Object: Provide a richer means of describing auxiliary symbols

The current state of affairs has auxiliary symbols described as a big
bag of bytes. This is less than satisfying, it detracts from the YAML
file as being human readable.

Instead, allow for symbols to optionally contain their auxiliary data.
This allows us to have a much higher level way of describing things like
weak symbols, function definitions and section definitions.

This depends on D3105.

Differential Revision: http://llvm-reviews.chandlerc.com/D3092

llvm-svn: 204214

5 files changed, 192 insertions, 31 deletions

diff --git a/llvm/tools/llvm-nm/llvm-nm.cpp b/llvm/tools/llvm-nm/llvm-nm.cpp
index 100a1f20a11..22e019a8a52 100644
--- a/llvm/tools/llvm-nm/llvm-nm.cpp
+++ b/llvm/tools/llvm-nm/llvm-nm.cpp
@@ -344,8 +344,7 @@ static char getSymbolNMTypeChar(COFFObjectFile &Obj, symbol_iterator I) {
       return 'i';
 
     // Check for section symbol.
-    else if (Symb->StorageClass == COFF::IMAGE_SYM_CLASS_STATIC &&
-             Symb->Value == 0)
+    else if (Symb->isSectionDefinition())
       return 's';
   }
 
diff --git a/llvm/tools/llvm-objdump/llvm-objdump.cpp b/llvm/tools/llvm-objdump/llvm-objdump.cpp
index 3f871471bca..f5328a959e9 100644
--- a/llvm/tools/llvm-objdump/llvm-objdump.cpp
+++ b/llvm/tools/llvm-objdump/llvm-objdump.cpp
@@ -663,8 +663,7 @@ static void PrintCOFFSymbolTable(const COFFObjectFile *coff) {
   for (int i = 0, e = header->NumberOfSymbols; i != e; ++i) {
     if (aux_count--) {
       // Figure out which type of aux this is.
-      if (symbol->StorageClass == COFF::IMAGE_SYM_CLASS_STATIC
-          && symbol->Value == 0) { // Section definition.
+      if (symbol->isSectionDefinition()) { // Section definition.
         const coff_aux_section_definition *asd;
         if (error(coff->getAuxSymbol<coff_aux_section_definition>(i, asd)))
           return;
diff --git a/llvm/tools/llvm-readobj/COFFDumper.cpp b/llvm/tools/llvm-readobj/COFFDumper.cpp
index d2275f0aab8..cd40da7d6cc 100644
--- a/llvm/tools/llvm-readobj/COFFDumper.cpp
+++ b/llvm/tools/llvm-readobj/COFFDumper.cpp
@@ -946,12 +946,7 @@ void COFFDumper::printSymbol(const SymbolRef &Sym) {
   W.printNumber("AuxSymbolCount", Symbol->NumberOfAuxSymbols);
 
   for (unsigned I = 0; I < Symbol->NumberOfAuxSymbols; ++I) {
-    if (Symbol->StorageClass     == COFF::IMAGE_SYM_CLASS_EXTERNAL &&
-        Symbol->getBaseType()    == COFF::IMAGE_SYM_TYPE_NULL &&
-        Symbol->getComplexType() == COFF::IMAGE_SYM_DTYPE_FUNCTION &&
-        Symbol->SectionNumber != COFF::IMAGE_SYM_DEBUG &&
-        Symbol->SectionNumber != COFF::IMAGE_SYM_ABSOLUTE &&
-        Symbol->SectionNumber != COFF::IMAGE_SYM_UNDEFINED) {
+    if (Symbol->isFunctionDefinition()) {
       const coff_aux_function_definition *Aux;
       if (error(getSymbolAuxData(Obj, Symbol + I, Aux)))
         break;
@@ -963,11 +958,7 @@ void COFFDumper::printSymbol(const SymbolRef &Sym) {
       W.printHex("PointerToNextFunction", Aux->PointerToNextFunction);
       W.printBinary("Unused", makeArrayRef(Aux->Unused));
 
-    } else if (
-        Symbol->StorageClass   == COFF::IMAGE_SYM_CLASS_WEAK_EXTERNAL ||
-        (Symbol->StorageClass  == COFF::IMAGE_SYM_CLASS_EXTERNAL &&
-         Symbol->SectionNumber == COFF::IMAGE_SYM_UNDEFINED &&
-         Symbol->Value         == 0)) {
+    } else if (Symbol->isWeakExternal()) {
       const coff_aux_weak_external *Aux;
       if (error(getSymbolAuxData(Obj, Symbol + I, Aux)))
         break;
@@ -987,7 +978,7 @@ void COFFDumper::printSymbol(const SymbolRef &Sym) {
                     makeArrayRef(WeakExternalCharacteristics));
       W.printBinary("Unused", makeArrayRef(Aux->Unused));
 
