+  BinaryContext &BC;
 };
 
 using DebugStrBufferVector = SmallVector<char, 16>;
 class DebugStrWriter {
 public:
   DebugStrWriter() = delete;
-  DebugStrWriter(BinaryContext &BC) : BC(BC) { create(); }
+  DebugStrWriter(DWARFContext &DwCtx, bool IsDWO) : DwCtx(DwCtx), IsDWO(IsDWO) {
+    create();
+  }
   std::unique_ptr<DebugStrBufferVector> releaseBuffer() {
     return std::move(StrBuffer);
   }
@@ -495,7 +502,8 @@ class DebugStrWriter {
   void create();
   std::unique_ptr<DebugStrBufferVector> StrBuffer;
   std::unique_ptr<raw_svector_ostream> StrStream;
-  BinaryContext &BC;
+  DWARFContext &DwCtx;
+  bool IsDWO;
 };
 
 enum class LocWriterKind { DebugLocWriter, DebugLoclistWriter };

bolt/include/bolt/Rewrite/DWARFRewriter.h

@@ -203,13 +203,16 @@ class DWARFRewriter {
   using OverriddenSectionsMap = std::unordered_map<DWARFSectionKind, StringRef>;
   /// Output .dwo files.
   void writeDWOFiles(DWARFUnit &, const OverriddenSectionsMap &,
-                     const std::string &, DebugLocWriter &);
+                     const std::string &, DebugLocWriter &,
+                     DebugStrOffsetsWriter &, DebugStrWriter &);
   using KnownSectionsEntry = std::pair<MCSection *, DWARFSectionKind>;
   struct DWPState {
     std::unique_ptr<ToolOutputFile> Out;
     std::unique_ptr<BinaryContext> TmpBC;
     std::unique_ptr<MCStreamer> Streamer;
     std::unique_ptr<DWPStringPool> Strings;
+    /// Used to store String sections for .dwo files if they are being modified.
+    std::vector<std::unique_ptr<DebugBufferVector>> StrSections;
     const MCObjectFileInfo *MCOFI = nullptr;
     const DWARFUnitIndex *CUIndex = nullptr;
     std::deque<SmallString<32>> UncompressedSections;
@@ -230,7 +233,8 @@ class DWARFRewriter {
 
   /// add content of dwo to .dwp file.
   void updateDWP(DWARFUnit &, const OverriddenSectionsMap &, const UnitMeta &,
-                 UnitMetaVectorType &, DWPState &, DebugLocWriter &);
+                 UnitMetaVectorType &, DWPState &, DebugLocWriter &,
+                 DebugStrOffsetsWriter &, DebugStrWriter &);
 };
 
 } // namespace bolt

bolt/lib/Core/DIEBuilder.cpp

@@ -22,6 +22,7 @@
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/FileSystem.h"
 #include "llvm/Support/LEB128.h"
 
 #include <algorithm>
@@ -41,6 +42,90 @@ extern cl::opt<unsigned> Verbosity;
 namespace llvm {
 namespace bolt {
 
