[DWARF] - Introduce DWARFDebugPubTable class for dumping pub* sections.

[grimar]

Summary

It was requested during review of D26283 (which already landed) to share the code
for parsing pubnames/pubtypes.

Patch implements that parser instead of static function which was used before.
So it should be possible to reuse in LLD now and clean up the duplication code.

[dblaikie]

include/llvm/DebugInfo/DWARF/DWARFDebugPubTable.h

73–74

Probably no need to worry about using a separate range type here - just return a const ref to 'Sets' directly?

[aprantl]

include/llvm/DebugInfo/DWARF/DWARFDebugPubTable.h

22

Please doxygen-ify these comments.

[clayborg]

Thanks for doing this, LGTM

[davide]

The logic looks correct, some minor comments inline.

include/llvm/DebugInfo/DWARF/DWARFDebugPubTable.h

58

s/Entires/Entries/ maybe?

64–67

Add comments to explain what GnuStyle means.

lib/DebugInfo/DWARF/DWARFContext.cpp

201–217

You can probably express this in a more compact form (and avoid some duplication) with a mapping from DumpType to the section name. Up to you.

lib/DebugInfo/DWARF/DWARFDebugPubTable.cpp

35–37

uint8_t IndexEntryValue = (GnuStyle) ? ... : 0;

53–55

Maybe use the ternary operator? As in

OS << (GnuStyle ? ... : ...);

65

ehm, is this clang formatted? Probably yes, but I'm double checking because it looks weird.

[grimar]

include/llvm/DebugInfo/DWARF/DWARFDebugPubTable.h

22

Done.

58

Done.

64–67

Done.

73–74

Ok. I am returning ArrayRef now.

lib/DebugInfo/DWARF/DWARFContext.cpp

201–217

I'll try to think how to improve that place and suggest a separate patch.

lib/DebugInfo/DWARF/DWARFDebugPubTable.cpp

35–37

Done.

53–55

Done.

65

Yes, it was formatted. I rewrote this part a bit to look more straightforward,
hopefully it is better looking now.

include/llvm/DebugInfo/DWARF/DWARFDebugPubTable.h

+//===-- DWARFDebugPubTable.h ------------------------------------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_DEBUGINFO_DWARFDEBUGPUBTABLE_H
+#define LLVM_LIB_DEBUGINFO_DWARFDEBUGPUBTABLE_H
+
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/DataExtractor.h"
+#include "llvm/Support/Dwarf.h"
+#include <vector>
+
+namespace llvm {
+
+class raw_ostream;
+
+// Represents structure for holding and parsing .debug_pub* tables.
+class DWARFDebugPubTable {
+public:
+  struct Entry {
+    // Section offset from the beginning of the compilation unit.
+    uint32_t SecOffset;
+
+    // An entry of the various gnu_pub* debug sections.
+    llvm::dwarf::PubIndexEntryDescriptor Descriptor;
+
+    // The name of the object as given by the DW_AT_name attribute of the
+    // referenced DIE.
+    const char *Name;
+  };
+
+  // Each table consists of sets of variable length entries. Each set describes
+  // the names of global objects and functions, or global types, respectively,
+  // whose definitions are represented by debugging information entries owned by
+  // a single compilation unit.
+  struct Set {
+    // The total length of the entries for that set, not including the length
+    // field itself.
+    uint32_t Length;
+
+    // This number is specific to the name lookup table and is independent of
+    // the DWARF version number.
+    uint16_t Version;
+
+    // The offset from the beginning of the .debug_info section of the
+    // compilation unit header referenced by the set.
+    uint32_t Offset;
+
+    // The size in bytes of the contents of the .debug_info section generated to
+    // represent that compilation unit.
+    uint32_t Size;
+
+    std::vector<Entry> Entires;
+  };
+
+private:
+  typedef std::vector<Set> SetColl;
+  typedef iterator_range<SetColl::const_iterator> entry_iterator_range;
+
+  SetColl Sets;
+
+  bool GnuStyle;
+
+public:
+  DWARFDebugPubTable(StringRef Data, bool LittleEndian, bool GnuStyle);
+  void dump(StringRef Name, raw_ostream &OS) const;
+
+  entry_iterator_range entries() const {
+    return entry_iterator_range(Sets.begin(), Sets.end());
+  }
+};
+}
+
+#endif

lib/DebugInfo/DWARF/CMakeLists.txt

 add_llvm_library(LLVMDebugInfoDWARF
   DWARFAbbreviationDeclaration.cpp
   DWARFAcceleratorTable.cpp
   DWARFCompileUnit.cpp
   DWARFContext.cpp
   DWARFDebugAbbrev.cpp
   DWARFDebugArangeSet.cpp
   DWARFDebugAranges.cpp
   DWARFDebugFrame.cpp
   DWARFDebugInfoEntry.cpp
   DWARFDebugLine.cpp
   DWARFDebugLoc.cpp
   DWARFDebugMacro.cpp
+  DWARFDebugPubTable.cpp
   DWARFDebugRangeList.cpp
   DWARFDie.cpp
   DWARFFormValue.cpp
   DWARFGdbIndex.cpp
   DWARFTypeUnit.cpp
   DWARFUnitIndex.cpp
   DWARFUnit.cpp
   SyntaxHighlighting.cpp
 
