Add support for separating strings for the split debug info DWARF5

proposal. This leaves the strings in the skeleton die as strp,
but in all dwo files they're accessed now via DW_FORM_GNU_str_index.

Add support for dumping these sections and modify the fission-cu.ll
testcase to have the correct strings and form. Fix a small bug
in the fixed form sizes routine that involved out of array accesses
for the table and add a FIXME in the extractFast routine to fix
this up.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@171779 91177308-0d34-0410-b5e6-96231b3b80d8

12 files changed, 120 insertions, 19 deletions

diff --git a/lib/CodeGen/AsmPrinter/DIE.cpp b/lib/CodeGen/AsmPrinter/DIE.cpp
index 32d6cb2..f33f0f9 100644
--- a/lib/CodeGen/AsmPrinter/DIE.cpp
+++ b/lib/CodeGen/AsmPrinter/DIE.cpp
@@ -197,6 +197,7 @@ void DIEInteger::EmitValue(AsmPrinter *Asm, unsigned Form) const {
   case dwarf::DW_FORM_data4: Size = 4; break;
   case dwarf::DW_FORM_ref8:  // Fall thru
   case dwarf::DW_FORM_data8: Size = 8; break;
+  case dwarf::DW_FORM_GNU_str_index: Asm->EmitULEB128(Integer); return;
   case dwarf::DW_FORM_udata: Asm->EmitULEB128(Integer); return;
   case dwarf::DW_FORM_sdata: Asm->EmitSLEB128(Integer); return;
   case dwarf::DW_FORM_addr:
@@ -220,6 +221,7 @@ unsigned DIEInteger::SizeOf(AsmPrinter *AP, unsigned Form) const {
   case dwarf::DW_FORM_data4: return sizeof(int32_t);
   case dwarf::DW_FORM_ref8:  // Fall thru
   case dwarf::DW_FORM_data8: return sizeof(int64_t);
+  case dwarf::DW_FORM_GNU_str_index: return MCAsmInfo::getULEB128Size(Integer);
   case dwarf::DW_FORM_udata: return MCAsmInfo::getULEB128Size(Integer);
   case dwarf::DW_FORM_sdata: return MCAsmInfo::getSLEB128Size(Integer);
   case dwarf::DW_FORM_addr:  return AP->getDataLayout().getPointerSize();
diff --git a/lib/CodeGen/AsmPrinter/DwarfCompileUnit.cpp b/lib/CodeGen/AsmPrinter/DwarfCompileUnit.cpp
index 854d82c..60d7cd3 100644
--- a/lib/CodeGen/AsmPrinter/DwarfCompileUnit.cpp
+++ b/lib/CodeGen/AsmPrinter/DwarfCompileUnit.cpp
@@ -128,6 +128,27 @@ void CompileUnit::addSInt(DIE *Die, unsigned Attribute,
 /// reference to the string pool instead of immediate strings so that DIEs have
 /// more predictable sizes.
 void CompileUnit::addString(DIE *Die, unsigned Attribute, StringRef String) {
+  if (!DD->useSplitDwarf()) {
+    MCSymbol *Symb = DU->getStringPoolEntry(String);
+    DIEValue *Value;
+    if (Asm->needsRelocationsForDwarfStringPool())
+      Value = new (DIEValueAllocator) DIELabel(Symb);
+    else {
+      MCSymbol *StringPool = DU->getStringPoolSym();
+      Value = new (DIEValueAllocator) DIEDelta(Symb, StringPool);
+    }
+    Die->addValue(Attribute, dwarf::DW_FORM_strp, Value);
+  } else {
+    unsigned idx = DU->getStringPoolIndex(String);
+    DIEValue *Value = new (DIEValueAllocator) DIEInteger(idx);
+    Die->addValue(Attribute, dwarf::DW_FORM_GNU_str_index, Value);
+  }
+}
+
+/// addLocalString - Add a string attribute data and value. This is guaranteed
+/// to be in the local string pool instead of indirected.
+void CompileUnit::addLocalString(DIE *Die, unsigned Attribute,
+                                 StringRef String) {
   MCSymbol *Symb = DU->getStringPoolEntry(String);
   DIEValue *Value;
   if (Asm->needsRelocationsForDwarfStringPool())
diff --git a/lib/CodeGen/AsmPrinter/DwarfCompileUnit.h b/lib/CodeGen/AsmPrinter/DwarfCompileUnit.h
index 368234b..f210dcc 100644
--- a/lib/CodeGen/AsmPrinter/DwarfCompileUnit.h
+++ b/lib/CodeGen/AsmPrinter/DwarfCompileUnit.h
@@ -198,6 +198,10 @@ public:
   ///
   void addString(DIE *Die, unsigned Attribute, const StringRef Str);
 