-    } else if (Symbol->StorageClass == COFF::IMAGE_SYM_CLASS_FILE) {
+    } else if (Symbol->isFileRecord()) {
       const coff_aux_file_record *Aux;
       if (error(getSymbolAuxData(Obj, Symbol + I, Aux)))
         break;
@@ -995,11 +986,7 @@ void COFFDumper::printSymbol(const SymbolRef &Sym) {
       DictScope AS(W, "AuxFileRecord");
       W.printString("FileName", StringRef(Aux->FileName));
 
-    // C++/CLI creates external ABS symbols for non-const appdomain globals.
-    // These are also followed by an auxiliary section definition.
-    } else if (Symbol->StorageClass == COFF::IMAGE_SYM_CLASS_STATIC ||
-               (Symbol->StorageClass == COFF::IMAGE_SYM_CLASS_EXTERNAL &&
-                Symbol->SectionNumber == COFF::IMAGE_SYM_ABSOLUTE)) {
+    } else if (Symbol->isSectionDefinition()) {
       const coff_aux_section_definition *Aux;
       if (error(getSymbolAuxData(Obj, Symbol + I, Aux)))
         break;
@@ -1026,7 +1013,7 @@ void COFFDumper::printSymbol(const SymbolRef &Sym) {
 
         W.printNumber("AssocSection", AssocName, Aux->Number);
       }
-    } else if (Symbol->StorageClass == COFF::IMAGE_SYM_CLASS_CLR_TOKEN) {
+    } else if (Symbol->isCLRToken()) {
       const coff_aux_clr_token *Aux;
       if (error(getSymbolAuxData(Obj, Symbol + I, Aux)))
         break;
diff --git a/llvm/tools/obj2yaml/coff2yaml.cpp b/llvm/tools/obj2yaml/coff2yaml.cpp
index 6cf79aeabc7..b003029407c 100644
--- a/llvm/tools/obj2yaml/coff2yaml.cpp
+++ b/llvm/tools/obj2yaml/coff2yaml.cpp
@@ -78,6 +78,61 @@ void COFFDumper::dumpSections(unsigned NumSections) {
   }
 }
 
