[readobj] Expand CodeView dumping functionality

This rewrites and expands the existing codeview dumping functionality in
llvm-readobj using techniques similar to those in lib/Object. This defines a
number of new records and enums useful for reading memory mapped codeview
sections in COFF objects.

The dumper is intended as a testing tool for LLVM as it grows more codeview
output capabilities.

Reviewers: majnemer

Differential Revision: http://reviews.llvm.org/D16104

llvm-svn: 257658

llvm/include/llvm/DebugInfo/CodeView/CodeView.h

@@ -15,6 +15,97 @@
 namespace llvm {
 namespace codeview {
 
+/// These values correspond to the CV_CPU_TYPE_e enumeration, and are documented
+/// here: https://msdn.microsoft.com/en-us/library/b2fc64ek.aspx
+enum class CPUType : uint16_t {
+  Intel8080 = 0x0,
+  Intel8086 = 0x1,
+  Intel80286 = 0x2,
+  Intel80386 = 0x3,
+  Intel80486 = 0x4,
+  Pentium = 0x5,
+  PentiumPro = 0x6,
+  Pentium3 = 0x7,
+  MIPS = 0x10,
+  MIPS16 = 0x11,
+  MIPS32 = 0x12,
+  MIPS64 = 0x13,
+  MIPSI = 0x14,
+  MIPSII = 0x15,
+  MIPSIII = 0x16,
+  MIPSIV = 0x17,
+  MIPSV = 0x18,
+  M68000 = 0x20,
+  M68010 = 0x21,
+  M68020 = 0x22,
+  M68030 = 0x23,
+  M68040 = 0x24,
+  Alpha = 0x30,
+  Alpha21164 = 0x31,
+  Alpha21164A = 0x32,
+  Alpha21264 = 0x33,
+  Alpha21364 = 0x34,
+  PPC601 = 0x40,
+  PPC603 = 0x41,
+  PPC604 = 0x42,
+  PPC620 = 0x43,
+  PPCFP = 0x44,
+  PPCBE = 0x45,
+  SH3 = 0x50,
+  SH3E = 0x51,
+  SH3DSP = 0x52,
+  SH4 = 0x53,
+  SHMedia = 0x54,
+  ARM3 = 0x60,
+  ARM4 = 0x61,
+  ARM4T = 0x62,
+  ARM5 = 0x63,
+  ARM5T = 0x64,
+  ARM6 = 0x65,
+  ARM_XMAC = 0x66,
+  ARM_WMMX = 0x67,
+  ARM7 = 0x68,
+  Omni = 0x70,
+  Ia64 = 0x80,
+  Ia64_2 = 0x81,
+  CEE = 0x90,
+  AM33 = 0xa0,
+  M32R = 0xb0,
+  TriCore = 0xc0,
+  X64 = 0xd0,
+  EBC = 0xe0,
+  Thumb = 0xf0,
+  ARMNT = 0xf4,
+  D3D11_Shader = 0x100,
+};
+
+/// These values correspond to the CV_CFL_LANG enumeration, and are documented
+/// here: https://msdn.microsoft.com/en-us/library/bw3aekw6.aspx
+enum SourceLanguage : uint8_t {
+  C = 0x00,
+  Cpp = 0x01,
+  Fortran = 0x02,
+  Masm = 0x03,
+  Pascal = 0x04,
+  Basic = 0x05,
+  Cobol = 0x06,
+  Link = 0x07,
+  Cvtres = 0x08,
+  Cvtpgd = 0x09,
+  CSharp = 0x0a,
+  VB = 0x0b,
+  ILAsm = 0x0c,
+  Java = 0x0d,
+  JScript = 0x0e,
+  MSIL = 0x0f,
+  HLSL = 0x10
+};
+
+/// These values correspond to the CV_call_e enumeration, and are documented
+/// at the following locations:
+///   https://msdn.microsoft.com/en-us/library/b2fc64ek.aspx
+///   https://msdn.microsoft.com/en-us/library/windows/desktop/ms680207(v=vs.85).aspx
+///
 enum class CallingConvention : uint8_t {
   NearC = 0x00,       // near right to left push, caller pops stack
   FarC = 0x01,        // far right to left push, caller pops stack
@@ -140,13 +231,15 @@ enum class HfaKind : uint8_t {
   Other = 0x03
 };
 