+  /// addLocalString - Add a string attribute data and value.
+  ///
+  void addLocalString(DIE *Die, unsigned Attribute, const StringRef Str);
+
   /// addLabel - Add a Dwarf label attribute data and value.
   ///
   void addLabel(DIE *Die, unsigned Attribute, unsigned Form,
diff --git a/lib/CodeGen/AsmPrinter/DwarfDebug.cpp b/lib/CodeGen/AsmPrinter/DwarfDebug.cpp
index 321256a..11b853f 100644
--- a/lib/CodeGen/AsmPrinter/DwarfDebug.cpp
+++ b/lib/CodeGen/AsmPrinter/DwarfDebug.cpp
@@ -228,6 +228,16 @@ MCSymbol *DwarfUnits::getStringPoolEntry(StringRef Str) {
   return Entry.first = Asm->GetTempSymbol(StringPref, Entry.second);
 }
 
+unsigned DwarfUnits::getStringPoolIndex(StringRef Str) {
+  std::pair<MCSymbol*, unsigned> &Entry =
+    StringPool->GetOrCreateValue(Str).getValue();
+  if (Entry.first) return Entry.second;
+
+  Entry.second = NextStringPoolNumber++;
+  Entry.first = Asm->GetTempSymbol(StringPref, Entry.second);
+  return Entry.second;
+}
+
 // Define a unique number for the abbreviation.
 //
 void DwarfUnits::assignAbbrevNumber(DIEAbbrev &Abbrev) {
@@ -2116,12 +2126,14 @@ void DwarfDebug::emitDebugPubTypes() {
 }
 
 // Emit strings into a string section.
-void DwarfUnits::emitStrings(const MCSection *Section) {
+void DwarfUnits::emitStrings(const MCSection *StrSection,
+                             const MCSection *OffsetSection = NULL,
+                             const MCSymbol *StrSecSym = NULL) {
 
   if (StringPool->empty()) return;
 
   // Start the dwarf str section.
-  Asm->OutStreamer.SwitchSection(Section);
+  Asm->OutStreamer.SwitchSection(StrSection);
 
   // Get all of the string pool entries and put them in an array by their ID so
   // we can sort them.
@@ -2144,6 +2156,17 @@ void DwarfUnits::emitStrings(const MCSection *Section) {
                                          Entries[i].second->getKeyLength()+1),
                                0/*addrspace*/);
   }
+
+  // If we've got an offset section go ahead and emit that now as well.
+  if (OffsetSection) {
+    Asm->OutStreamer.SwitchSection(OffsetSection);
+    unsigned offset = 0;
+    unsigned size = 4;
+    for (unsigned i = 0, e = Entries.size(); i != e; ++i) {
+      Asm->OutStreamer.EmitIntValue(offset, size, 0);
+      offset += Entries[i].second->getKeyLength() + 1;
+    }
+  }
 }
 
 // Emit visible names into a debug str section.
@@ -2377,7 +2400,7 @@ CompileUnit *DwarfDebug::constructSkeletonCU(const MDNode *N) {
                                        DIUnit.getLanguage(), Die, Asm,
                                        this, &SkeletonHolder);
   // FIXME: This should be the .dwo file.
-  NewCU->addString(Die, dwarf::DW_AT_GNU_dwo_name, FN);
+  NewCU->addLocalString(Die, dwarf::DW_AT_GNU_dwo_name, FN);
 
   // FIXME: We also need DW_AT_addr_base and DW_AT_dwo_id.
 