+static void
+dumpFunctionDefinition(COFFYAML::Symbol *Sym,
+                       const object::coff_aux_function_definition *ObjFD) {
+  COFF::AuxiliaryFunctionDefinition YAMLFD;
+  YAMLFD.TagIndex = ObjFD->TagIndex;
+  YAMLFD.TotalSize = ObjFD->TotalSize;
+  YAMLFD.PointerToLinenumber = ObjFD->PointerToLinenumber;
+  YAMLFD.PointerToNextFunction = ObjFD->PointerToNextFunction;
+
+  Sym->FunctionDefinition = YAMLFD;
+}
+
+static void
+dumpbfAndEfLineInfo(COFFYAML::Symbol *Sym,
+                    const object::coff_aux_bf_and_ef_symbol *ObjBES) {
+  COFF::AuxiliarybfAndefSymbol YAMLAAS;
+  YAMLAAS.Linenumber = ObjBES->Linenumber;
+  YAMLAAS.PointerToNextFunction = ObjBES->PointerToNextFunction;
+
+  Sym->bfAndefSymbol = YAMLAAS;
+}
+
+static void dumpWeakExternal(COFFYAML::Symbol *Sym,
+                             const object::coff_aux_weak_external *ObjWE) {
+  COFF::AuxiliaryWeakExternal YAMLWE;
+  YAMLWE.TagIndex = ObjWE->TagIndex;
+  YAMLWE.Characteristics = ObjWE->Characteristics;
+
+  Sym->WeakExternal = YAMLWE;
+}
+
+static void
+dumpSectionDefinition(COFFYAML::Symbol *Sym,
+                      const object::coff_aux_section_definition *ObjSD) {
+  COFF::AuxiliarySectionDefinition YAMLASD;
+  YAMLASD.Length = ObjSD->Length;
+  YAMLASD.NumberOfRelocations = ObjSD->NumberOfRelocations;
+  YAMLASD.NumberOfLinenumbers = ObjSD->NumberOfLinenumbers;
+  YAMLASD.CheckSum = ObjSD->CheckSum;
+  YAMLASD.Number = ObjSD->Number;
+  YAMLASD.Selection = ObjSD->Selection;
+
+  Sym->SectionDefinition = YAMLASD;
+}
+
+static void
+dumpCLRTokenDefinition(COFFYAML::Symbol *Sym,
+                       const object::coff_aux_clr_token *ObjCLRToken) {
+  COFF::AuxiliaryCLRToken YAMLCLRToken;
+  YAMLCLRToken.AuxType = ObjCLRToken->AuxType;
+  YAMLCLRToken.SymbolTableIndex = ObjCLRToken->SymbolTableIndex;
+
+  Sym->CLRToken = YAMLCLRToken;
+}
+
 void COFFDumper::dumpSymbols(unsigned NumSymbols) {
   std::vector<COFFYAML::Symbol> &Symbols = YAMLObj.Symbols;
   for (const auto &S : Obj.symbols()) {
@@ -90,7 +145,62 @@ void COFFDumper::dumpSymbols(unsigned NumSymbols) {
     Sym.Header.Value = Symbol->Value;
     Sym.Header.SectionNumber = Symbol->SectionNumber;
     Sym.Header.NumberOfAuxSymbols = Symbol->NumberOfAuxSymbols;
-    Sym.AuxiliaryData = object::yaml::BinaryRef(Obj.getSymbolAuxData(Symbol));
+
+    if (Symbol->NumberOfAuxSymbols > 0) {
+      ArrayRef<uint8_t> AuxData = Obj.getSymbolAuxData(Symbol);
+      if (Symbol->isFunctionDefinition()) {
+        // This symbol represents a function definition.
+        assert(Symbol->NumberOfAuxSymbols == 1 &&
+               "Expected a single aux symbol to describe this function!");
+
+        const object::coff_aux_function_definition *ObjFD =
+            reinterpret_cast<const object::coff_aux_function_definition *>(
+                AuxData.data());
+        dumpFunctionDefinition(&Sym, ObjFD);
+      } else if (Symbol->StorageClass == COFF::IMAGE_SYM_CLASS_FUNCTION) {
+        // This symbol describes function line number information.
+        assert(Symbol->NumberOfAuxSymbols == 1 &&
+               "Exepected a single aux symbol to describe this section!");
+
+        const object::coff_aux_bf_and_ef_symbol *ObjBES =
+            reinterpret_cast<const object::coff_aux_bf_and_ef_symbol *>(
+                AuxData.data());
+        dumpbfAndEfLineInfo(&Sym, ObjBES);
+      } else if (Symbol->isWeakExternal()) {
+        // This symbol represents a weak external definition.
+        assert(Symbol->NumberOfAuxSymbols == 1 &&
+               "Exepected a single aux symbol to describe this section!");
+
+        const object::coff_aux_weak_external *ObjWE =
+            reinterpret_cast<const object::coff_aux_weak_external *>(
+                AuxData.data());
+        dumpWeakExternal(&Sym, ObjWE);
+      } else if (Symbol->isFileRecord()) {
+        // This symbol represents a file record.
+        Sym.File = StringRef(reinterpret_cast<const char *>(AuxData.data()),
+                             Symbol->NumberOfAuxSymbols * COFF::SymbolSize);
+      } else if (Symbol->isSectionDefinition()) {
+        // This symbol represents a section definition.
+        assert(Symbol->NumberOfAuxSymbols == 1 &&
+               "Expected a single aux symbol to describe this section!");
+
+        const object::coff_aux_section_definition *ObjSD =
+            reinterpret_cast<const object::coff_aux_section_definition *>(
+                AuxData.data());
+        dumpSectionDefinition(&Sym, ObjSD);
+      } else if (Symbol->isCLRToken()) {
+        // This symbol represents a CLR token definition.
+        assert(Symbol->NumberOfAuxSymbols == 1 &&
+               "Expected a single aux symbol to describe this CLR Token");
+
+        const object::coff_aux_clr_token *ObjCLRToken =
+            reinterpret_cast<const object::coff_aux_clr_token *>(
+                AuxData.data());
+        dumpCLRTokenDefinition(&Sym, ObjCLRToken);
+      } else {
+        llvm_unreachable("Unhandled auxiliary symbol!");
+      }
+    }
     Symbols.push_back(Sym);
   }
 }
diff --git a/llvm/tools/yaml2obj/yaml2coff.cpp b/llvm/tools/yaml2obj/yaml2coff.cpp
index c757eb64fb7..a0ede246bd6 100644
--- a/llvm/tools/yaml2obj/yaml2coff.cpp
+++ b/llvm/tools/yaml2obj/yaml2coff.cpp
@@ -14,6 +14,7 @@
 