+/// Source-level access specifier. (CV_access_e)
 enum class MemberAccess : uint8_t {
   None = 0,
   Private = 1,
   Protected = 2,
   Public = 3
 };
 
+/// Part of member attribute flags. (CV_methodprop_e)
 enum class MethodKind : uint8_t {
   Vanilla = 0x00,
   Virtual = 0x01,
@@ -157,9 +250,14 @@ enum class MethodKind : uint8_t {
   PureIntroducingVirtual = 0x06
 };
 
+/// Equivalent to CV_fldattr_t bitfield.
 enum class MethodOptions : uint16_t {
   None = 0x0000,
+  AccessMask = 0x0003,
+  MethodKindMask = 0x001c,
   Pseudo = 0x0020,
+  NoInherit = 0x0040,
+  NoConstruct = 0x0080,
   CompilerGenerated = 0x0100,
   Sealed = 0x0200
 };
@@ -178,6 +276,7 @@ inline MethodOptions operator~(MethodOptions a) {
   return static_cast<MethodOptions>(~static_cast<uint16_t>(a));
 }
 
+/// Equivalent to CV_modifier_t.
 enum class ModifierOptions : uint16_t {
   None = 0x0000,
   Const = 0x0001,
@@ -207,9 +306,18 @@ enum class ModuleSubstreamKind : uint32_t {
   FrameData = 0xf5,
   InlineeLines = 0xf6,
   CrossScopeImports = 0xf7,
-  CrossScopeExports = 0xf8
+  CrossScopeExports = 0xf8,
+
+  // These appear to relate to .Net assembly info.
+  ILLines = 0xf9,
+  FuncMDTokenMap = 0xfa,
+  TypeMDTokenMap = 0xfb,
+  MergedAssemblyInput = 0xfc,
+
+  CoffSymbolRVA = 0xfd,
 };
 
+/// Equivalent to CV_ptrtype_e.
 enum class PointerKind : uint8_t {
   Near16 = 0x00,                // 16 bit pointer
   Far16 = 0x01,                 // 16:16 far pointer
@@ -226,6 +334,7 @@ enum class PointerKind : uint8_t {
   Near64 = 0x0c                 // 64 bit pointer
 };
 
+/// Equivalent to CV_ptrmode_e.
 enum class PointerMode : uint8_t {
   Pointer = 0x00,                 // "normal" pointer
   LValueReference = 0x01,         // "old" reference
@@ -234,6 +343,7 @@ enum class PointerMode : uint8_t {
   RValueReference = 0x04          // r-value reference
 };
 
+/// Equivalent to misc lfPointerAttr bitfields.
 enum class PointerOptions : uint32_t {
   None = 0x00000000,
   Flat32 = 0x00000100,
@@ -258,6 +368,7 @@ inline PointerOptions operator~(PointerOptions a) {
   return static_cast<PointerOptions>(~static_cast<uint16_t>(a));
 }
 
+/// Equivalent to CV_pmtype_e.
 enum class PointerToMemberRepresentation : uint16_t {
   Unknown = 0x00,                     // not specified (pre VC8)
   SingleInheritanceData = 0x01,       // member data, single inheritance

llvm/include/llvm/DebugInfo/CodeView/TypeIndex.h

@@ -74,6 +74,9 @@ enum class SimpleTypeMode : uint32_t {
   NearPointer128 = 0x00000700 // 128 bit near pointer
 };
 
+/// A 32-bit type reference. Types are indexed by their order of appearance in
+/// .debug$T plus 0x1000. Type indices less than 0x1000 are "simple" types,
+/// composed of a SimpleTypeMode byte followed by a SimpleTypeKind byte.
 class TypeIndex {
 public:
   static const uint32_t FirstNonSimpleIndex = 0x1000;
@@ -91,6 +94,8 @@ class TypeIndex {
   uint32_t getIndex() const { return Index; }
   bool isSimple() const { return Index < FirstNonSimpleIndex; }
 
