yaml2obj: Support bigobj

Teach yaml2obj how to make a bigobj COFF file.  Like the rest of LLVM,
we automatically decide whether or not to use regular COFF or bigobj
COFF on the fly depending on how many sections the resulting object
would have.

This ends the task of adding bigobj support to LLVM.

N.B. This was tested by forcing yaml2obj to be used in bigobj mode
regardless of the number of sections.  While a dedicated test was
written, the smallest I could make it was 36 MB (!) of yaml and it still
took a significant amount of time to execute on a powerful machine.

llvm-svn: 217858

include/llvm/Support/COFF.h

@@ -63,7 +63,7 @@ namespace COFF {
   };
 
   struct BigObjHeader {
-    enum { MinBigObjectVersion = 2 };
+    enum : uint16_t { MinBigObjectVersion = 2 };
 
     uint16_t Sig1; ///< Must be IMAGE_FILE_MACHINE_UNKNOWN (0).
     uint16_t Sig2; ///< Must be 0xFFFF.

test/tools/llvm-readobj/Inputs/file-aux-record.yaml

@@ -6,7 +6,7 @@ symbols:
   - !Symbol
     Name: .file
     Value: 0
-    SectionNumber: 65534
+    SectionNumber: -2
     SimpleType: IMAGE_SYM_TYPE_NULL
     ComplexType: IMAGE_SYM_DTYPE_NULL
     StorageClass: IMAGE_SYM_CLASS_FILE
@@ -14,7 +14,7 @@ symbols:
   - !Symbol
     Name: '@comp.id'
     Value: 13485607
-    SectionNumber: 65535
+    SectionNumber: -1
     SimpleType: IMAGE_SYM_TYPE_NULL
     ComplexType: IMAGE_SYM_DTYPE_NULL
     StorageClass: IMAGE_SYM_CLASS_STATIC

test/tools/llvm-readobj/Inputs/file-multiple-aux-records.yaml

@@ -6,7 +6,7 @@ symbols:
   - !Symbol
     Name: .file
     Value: 0
-    SectionNumber: 65534
+    SectionNumber: -2
     SimpleType: IMAGE_SYM_TYPE_NULL
     ComplexType: IMAGE_SYM_DTYPE_NULL
     StorageClass: IMAGE_SYM_CLASS_FILE
@@ -14,7 +14,7 @@ symbols:
   - !Symbol
     Name: '@comp.id'
     Value: 13485607
-    SectionNumber: 65535
+    SectionNumber: -1
     SimpleType: IMAGE_SYM_TYPE_NULL
     ComplexType: IMAGE_SYM_DTYPE_NULL
     StorageClass: IMAGE_SYM_CLASS_STATIC

tools/yaml2obj/yaml2coff.cpp

@@ -36,6 +36,18 @@ struct COFFParser {
     StringTable.append(4, char(0));
   }
 
+  bool useBigObj() const {
+    return Obj.Sections.size() > COFF::MaxNumberOfSections16;
+  }
+
+  unsigned getHeaderSize() const {
+    return useBigObj() ? COFF::Header32Size : COFF::Header16Size;
+  }
+
+  unsigned getSymbolSize() const {
+    return useBigObj() ? COFF::Symbol32Size : COFF::Symbol16Size;
+  }
+
   bool parseSections() {
     for (std::vector<COFFYAML::Section>::iterator i = Obj.Sections.begin(),
            e = Obj.Sections.end(); i != e; ++i) {
@@ -121,7 +133,7 @@ static bool layoutCOFF(COFFParser &CP) {
 
   // The section table starts immediately after the header, including the
   // optional header.
-  SectionTableStart = COFF::Header16Size + CP.Obj.Header.SizeOfOptionalHeader;
+  SectionTableStart = CP.getHeaderSize() + CP.Obj.Header.SizeOfOptionalHeader;
   SectionTableSize = COFF::SectionSize * CP.Obj.Sections.size();
 
   uint32_t CurrentSectionDataOffset = SectionTableStart + SectionTableSize;
@@ -163,7 +175,7 @@ static bool layoutCOFF(COFFParser &CP) {
       NumberOfAuxSymbols += 1;
     if (!i->File.empty())
       NumberOfAuxSymbols +=
-          (i->File.size() + COFF::Symbol16Size - 1) / COFF::Symbol16Size;
+          (i->File.size() + CP.getSymbolSize() - 1) / CP.getSymbolSize();
     if (i->SectionDefinition)
       NumberOfAuxSymbols += 1;
     if (i->CLRToken)
@@ -222,14 +234,46 @@ zeros_impl<sizeof(T)> zeros(const T &) {
   return zeros_impl<sizeof(T)>();
 }
 