 #include "yaml2obj.h"
 #include "llvm/ADT/SmallString.h"
+#include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/ADT/StringSwitch.h"
@@ -153,13 +154,22 @@ static bool layoutCOFF(COFFParser &CP) {
   for (std::vector<COFFYAML::Symbol>::iterator i = CP.Obj.Symbols.begin(),
                                                e = CP.Obj.Symbols.end();
                                                i != e; ++i) {
-    unsigned AuxBytes = i->AuxiliaryData.binary_size();
-    if (AuxBytes % COFF::SymbolSize != 0) {
-      errs() << "AuxiliaryData size not a multiple of symbol size!\n";
-      return false;
-    }
-    i->Header.NumberOfAuxSymbols = AuxBytes / COFF::SymbolSize;
-    NumberOfSymbols += 1 + i->Header.NumberOfAuxSymbols;
+    uint32_t NumberOfAuxSymbols = 0;
+    if (i->FunctionDefinition)
+      NumberOfAuxSymbols += 1;
+    if (i->bfAndefSymbol)
+      NumberOfAuxSymbols += 1;
+    if (i->WeakExternal)
+      NumberOfAuxSymbols += 1;
+    if (!i->File.empty())
+      NumberOfAuxSymbols +=
+          (i->File.size() + COFF::SymbolSize - 1) / COFF::SymbolSize;
+    if (i->SectionDefinition)
+      NumberOfAuxSymbols += 1;
+    if (i->CLRToken)
+      NumberOfAuxSymbols += 1;
+    i->Header.NumberOfAuxSymbols = NumberOfAuxSymbols;
+    NumberOfSymbols += 1 + NumberOfAuxSymbols;
   }
 
   // Store all the allocated start addresses in the header.
@@ -194,6 +204,24 @@ binary_le_impl<value_type> binary_le(value_type V) {
   return binary_le_impl<value_type>(V);
 }
 
+template <size_t NumBytes>
+struct zeros_impl {
+  zeros_impl() {}
+};
+
+template <size_t NumBytes>
+raw_ostream &operator<<(raw_ostream &OS, const zeros_impl<NumBytes> &) {
+  char Buffer[NumBytes];
+  memset(Buffer, 0, sizeof(Buffer));
+  OS.write(Buffer, sizeof(Buffer));
+  return OS;
+}
+
+template <typename T>
+zeros_impl<sizeof(T)> zeros(const T &) {
+  return zeros_impl<sizeof(T)>();
+}
+
 bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
   OS << binary_le(CP.Obj.Header.Machine)
      << binary_le(CP.Obj.Header.NumberOfSections)
@@ -253,7 +281,45 @@ bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
        << binary_le(i->Header.Type)
        << binary_le(i->Header.StorageClass)
        << binary_le(i->Header.NumberOfAuxSymbols);
-    i->AuxiliaryData.writeAsBinary(OS);
+
+    if (i->FunctionDefinition)
+      OS << binary_le(i->FunctionDefinition->TagIndex)
+         << binary_le(i->FunctionDefinition->TotalSize)
+         << binary_le(i->FunctionDefinition->PointerToLinenumber)
+         << binary_le(i->FunctionDefinition->PointerToNextFunction)
+         << zeros(i->FunctionDefinition->unused);
+    if (i->bfAndefSymbol)
+      OS << zeros(i->bfAndefSymbol->unused1)
+         << binary_le(i->bfAndefSymbol->Linenumber)
+         << zeros(i->bfAndefSymbol->unused2)
+         << binary_le(i->bfAndefSymbol->PointerToNextFunction)
+         << zeros(i->bfAndefSymbol->unused3);
+    if (i->WeakExternal)
+      OS << binary_le(i->WeakExternal->TagIndex)
+         << binary_le(i->WeakExternal->Characteristics)
+         << zeros(i->WeakExternal->unused);
+    if (!i->File.empty()) {
+      uint32_t NumberOfAuxRecords =
+          (i->File.size() + COFF::SymbolSize - 1) / COFF::SymbolSize;
+      uint32_t NumberOfAuxBytes = NumberOfAuxRecords * COFF::SymbolSize;
+      uint32_t NumZeros = NumberOfAuxBytes - i->File.size();
+      OS.write(i->File.data(), i->File.size());
+      for (uint32_t Padding = 0; Padding < NumZeros; ++Padding)
+        OS.write(0);
+    }
+    if (i->SectionDefinition)
+      OS << binary_le(i->SectionDefinition->Length)
+         << binary_le(i->SectionDefinition->NumberOfRelocations)
+         << binary_le(i->SectionDefinition->NumberOfLinenumbers)
+         << binary_le(i->SectionDefinition->CheckSum)
+         << binary_le(i->SectionDefinition->Number)
+         << binary_le(i->SectionDefinition->Selection)
+         << zeros(i->SectionDefinition->unused);
+    if (i->CLRToken)
+      OS << binary_le(i->CLRToken->AuxType)
+         << zeros(i->CLRToken->unused1)
+         << binary_le(i->CLRToken->SymbolTableIndex)
+         << zeros(i->CLRToken->unused2);
   }
 
   // Output string table.