+  bool isNoType() const { return Index == 0; }
+
   SimpleTypeKind getSimpleKind() const {
     assert(isSimple());
     return static_cast<SimpleTypeKind>(Index & SimpleKindMask);
@@ -144,7 +149,7 @@ class TypeIndex {
   static TypeIndex Float64() { return TypeIndex(SimpleTypeKind::Float64); }
 
 private:
-  uint32_t Index;
+  ulittle32_t Index;
 };
 
 inline bool operator==(const TypeIndex &A, const TypeIndex &B) {

llvm/include/llvm/DebugInfo/PDB/PDBTypes.h

@@ -11,6 +11,7 @@
 #define LLVM_DEBUGINFO_PDB_PDBTYPES_H
 
 #include "llvm/Config/llvm-config.h"
+#include "llvm/DebugInfo/CodeView/CodeView.h"
 #include "llvm/Support/Endian.h"
 #include <functional>
 #include <stdint.h>
@@ -108,67 +109,7 @@ enum class PDB_Checksum { None = 0, MD5 = 1, SHA1 = 2 };
 
 /// These values correspond to the CV_CPU_TYPE_e enumeration, and are documented
 /// here: https://msdn.microsoft.com/en-us/library/b2fc64ek.aspx
-enum class PDB_Cpu {
-  Intel8080 = 0x0,
-  Intel8086 = 0x1,
-  Intel80286 = 0x2,
-  Intel80386 = 0x3,
-  Intel80486 = 0x4,
-  Pentium = 0x5,
-  PentiumPro = 0x6,
-  Pentium3 = 0x7,
-  MIPS = 0x10,
-  MIPS16 = 0x11,
-  MIPS32 = 0x12,
-  MIPS64 = 0x13,
-  MIPSI = 0x14,
-  MIPSII = 0x15,
-  MIPSIII = 0x16,
-  MIPSIV = 0x17,
-  MIPSV = 0x18,
-  M68000 = 0x20,
-  M68010 = 0x21,
-  M68020 = 0x22,
-  M68030 = 0x23,
-  M68040 = 0x24,
-  Alpha = 0x30,
-  Alpha21164 = 0x31,
-  Alpha21164A = 0x32,
-  Alpha21264 = 0x33,
-  Alpha21364 = 0x34,
-  PPC601 = 0x40,
-  PPC603 = 0x41,
-  PPC604 = 0x42,
-  PPC620 = 0x43,
-  PPCFP = 0x44,
-  PPCBE = 0x45,
-  SH3 = 0x50,
-  SH3E = 0x51,
-  SH3DSP = 0x52,
-  SH4 = 0x53,
-  SHMedia = 0x54,
-  ARM3 = 0x60,
-  ARM4 = 0x61,
-  ARM4T = 0x62,
-  ARM5 = 0x63,
-  ARM5T = 0x64,
-  ARM6 = 0x65,
-  ARM_XMAC = 0x66,
-  ARM_WMMX = 0x67,
-  ARM7 = 0x68,
-  Omni = 0x70,
-  Ia64 = 0x80,
-  Ia64_2 = 0x81,
-  CEE = 0x90,
-  AM33 = 0xa0,
-  M32R = 0xb0,
-  TriCore = 0xc0,
-  X64 = 0xd0,
-  EBC = 0xe0,
-  Thumb = 0xf0,
-  ARMNT = 0xf4,
-  D3D11_Shader = 0x100,
-};
+typedef codeview::CPUType PDB_Cpu;
 