@@ -2393,7 +2416,7 @@ CompileUnit *DwarfDebug::constructSkeletonCU(const MDNode *N) {
     NewCU->addUInt(Die, dwarf::DW_AT_stmt_list, dwarf::DW_FORM_data4, 0);
 
   if (!CompilationDir.empty())
-    NewCU->addString(Die, dwarf::DW_AT_comp_dir, CompilationDir);
+    NewCU->addLocalString(Die, dwarf::DW_AT_comp_dir, CompilationDir);
 
   SkeletonHolder.addUnit(NewCU);
 
@@ -2458,5 +2481,8 @@ void DwarfDebug::emitDebugAbbrevDWO() {
 // sections.
 void DwarfDebug::emitDebugStrDWO() {
   assert(useSplitDwarf() && "No split dwarf?");
-  InfoHolder.emitStrings(Asm->getObjFileLowering().getDwarfStrDWOSection());
+  const MCSection *OffSec = Asm->getObjFileLowering().getDwarfStrOffDWOSection();
+  const MCSymbol *StrSym = DwarfStrSectionSym;
+  InfoHolder.emitStrings(Asm->getObjFileLowering().getDwarfStrDWOSection(),
+                         OffSec, StrSym);
 }
diff --git a/lib/CodeGen/AsmPrinter/DwarfDebug.h b/lib/CodeGen/AsmPrinter/DwarfDebug.h
index 59e4890..cc6c0f7 100644
--- a/lib/CodeGen/AsmPrinter/DwarfDebug.h
+++ b/lib/CodeGen/AsmPrinter/DwarfDebug.h
@@ -240,7 +240,7 @@ public:
                  const MCSymbol *);
 
   /// \brief Emit all of the strings to the section given.
-  void emitStrings(const MCSection *);
+  void emitStrings(const MCSection *, const MCSection *, const MCSymbol *);
 
   /// \brief Returns the entry into the start of the pool.
   MCSymbol *getStringPoolSym();
@@ -249,6 +249,10 @@ public:
   /// string text.
   MCSymbol *getStringPoolEntry(StringRef Str);
 
+  /// \brief Returns the index into the string pool with the given
+  /// string text.
+  unsigned getStringPoolIndex(StringRef Str);
+
   /// \brief Returns the string pool.
   StrPool *getStringPool() { return StringPool; }
 };
diff --git a/lib/DebugInfo/DWARFCompileUnit.h b/lib/DebugInfo/DWARFCompileUnit.h
index ba638df..c58664f 100644
--- a/lib/DebugInfo/DWARFCompileUnit.h
+++ b/lib/DebugInfo/DWARFCompileUnit.h
@@ -28,6 +28,7 @@ class DWARFCompileUnit {
   StringRef AbbrevSection;
   StringRef RangeSection;
   StringRef StringSection;
+  StringRef StringOffsetSection;
   const RelocAddrMap *RelocMap;
   bool isLittleEndian;
 
@@ -42,13 +43,16 @@ class DWARFCompileUnit {
 public:
 
   DWARFCompileUnit(const DWARFDebugAbbrev *DA, StringRef IS, StringRef AS,
-                   StringRef RS, StringRef SS, const RelocAddrMap *M, bool LE) :
+                   StringRef RS, StringRef SS, StringRef SOS,
+                   const RelocAddrMap *M, bool LE) :
     Abbrev(DA), InfoSection(IS), AbbrevSection(AS),
-    RangeSection(RS), StringSection(SS), RelocMap(M), isLittleEndian(LE) {
+    RangeSection(RS), StringSection(SS), StringOffsetSection(SOS),
+    RelocMap(M), isLittleEndian(LE) {
     clear();
   }
 