+/// Returns DWO Name to be used to update DW_AT_dwo_name/DW_AT_GNU_dwo_name
+/// either in CU or TU unit die. Handles case where user specifies output DWO
+/// directory, and there are duplicate names. Assumes DWO ID is unique.
+static std::string
+getDWOName(llvm::DWARFUnit &CU,
+           std::unordered_map<std::string, uint32_t> &NameToIndexMap,
+           std::optional<StringRef> &DwarfOutputPath) {
+  assert(CU.getDWOId() && "DWO ID not found.");
+  std::string DWOName = dwarf::toString(
+      CU.getUnitDIE().find({dwarf::DW_AT_dwo_name, dwarf::DW_AT_GNU_dwo_name}),
+      "");
+  assert(!DWOName.empty() &&
+         "DW_AT_dwo_name/DW_AT_GNU_dwo_name does not exist.");
+  if (DwarfOutputPath) {
+    DWOName = std::string(sys::path::filename(DWOName));
+    auto Iter = NameToIndexMap.find(DWOName);
+    if (Iter == NameToIndexMap.end())
+      Iter = NameToIndexMap.insert({DWOName, 0}).first;
+    DWOName.append(std::to_string(Iter->second));
+    ++Iter->second;
+  }
+  DWOName.append(".dwo");
+  return DWOName;
+}
+
+/// Adds a \p Str to .debug_str section.
+/// Uses \p AttrInfoVal to either update entry in a DIE for legacy DWARF using
+/// \p DebugInfoPatcher, or for DWARF5 update an index in .debug_str_offsets
+/// for this contribution of \p Unit.
+static void addStringHelper(DebugStrOffsetsWriter &StrOffstsWriter,
+                            DebugStrWriter &StrWriter, DIEBuilder &DIEBldr,
+                            DIE &Die, const DWARFUnit &Unit,
+                            DIEValue &DIEAttrInfo, StringRef Str) {
+  uint32_t NewOffset = StrWriter.addString(Str);
+  if (Unit.getVersion() >= 5) {
+    StrOffstsWriter.updateAddressMap(DIEAttrInfo.getDIEInteger().getValue(),
+                                     NewOffset);
+    return;
+  }
+  DIEBldr.replaceValue(&Die, DIEAttrInfo.getAttribute(), DIEAttrInfo.getForm(),
+                       DIEInteger(NewOffset));
+}
+
+std::string DIEBuilder::updateDWONameCompDir(
+    DebugStrOffsetsWriter &StrOffstsWriter, DebugStrWriter &StrWriter,
+    DWARFUnit &SkeletonCU, std::optional<StringRef> DwarfOutputPath,
+    std::optional<StringRef> DWONameToUse) {
+  DIE &UnitDIE = *getUnitDIEbyUnit(SkeletonCU);
+  DIEValue DWONameAttrInfo = UnitDIE.findAttribute(dwarf::DW_AT_dwo_name);
+  if (!DWONameAttrInfo)
+    DWONameAttrInfo = UnitDIE.findAttribute(dwarf::DW_AT_GNU_dwo_name);
+  if (!DWONameAttrInfo)
+    return "";
+  std::string ObjectName;
+  if (DWONameToUse)
+    ObjectName = *DWONameToUse;
+  else
+    ObjectName = getDWOName(SkeletonCU, NameToIndexMap, DwarfOutputPath);
+  addStringHelper(StrOffstsWriter, StrWriter, *this, UnitDIE, SkeletonCU,
+                  DWONameAttrInfo, ObjectName);
+
+  DIEValue CompDirAttrInfo = UnitDIE.findAttribute(dwarf::DW_AT_comp_dir);
+  assert(CompDirAttrInfo && "DW_AT_comp_dir is not in Skeleton CU.");
+
+  if (DwarfOutputPath) {
+    if (!sys::fs::exists(*DwarfOutputPath))
+      sys::fs::create_directory(*DwarfOutputPath);
+    addStringHelper(StrOffstsWriter, StrWriter, *this, UnitDIE, SkeletonCU,
+                    CompDirAttrInfo, *DwarfOutputPath);
+  }
+  return ObjectName;
+}
+
+void DIEBuilder::updateDWONameCompDirForTypes(
+    DebugStrOffsetsWriter &StrOffstsWriter, DebugStrWriter &StrWriter,
+    DWARFUnit &Unit, std::optional<StringRef> DwarfOutputPath,
+    const StringRef DWOName) {
+  for (DWARFUnit *DU : getState().DWARF5TUVector)
+    updateDWONameCompDir(StrOffstsWriter, StrWriter, *DU, DwarfOutputPath,
+                         DWOName);
+  if (StrOffstsWriter.isStrOffsetsSectionModified())
+    StrOffstsWriter.finalizeSection(Unit, *this);
+}
+
 void DIEBuilder::updateReferences() {
   for (auto &[SrcDIEInfo, ReferenceInfo] : getState().AddrReferences) {
     DIEInfo *DstDIEInfo = ReferenceInfo.Dst;

bolt/lib/Core/DebugData.cpp

@@ -31,6 +31,7 @@
 #include <cstdint>
 #include <functional>
 #include <memory>
+#include <optional>
 #include <unordered_map>
 #include <vector>
 
@@ -867,10 +868,17 @@ void DebugStrOffsetsWriter::finalizeSection(DWARFUnit &Unit,
                                             DIEBuilder &DIEBldr) {
   std::optional<AttrInfo> AttrVal =
       findAttributeInfo(Unit.getUnitDIE(), dwarf::DW_AT_str_offsets_base);
-  if (!AttrVal)
+  if (!AttrVal && !Unit.isDWOUnit())
     return;
-  std::optional<uint64_t> Val = AttrVal->V.getAsSectionOffset();
-  assert(Val && "DW_AT_str_offsets_base Value not present.");
+  std::optional<uint64_t> Val = std::nullopt;
+  if (AttrVal) {
+    Val = AttrVal->V.getAsSectionOffset();
+  } else {
+    if (!Unit.isDWOUnit())
+      BC.errs() << "BOLT-WARNING: [internal-dwarf-error]: "
+                   "DW_AT_str_offsets_base Value not present\n";
+    Val = 0;
+  }
   DIE &Die = *DIEBldr.getUnitDIEbyUnit(Unit);
   DIEValue StrListBaseAttrInfo =
       Die.findAttribute(dwarf::DW_AT_str_offsets_base);
@@ -915,7 +923,11 @@ void DebugStrWriter::create() {
 }
 
 void DebugStrWriter::initialize() {
-  auto StrSection = BC.DwCtx->getDWARFObj().getStrSection();
+  StringRef StrSection;
+  if (IsDWO)
+    StrSection = DwCtx.getDWARFObj().getStrDWOSection();
+  else
+    StrSection = DwCtx.getDWARFObj().getStrSection();
   (*StrStream) << StrSection;
 }