[llvm-readobj] add support for macho universal binary

[chilledheart]

Summary

MachO Universal Binary is a commonly used format in OS X and iOS, e.g.
frameworks in Apple's iOS SDK. Adding support for MachO Universal Binary
makes llvm-readobj a more convenient tool for the daily use.

And because MachO Universal Binary is rather a container format than binary
format, this patch adds support for handling this format in the same way as
in Archive files, i.e., dumping required information for MachO Objects in it
one by one.

[• rafael]

tools/llvm-readobj/llvm-readobj.cpp

326

Pity we don't have ' objects() to use with range loop. Would you mind adding it?

331

Why? Do we really get universal binaries with non Object Members?

[chilledheart]

add MachOUniversalBinary::objects() method and remove unnecessary checks for non-object members

[chilledheart]

I am not sure if explicit initialization of MachOUniversalBinary::Objects is good enough. Maybe we need to change the MachOUniversalBinary::object_iterator to typedef ObjectForArch* object_iterator and related code as well to use MachOUniversalBinary::Objects directly?

tools/llvm-readobj/llvm-readobj.cpp

331

Sorry, I'll remove the dead code in the next patch.

[• rafael]

include/llvm/Object/MachOUniversal.h

117

Please don't add this. You can use iterator_range to implement objects().

[chilledheart]

use llvm::iterator_range

[chilledheart]

improve code style

[• rafael]

tools/llvm-readobj/llvm-readobj.cpp

329

This now ignores every error. Shouldn't we call reportError for every error instead?

[chilledheart]

report error if any

[chilledheart]

report error if any

include/llvm/Object/MachOUniversal.h

117

Thanks for pointing it out. It should be corrected in the next patch.

tools/llvm-readobj/llvm-readobj.cpp

329

reportError code is added back

[• rafael]

LGTM, thanks!

[chilledheart]

You're welcome.

[chilledheart]

Hi rafael,

If everything is okay, could you commit this patch for me? I don't have the commit permission to the source tree yet.

Cheers,

Chilledheart

[chilledheart]

Thank you, rafael!

[chilledheart]

"Committed as r233107"

include/llvm/Object/MachOUniversal.h

 //
 //===----------------------------------------------------------------------===//
 
 #ifndef LLVM_OBJECT_MACHOUNIVERSAL_H
 #define LLVM_OBJECT_MACHOUNIVERSAL_H
 
 #include "llvm/ADT/StringRef.h"
 #include "llvm/ADT/Triple.h"
+#include "llvm/ADT/iterator_range.h"
 #include "llvm/Object/Archive.h"
 #include "llvm/Object/Binary.h"
 #include "llvm/Object/MachO.h"
 #include "llvm/Support/ErrorOr.h"
 #include "llvm/Support/MachO.h"
 
 namespace llvm {
 namespace object {
public:
 
     ErrorOr<std::unique_ptr<Archive>> getAsArchive() const;
   };
 
   class object_iterator {
     ObjectForArch Obj;
   public:
     object_iterator(const ObjectForArch &Obj) : Obj(Obj) {}
-    const ObjectForArch* operator->() const {
-      return &Obj;
-    }
+    const ObjectForArch *operator->() const { return &Obj; }
+    const ObjectForArch &operator*() const { return Obj; }
 
     bool operator==(const object_iterator &Other) const {
       return Obj == Other.Obj;
     }
     bool operator!=(const object_iterator &Other) const {
       return !(*this == Other);
     }
 
public:
 
   object_iterator begin_objects() const {
     return ObjectForArch(this, 0);
   }
   object_iterator end_objects() const {
     return ObjectForArch(nullptr, 0);
   }
 
+  iterator_range<object_iterator> objects() const {
+    return make_range(begin_objects(), end_objects());
+  }
+
   uint32_t getNumberOfObjects() const { return NumberOfObjects; }
 
   // Cast methods.
   static inline bool classof(Binary const *V) {
     return V->isMachOUniversalBinary();
   }
 
   ErrorOr<std::unique_ptr<MachOObjectFile>>
   getObjectForArch(Triple::ArchType Arch) const;
 };
 