   StringRef getStringSection() const { return StringSection; }
+  StringRef getStringOffsetSection() const { return StringOffsetSection; }
   const RelocAddrMap *getRelocMap() const { return RelocMap; }
   DataExtractor getDebugInfoExtractor() const;
 
diff --git a/lib/DebugInfo/DWARFContext.cpp b/lib/DebugInfo/DWARFContext.cpp
index 1270e6e..d6d9fcf 100644
--- a/lib/DebugInfo/DWARFContext.cpp
+++ b/lib/DebugInfo/DWARFContext.cpp
@@ -152,7 +152,8 @@ void DWARFContext::parseCompileUnits() {
   while (DIData.isValidOffset(offset)) {
     CUs.push_back(DWARFCompileUnit(getDebugAbbrev(), getInfoSection(),
                                    getAbbrevSection(), getRangeSection(),
-                                   getStringSection(), &infoRelocMap(),
+                                   getStringSection(), "",
+                                   &infoRelocMap(),
                                    isLittleEndian()));
     if (!CUs.back().extract(DIData, &offset)) {
       CUs.pop_back();
@@ -172,6 +173,7 @@ void DWARFContext::parseDWOCompileUnits() {
                                       getAbbrevDWOSection(),
                                       getRangeDWOSection(),
                                       getStringDWOSection(),
+                                      getStringOffsetDWOSection(),
                                       &infoDWORelocMap(),
                                       isLittleEndian()));
     if (!DWOCUs.back().extract(DIData, &offset)) {
@@ -382,6 +384,8 @@ DWARFContextInMemory::DWARFContextInMemory(object::ObjectFile *Obj) :
       AbbrevDWOSection = data;
     else if (name == "debug_str.dwo")
       StringDWOSection = data;
+    else if (name == "debug_str_offsets.dwo")
+      StringOffsetDWOSection = data;
     // Any more debug info sections go here.
     else
       continue;
diff --git a/lib/DebugInfo/DWARFContext.h b/lib/DebugInfo/DWARFContext.h
index 0c3eb30..c6d6283 100644
--- a/lib/DebugInfo/DWARFContext.h
+++ b/lib/DebugInfo/DWARFContext.h
@@ -106,6 +106,7 @@ public:
   virtual StringRef getInfoDWOSection() = 0;
   virtual StringRef getAbbrevDWOSection() = 0;
   virtual StringRef getStringDWOSection() = 0;
+  virtual StringRef getStringOffsetDWOSection() = 0;
   virtual StringRef getRangeDWOSection() = 0;
   virtual const RelocAddrMap &infoDWORelocMap() const = 0;
 
@@ -140,6 +141,7 @@ class DWARFContextInMemory : public DWARFContext {
   StringRef InfoDWOSection;
   StringRef AbbrevDWOSection;
   StringRef StringDWOSection;
+  StringRef StringOffsetDWOSection;
   StringRef RangeDWOSection;
 
 public:
@@ -157,6 +159,9 @@ public:
   virtual StringRef getInfoDWOSection() { return InfoDWOSection; }
   virtual StringRef getAbbrevDWOSection() { return AbbrevDWOSection; }
   virtual StringRef getStringDWOSection() { return StringDWOSection; }
+  virtual StringRef getStringOffsetDWOSection() {
+    return StringOffsetDWOSection;
+  }
   virtual StringRef getRangeDWOSection() { return RangeDWOSection; }
   virtual const RelocAddrMap &infoDWORelocMap() const { return InfoDWORelocMap; }
 };
diff --git a/lib/DebugInfo/DWARFDebugInfoEntry.cpp b/lib/DebugInfo/DWARFDebugInfoEntry.cpp
index c083b8c..bb11850 100644
--- a/lib/DebugInfo/DWARFDebugInfoEntry.cpp
+++ b/lib/DebugInfo/DWARFDebugInfoEntry.cpp
@@ -12,6 +12,7 @@
 #include "DWARFContext.h"
 #include "DWARFDebugAbbrev.h"
 #include "DWARFFormValue.h"
+#include "llvm/Support/Debug.h"
 #include "llvm/Support/Dwarf.h"
 #include "llvm/Support/Format.h"
 #include "llvm/Support/raw_ostream.h"
@@ -113,9 +114,14 @@ bool DWARFDebugInfoEntryMinimal::extractFast(const DWARFCompileUnit *cu,
     uint32_t i;
     uint16_t form;
     for (i=0; i<numAttributes; ++i) {
+
       form = AbbrevDecl->getFormByIndex(i);
 