   ADDITIONAL_HEADER_DIRS
   ${LLVM_MAIN_INCLUDE_DIR}/llvm/DebugInfo/DWARF
   ${LLVM_MAIN_INCLUDE_DIR}/llvm/DebugInfo
   )

lib/DebugInfo/DWARF/DWARFContext.cpp

 //===-- DWARFContext.cpp --------------------------------------------------===//
 //
 //                     The LLVM Compiler Infrastructure
 //
 // This file is distributed under the University of Illinois Open Source
 // License. See LICENSE.TXT for details.
 //
 //===----------------------------------------------------------------------===//
 
 #include "llvm/DebugInfo/DWARF/DWARFContext.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringSwitch.h"
 #include "llvm/DebugInfo/DWARF/DWARFAcceleratorTable.h"
 #include "llvm/DebugInfo/DWARF/DWARFDebugArangeSet.h"
+#include "llvm/DebugInfo/DWARF/DWARFDebugPubTable.h"
 #include "llvm/DebugInfo/DWARF/DWARFUnitIndex.h"
 #include "llvm/Object/MachO.h"
 #include "llvm/Object/RelocVisitor.h"
 #include "llvm/Support/Compression.h"
 #include "llvm/Support/Dwarf.h"
 #include "llvm/Support/ELF.h"
 #include "llvm/Support/Format.h"
 #include "llvm/Support/Path.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
 using namespace llvm;
 using namespace dwarf;
 using namespace object;
 
 #define DEBUG_TYPE "dwarf"
 
 typedef DWARFDebugLine::LineTable DWARFLineTable;
 typedef DILineInfoSpecifier::FileLineInfoKind FileLineInfoKind;
 typedef DILineInfoSpecifier::FunctionNameKind FunctionNameKind;
 
-static void dumpPubSection(raw_ostream &OS, StringRef Name, StringRef Data,
-                           bool LittleEndian, bool GnuStyle) {
-  OS << "\n." << Name << " contents:\n";
-  DataExtractor pubNames(Data, LittleEndian, 0);
-  uint32_t offset = 0;
-  while (pubNames.isValidOffset(offset)) {
-    OS << "length = " << format("0x%08x", pubNames.getU32(&offset));
-    OS << " version = " << format("0x%04x", pubNames.getU16(&offset));
-    OS << " unit_offset = " << format("0x%08x", pubNames.getU32(&offset));
-    OS << " unit_size = " << format("0x%08x", pubNames.getU32(&offset)) << '\n';
-    if (GnuStyle)
-      OS << "Offset     Linkage  Kind     Name\n";
-    else
-      OS << "Offset     Name\n";
-
-    while (offset < Data.size()) {
-      uint32_t dieRef = pubNames.getU32(&offset);
-      if (dieRef == 0)
-        break;
-      OS << format("0x%8.8x ", dieRef);
-      if (GnuStyle) {
-        PubIndexEntryDescriptor desc(pubNames.getU8(&offset));
-        OS << format("%-8s",
-                     dwarf::GDBIndexEntryLinkageString(desc.Linkage).data())
-           << ' '
-           << format("%-8s", dwarf::GDBIndexEntryKindString(desc.Kind).data())
-           << ' ';
-      }
-      OS << '\"' << pubNames.getCStr(&offset) << "\"\n";
-    }
-  }
-}
-
 static void dumpAccelSection(raw_ostream &OS, StringRef Name,
                              const DWARFSection& Section, StringRef StringSection,
                              bool LittleEndian) {
   DataExtractor AccelSection(Section.Data, LittleEndian, 0);
   DataExtractor StrData(StringSection, LittleEndian, 0);
   OS << "\n." << Name << " contents:\n";
   DWARFAcceleratorTable Accel(AccelSection, StrData, Section.Relocs);
   if (!Accel.extract())
if (DumpType == DIDT_All || DumpType == DIDT_Ranges) {
     DataExtractor rangesData(getRangeSection(), isLittleEndian(),
                              savedAddressByteSize);
     offset = 0;
     DWARFDebugRangeList rangeList;
     while (rangeList.extract(rangesData, &offset))
       rangeList.dump(OS);
   }
 