+struct num_zeros_impl {
+  size_t N;
+  num_zeros_impl(size_t N) : N(N) {}
+};
+
+raw_ostream &operator<<(raw_ostream &OS, const num_zeros_impl &NZI) {
+  for (size_t I = 0; I != NZI.N; ++I)
+    OS.write(0);
+  return OS;
+}
+
+num_zeros_impl num_zeros(size_t N) {
+  num_zeros_impl NZI(N);
+  return NZI;
+}
+
 bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
-  OS << binary_le(CP.Obj.Header.Machine)
+  if (CP.useBigObj()) {
-     << binary_le(static_cast<int16_t>(CP.Obj.Header.NumberOfSections))
+    OS << binary_le(static_cast<uint16_t>(COFF::IMAGE_FILE_MACHINE_UNKNOWN))
-     << binary_le(CP.Obj.Header.TimeDateStamp)
+       << binary_le(static_cast<uint16_t>(0xffff))
-     << binary_le(CP.Obj.Header.PointerToSymbolTable)
+       << binary_le(static_cast<uint16_t>(COFF::BigObjHeader::MinBigObjectVersion))
-     << binary_le(CP.Obj.Header.NumberOfSymbols)
+       << binary_le(CP.Obj.Header.Machine)
-     << binary_le(CP.Obj.Header.SizeOfOptionalHeader)
+       << binary_le(CP.Obj.Header.TimeDateStamp);
-     << binary_le(CP.Obj.Header.Characteristics);
+    OS.write(COFF::BigObjMagic, sizeof(COFF::BigObjMagic));
+    OS << zeros(uint32_t(0))
+       << zeros(uint32_t(0))
+       << zeros(uint32_t(0))
+       << zeros(uint32_t(0))
+       << binary_le(CP.Obj.Header.NumberOfSections)
+       << binary_le(CP.Obj.Header.PointerToSymbolTable)
+       << binary_le(CP.Obj.Header.NumberOfSymbols);
+  } else {
+    OS << binary_le(CP.Obj.Header.Machine)
+       << binary_le(static_cast<int16_t>(CP.Obj.Header.NumberOfSections))
+       << binary_le(CP.Obj.Header.TimeDateStamp)
+       << binary_le(CP.Obj.Header.PointerToSymbolTable)
+       << binary_le(CP.Obj.Header.NumberOfSymbols)
+       << binary_le(CP.Obj.Header.SizeOfOptionalHeader)
+       << binary_le(CP.Obj.Header.Characteristics);
+  }
 
   // Output section table.
   for (std::vector<COFFYAML::Section>::iterator i = CP.Obj.Sections.begin(),
@@ -276,9 +320,12 @@ bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
                                                      e = CP.Obj.Symbols.end();
                                                      i != e; ++i) {
     OS.write(i->Header.Name, COFF::NameSize);
-    OS << binary_le(i->Header.Value)
+    OS << binary_le(i->Header.Value);
-       << binary_le(static_cast<int16_t>(i->Header.SectionNumber))
+    if (CP.useBigObj())
-       << binary_le(i->Header.Type)
+       OS << binary_le(i->Header.SectionNumber);
+    else
+       OS << binary_le(static_cast<int16_t>(i->Header.SectionNumber));
+    OS << binary_le(i->Header.Type)
        << binary_le(i->Header.StorageClass)
        << binary_le(i->Header.NumberOfAuxSymbols);
 
@@ -287,25 +334,28 @@ bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
          << binary_le(i->FunctionDefinition->TotalSize)
          << binary_le(i->FunctionDefinition->PointerToLinenumber)
          << binary_le(i->FunctionDefinition->PointerToNextFunction)
-         << zeros(i->FunctionDefinition->unused);
+         << zeros(i->FunctionDefinition->unused)
+         << num_zeros(CP.getSymbolSize() - COFF::Symbol16Size);
     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);
+         << zeros(i->bfAndefSymbol->unused3)
+         << num_zeros(CP.getSymbolSize() - COFF::Symbol16Size);
     if (i->WeakExternal)
       OS << binary_le(i->WeakExternal->TagIndex)
          << binary_le(i->WeakExternal->Characteristics)
-         << zeros(i->WeakExternal->unused);
+         << zeros(i->WeakExternal->unused)
+         << num_zeros(CP.getSymbolSize() - COFF::Symbol16Size);
     if (!i->File.empty()) {
+      unsigned SymbolSize = CP.getSymbolSize();
       uint32_t NumberOfAuxRecords =
-          (i->File.size() + COFF::Symbol16Size - 1) / COFF::Symbol16Size;
+          (i->File.size() + SymbolSize - 1) / SymbolSize;
-      uint32_t NumberOfAuxBytes = NumberOfAuxRecords * COFF::Symbol16Size;
+      uint32_t NumberOfAuxBytes = NumberOfAuxRecords * 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 << num_zeros(NumZeros);
-        OS.write(0);
     }
     if (i->SectionDefinition)
       OS << binary_le(i->SectionDefinition->Length)
@@ -315,12 +365,14 @@ bool writeCOFF(COFFParser &CP, raw_ostream &OS) {
          << binary_le(static_cast<int16_t>(i->SectionDefinition->Number))
          << binary_le(i->SectionDefinition->Selection)
          << zeros(i->SectionDefinition->unused)
-         << binary_le(static_cast<int16_t>(i->SectionDefinition->Number >> 16));
+         << binary_le(static_cast<int16_t>(i->SectionDefinition->Number >> 16))
+         << num_zeros(CP.getSymbolSize() - COFF::Symbol16Size);
     if (i->CLRToken)
       OS << binary_le(i->CLRToken->AuxType)
          << zeros(i->CLRToken->unused1)
          << binary_le(i->CLRToken->SymbolTableIndex)
-         << zeros(i->CLRToken->unused2);
+         << zeros(i->CLRToken->unused2)
+         << num_zeros(CP.getSymbolSize() - COFF::Symbol16Size);
   }
 
   // Output string table.