-      const uint8_t fixed_skip_size = fixed_form_sizes[form];
+      // FIXME: Currently we're checking if this is less than the last
+      // entry in the fixed_form_sizes table, but this should be changed
+      // to use dynamic dispatch.
+      const uint8_t fixed_skip_size = (form < DW_FORM_ref_sig8) ?
+                                       fixed_form_sizes[form] : 0;
       if (fixed_skip_size)
         offset += fixed_skip_size;
       else {
@@ -187,6 +193,8 @@ bool DWARFDebugInfoEntryMinimal::extractFast(const DWARFCompileUnit *cu,
           case DW_FORM_sdata:
           case DW_FORM_udata:
           case DW_FORM_ref_udata:
+          case DW_FORM_GNU_str_index:
+          case DW_FORM_GNU_addr_index:
             debug_info_data.getULEB128(&offset);
             break;
 
@@ -207,7 +215,6 @@ bool DWARFDebugInfoEntryMinimal::extractFast(const DWARFCompileUnit *cu,
             return false;
           }
           offset += form_size;
-
         } while (form_is_indirect);
       }
     }
@@ -327,6 +334,8 @@ DWARFDebugInfoEntryMinimal::extract(const DWARFCompileUnit *cu,
               case DW_FORM_sdata:
               case DW_FORM_udata:
               case DW_FORM_ref_udata:
+              case DW_FORM_GNU_str_index:
+              case DW_FORM_GNU_addr_index:
                 debug_info_data.getULEB128(&offset);
                 break;
 
diff --git a/lib/DebugInfo/DWARFFormValue.cpp b/lib/DebugInfo/DWARFFormValue.cpp
index 1610db2..47fad8b 100644
--- a/lib/DebugInfo/DWARFFormValue.cpp
+++ b/lib/DebugInfo/DWARFFormValue.cpp
@@ -46,8 +46,6 @@ static const uint8_t form_sizes_addr4[] = {
   0, // 0x18 DW_FORM_exprloc
   0, // 0x19 DW_FORM_flag_present
   8, // 0x20 DW_FORM_ref_sig8
-  4, // 0x1f01 DW_FORM_GNU_addr_index
-  4, // 0x1f02 DW_FORM_GNU_str_index
 };
 
 static const uint8_t form_sizes_addr8[] = {
@@ -78,8 +76,6 @@ static const uint8_t form_sizes_addr8[] = {
   0, // 0x18 DW_FORM_exprloc
   0, // 0x19 DW_FORM_flag_present
   8, // 0x20 DW_FORM_ref_sig8
-  8, // 0x1f01 DW_FORM_GNU_addr_index
-  8, // 0x1f01 DW_FORM_GNU_str_index
 };
 
 const uint8_t *
@@ -295,6 +291,8 @@ DWARFFormValue::skipValue(uint16_t form, DataExtractor debug_info_data,
     case DW_FORM_sdata:
     case DW_FORM_udata:
     case DW_FORM_ref_udata:
+    case DW_FORM_GNU_str_index:
+    case DW_FORM_GNU_addr_index:
       debug_info_data.getULEB128(offset_ptr);
       return true;
 
@@ -321,6 +319,7 @@ DWARFFormValue::skipValue(uint16_t form, DataExtractor debug_info_data,
 void
 DWARFFormValue::dump(raw_ostream &OS, const DWARFCompileUnit *cu) const {
   DataExtractor debug_str_data(cu->getStringSection(), true, 0);
+  DataExtractor debug_str_offset_data(cu->getStringOffsetSection(), true, 0);
   uint64_t uvalue = getUnsigned();
   bool cu_relative_offset = false;
 