   if (DumpType == DIDT_All || DumpType == DIDT_Pubnames)
-    dumpPubSection(OS, "debug_pubnames", getPubNamesSection(),
-                   isLittleEndian(), false);
+    DWARFDebugPubTable(getPubNamesSection(), isLittleEndian(), false)
+        .dump("debug_pubnames", OS);
 
   if (DumpType == DIDT_All || DumpType == DIDT_Pubtypes)
-    dumpPubSection(OS, "debug_pubtypes", getPubTypesSection(),
-                   isLittleEndian(), false);
+    DWARFDebugPubTable(getPubTypesSection(), isLittleEndian(), false)
+        .dump("debug_pubtypes", OS);
 
   if (DumpType == DIDT_All || DumpType == DIDT_GnuPubnames)
-    dumpPubSection(OS, "debug_gnu_pubnames", getGnuPubNamesSection(),
-                   isLittleEndian(), true /* GnuStyle */);
+    DWARFDebugPubTable(getGnuPubNamesSection(), isLittleEndian(),
+                       true /* GnuStyle */)
+        .dump("debug_gnu_pubnames", OS);
 
   if (DumpType == DIDT_All || DumpType == DIDT_GnuPubtypes)
-    dumpPubSection(OS, "debug_gnu_pubtypes", getGnuPubTypesSection(),
-                   isLittleEndian(), true /* GnuStyle */);
+    DWARFDebugPubTable(getGnuPubTypesSection(), isLittleEndian(),
+                       true /* GnuStyle */)
+        .dump("debug_gnu_pubtypes", OS);
 
   if ((DumpType == DIDT_All || DumpType == DIDT_StrOffsetsDwo) &&
       !getStringOffsetDWOSection().empty()) {
     OS << "\n.debug_str_offsets.dwo contents:\n";
     DataExtractor strOffsetExt(getStringOffsetDWOSection(), isLittleEndian(),
                                0);
     offset = 0;
     uint64_t size = getStringOffsetDWOSection().size();

lib/DebugInfo/DWARF/DWARFDebugPubTable.cpp

+//===-- DWARFDebugPubTable.cpp ---------------------------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/DebugInfo/DWARF/DWARFDebugPubTable.h"
+#include "llvm/Support/Format.h"
+#include "llvm/Support/raw_ostream.h"
+
+using namespace llvm;
+using namespace llvm::dwarf;
+
+DWARFDebugPubTable::DWARFDebugPubTable(StringRef Data, bool LittleEndian,
+                                       bool GnuStyle)
+    : GnuStyle(GnuStyle) {
+  DataExtractor PubNames(Data, LittleEndian, 0);
+  uint32_t Offset = 0;
+  while (PubNames.isValidOffset(Offset)) {
+    Sets.push_back({});
+    Set &SetData = Sets.back();
+
+    SetData.Length = PubNames.getU32(&Offset);
+    SetData.Version = PubNames.getU16(&Offset);
+    SetData.Offset = PubNames.getU32(&Offset);
+    SetData.Size = PubNames.getU32(&Offset);
+
+    while (Offset < Data.size()) {
+      uint32_t DieRef = PubNames.getU32(&Offset);
+      if (DieRef == 0)
+        break;
+      uint8_t IndexEntryValue = 0;
+      if (GnuStyle)
+        IndexEntryValue = PubNames.getU8(&Offset);
+      const char *Name = PubNames.getCStr(&Offset);
+      SetData.Entires.push_back(
+          {DieRef, PubIndexEntryDescriptor(IndexEntryValue), Name});
+    }
+  }
+}
+
+void DWARFDebugPubTable::dump(StringRef Name, raw_ostream &OS) const {
+  OS << "\n." << Name << " contents: a\n";
+  for (const Set &S : Sets) {
+    OS << "length = " << format("0x%08x", S.Length);
+    OS << " version = " << format("0x%04x", S.Version);
+    OS << " unit_offset = " << format("0x%08x", S.Offset);
+    OS << " unit_size = " << format("0x%08x", S.Size) << '\n';
+    if (GnuStyle)
+      OS << "Offset     Linkage  Kind     Name\n";
+    else
+      OS << "Offset     Name\n";
+
+    for (const Entry &E : S.Entires) {
+      OS << format("0x%8.8x ", E.SecOffset);
+      if (GnuStyle)
+        OS << format("%-8s",
+                     dwarf::GDBIndexEntryLinkageString(E.Descriptor.Linkage)
+                         .data())
+           << ' '
+           << format("%-8s",
+                     dwarf::GDBIndexEntryKindString(E.Descriptor.Kind).data())
+           << ' ';
+      OS << '\"' << E.Name << "\"\n";
+    }
+  }
+}