 enum class PDB_Machine {
   Invalid = 0xffff,
@@ -200,56 +141,11 @@ enum class PDB_Machine {
 ///   https://msdn.microsoft.com/en-us/library/b2fc64ek.aspx
 ///   https://msdn.microsoft.com/en-us/library/windows/desktop/ms680207(v=vs.85).aspx
 ///
-enum class PDB_CallingConv {
-  NearCdecl = 0x00,
-  FarCdecl = 0x01,
-  NearPascal = 0x02,
-  FarPascal = 0x03,
-  NearFastcall = 0x04,
-  FarFastcall = 0x05,
-  Skipped = 0x06,
-  NearStdcall = 0x07,
-  FarStdcall = 0x08,
-  NearSyscall = 0x09,
-  FarSyscall = 0x0a,
-  Thiscall = 0x0b,
-  MipsCall = 0x0c,
-  Generic = 0x0d,
-  Alphacall = 0x0e,
-  Ppccall = 0x0f,
-  SuperHCall = 0x10,
-  Armcall = 0x11,
-  AM33call = 0x12,
-  Tricall = 0x13,
-  Sh5call = 0x14,
-  M32R = 0x15,
-  Clrcall = 0x16,
-  Inline = 0x17,
-  NearVectorcall = 0x18,
-  Reserved = 0x19,
-};
+typedef codeview::CallingConvention PDB_CallingConv;
 
 /// These values correspond to the CV_CFL_LANG enumeration, and are documented
 /// here: https://msdn.microsoft.com/en-us/library/bw3aekw6.aspx
-enum class PDB_Lang {
-  C = 0x00,
-  Cpp = 0x01,
-  Fortran = 0x02,
-  Masm = 0x03,
-  Pascal = 0x04,
-  Basic = 0x05,
-  Cobol = 0x06,
-  Link = 0x07,
-  Cvtres = 0x08,
-  Cvtpgd = 0x09,
-  CSharp = 0x0a,
-  VB = 0x0b,
-  ILAsm = 0x0c,
-  Java = 0x0d,
-  JScript = 0x0e,
-  MSIL = 0x0f,
-  HLSL = 0x10
-};
+typedef codeview::SourceLanguage PDB_Lang;
 
 /// These values correspond to the DataKind enumeration, and are documented
 /// here: https://msdn.microsoft.com/en-us/library/b2x2t313.aspx

llvm/include/llvm/Support/COFF.h

@@ -674,6 +674,7 @@ namespace COFF {
     DEBUG_INDEX_SUBSECTION = 0xF4,
 
     // Symbol subsections are split into records of different types.
+    DEBUG_SYMBOL_TYPE_LOCAL_PROC_START = 0x1146,
     DEBUG_SYMBOL_TYPE_PROC_START = 0x1147,
     DEBUG_SYMBOL_TYPE_PROC_END = 0x114F
   };

llvm/include/llvm/Support/Endian.h

@@ -173,6 +173,10 @@ template<typename value_type,
          endianness endian,
          std::size_t alignment>
 struct packed_endian_specific_integral {
+  packed_endian_specific_integral() = default;
+
+  explicit packed_endian_specific_integral(value_type val) { *this = val; }
+
   operator value_type() const {
     return endian::read<value_type, endian, alignment>(
       (const void*)Value.buffer);

llvm/lib/CodeGen/AsmPrinter/WinCodeViewLineTables.cpp

@@ -199,7 +199,6 @@ void WinCodeViewLineTables::emitDebugInfoForFunction(const Function *GV) {
     return;
   assert(FI.End && "Don't know where the function ends?");
 
-  StringRef GVName = GV->getName();
   StringRef FuncName;
   if (auto *SP = getDISubprogram(GV))
     FuncName = SP->getDisplayName();
@@ -208,8 +207,8 @@ void WinCodeViewLineTables::emitDebugInfoForFunction(const Function *GV) {
   // "namespace_foo::bar" function, see PR21528.  Luckily, dbghelp.dll is trying
   // to demangle display names anyways, so let's just put a mangled name into
   // the symbols subsection until Clang gives us what we need.
-  if (GVName.startswith("\01?"))
-    FuncName = GVName.substr(1);
+  if (FuncName.empty())
+    FuncName = GlobalValue::getRealLinkageName(GV->getName());
   // Emit a symbol subsection, required by VS2012+ to find function boundaries.
   MCSymbol *SymbolsBegin = Asm->MMI->getContext().createTempSymbol(),
            *SymbolsEnd = Asm->MMI->getContext().createTempSymbol();