@@ -379,6 +378,17 @@ DWARFFormValue::dump(raw_ostream &OS, const DWARFCompileUnit *cu) const {
     }
     break;
   }
+  case DW_FORM_GNU_str_index: {
+    OS << format(" indexed (%8.8x) string = ", (uint32_t)uvalue);
+    const char *dbg_str = getIndirectCString(&debug_str_data,
+                                             &debug_str_offset_data);
+    if (dbg_str) {
+      OS << '"';
+      OS.write_escaped(dbg_str);
+      OS << '"';
+    }
+    break;
+  }
   case DW_FORM_ref_addr:
     OS << format("0x%016" PRIx64, uvalue);
     break;
@@ -436,6 +446,16 @@ DWARFFormValue::getAsCString(const DataExtractor *debug_str_data_ptr) const {
   return NULL;
 }
 
+const char*
+DWARFFormValue::getIndirectCString(const DataExtractor *DS,
+                                   const DataExtractor *DSO) const {
+  if (!DS || !DSO) return NULL;
+
+  uint32_t offset = Value.uval * 4;
+  uint32_t soffset = DSO->getULEB128(&offset);
+  return DS->getCStr(&soffset);
+}
+
 uint64_t DWARFFormValue::getReference(const DWARFCompileUnit *cu) const {
   uint64_t die_offset = Value.uval;
   switch (Form) {
diff --git a/lib/DebugInfo/DWARFFormValue.h b/lib/DebugInfo/DWARFFormValue.h
index c5b590d..7768c18 100644
--- a/lib/DebugInfo/DWARFFormValue.h
+++ b/lib/DebugInfo/DWARFFormValue.h
@@ -64,6 +64,8 @@ public:
   uint64_t getUnsigned() const { return Value.uval; }
   int64_t getSigned() const { return Value.sval; }
   const char *getAsCString(const DataExtractor *debug_str_data_ptr) const;
+  const char *getIndirectCString(const DataExtractor *,
+                                 const DataExtractor *) const;
   bool skipValue(DataExtractor debug_info_data, uint32_t *offset_ptr,
                  const DWARFCompileUnit *cu) const;
   static bool skipValue(uint16_t form, DataExtractor debug_info_data,
diff --git a/test/DebugInfo/X86/fission-cu.ll b/test/DebugInfo/X86/fission-cu.ll
index 22f59dd..3ada3ef 100644
--- a/test/DebugInfo/X86/fission-cu.ll
+++ b/test/DebugInfo/X86/fission-cu.ll
@@ -30,10 +30,10 @@
 ; FIXME: Strings will ultimately be a different form.
 ; CHECK: .debug_info.dwo contents:
 ; CHECK: DW_TAG_compile_unit
-; CHECK: DW_AT_producer [DW_FORM_strp]
+; CHECK: DW_AT_producer [DW_FORM_GNU_str_index] ( indexed (00000000) string = "clang version 3.3 (trunk 169021) (llvm/trunk 169020)")
 ; CHECK: DW_AT_language [DW_FORM_data2]        (0x000c)
-; CHECK: DW_AT_name [DW_FORM_strp]             ( .debug_str[0x00000035] = "baz.c")
+; CHECK: DW_AT_name [DW_FORM_GNU_str_index]    ( indexed (00000001) string = "baz.c")
 ; CHECK: DW_TAG_base_type
-; CHECK: DW_AT_name [DW_FORM_strp]      ( .debug_str[0x00000061] = "int")
+; CHECK: DW_AT_name [DW_FORM_GNU_str_index]     ( indexed (00000004) string = "int")
 ; CHECK: DW_TAG_variable
-; CHECK: DW_AT_name [DW_FORM_strp]     ( .debug_str[0x0000005f] = "a")
+; CHECK: DW_AT_name [DW_FORM_GNU_str_index]     ( indexed (00000003) string = "a")