 }
 }
 
 #endif

test/tools/llvm-readobj/macho-universal-x86_64.i386.test

+RUN: llvm-readobj -h %p/Inputs/macho-universal.x86_64.i386 \
+RUN: | FileCheck %s -check-prefix MULTIHEADER
+
+RUN: llvm-readobj -sections %p/Inputs/macho-universal.x86_64.i386 \
+RUN: | FileCheck %s -check-prefix MULTISECTIONS
+
+MULTIHEADER: Format: Mach-O 64-bit x86-64
+MULTIHEADER: Arch: x86_64
+MULTIHEADER: AddressSize: 64bit
+MULTIHEADER: MachHeader {
+MULTIHEADER:   Magic: Magic64 (0xFEEDFACF)
+MULTIHEADER:   CpuType: X86-64 (0x1000007)
+MULTIHEADER:   CpuSubType: CPU_SUBTYPE_X86_64_ALL (0x3)
+MULTIHEADER:   FileType: Executable (0x2)
+MULTIHEADER:   NumOfLoadCommands: 16
+MULTIHEADER:   SizeOfLoadCommands: 880
+MULTIHEADER:   Flags [ (0x200085)
+MULTIHEADER:     MH_DYLDLINK (0x4)
+MULTIHEADER:     MH_NOUNDEFS (0x1)
+MULTIHEADER:     MH_PIE (0x200000)
+MULTIHEADER:     MH_TWOLEVEL (0x80)
+MULTIHEADER:   ]
+MULTIHEADER:   Reserved: 0x0
+MULTIHEADER: }
+
+MULTIHEADER: Format: Mach-O 32-bit i386
+MULTIHEADER: Arch: i386
+MULTIHEADER: AddressSize: 32bit
+MULTIHEADER: MachHeader {
+MULTIHEADER:   Magic: Magic (0xFEEDFACE)
+MULTIHEADER:   CpuType: X86 (0x7)
+MULTIHEADER:   CpuSubType: CPU_SUBTYPE_I386_ALL (0x3)
+MULTIHEADER:   FileType: Executable (0x2)
+MULTIHEADER:   NumOfLoadCommands: 16
+MULTIHEADER:   SizeOfLoadCommands: 716
+MULTIHEADER:   Flags [ (0x1200085)
+MULTIHEADER:     MH_DYLDLINK (0x4)
+MULTIHEADER:     MH_NOUNDEFS (0x1)
+MULTIHEADER:     MH_NO_HEAP_EXECUTION (0x1000000)
+MULTIHEADER:     MH_PIE (0x200000)
+MULTIHEADER:     MH_TWOLEVEL (0x80)
+MULTIHEADER:   ]
+MULTIHEADER: }
+
+
+
+MULTISECTIONS: Format: Mach-O 64-bit x86-64
+MULTISECTIONS: Arch: x86_64
+MULTISECTIONS: AddressSize: 64bit
+MULTISECTIONS: Sections [
+MULTISECTIONS:   Section {
+MULTISECTIONS:     Index: 0
+MULTISECTIONS:     Name: __text (5F 5F 74 65 78 74 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Segment: __TEXT (5F 5F 54 45 58 54 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Address: 0x100000F60
+MULTISECTIONS:     Size: 0x12
+MULTISECTIONS:     Offset: 3936
+MULTISECTIONS:     Alignment: 4
+MULTISECTIONS:     RelocationOffset: 0x0
+MULTISECTIONS:     RelocationCount: 0
+MULTISECTIONS:     Type: 0x0
+MULTISECTIONS:     Attributes [ (0x800004)
+MULTISECTIONS:       PureInstructions (0x800000)
+MULTISECTIONS:       SomeInstructions (0x4)
+MULTISECTIONS:     ]
+MULTISECTIONS:     Reserved1: 0x0
+MULTISECTIONS:     Reserved2: 0x0
+MULTISECTIONS:   }
+MULTISECTIONS:   Section {
+MULTISECTIONS:     Index: 1
+MULTISECTIONS:     Name: __unwind_info (5F 5F 75 6E 77 69 6E 64 5F 69 6E 66 6F 00 00 00)
+MULTISECTIONS:     Segment: __TEXT (5F 5F 54 45 58 54 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Address: 0x100000F72
+MULTISECTIONS:     Size: 0x48
+MULTISECTIONS:     Offset: 3954
+MULTISECTIONS:     Alignment: 0
+MULTISECTIONS:     RelocationOffset: 0x0
+MULTISECTIONS:     RelocationCount: 0
+MULTISECTIONS:     Type: 0x0
+MULTISECTIONS:     Attributes [ (0x0)
+MULTISECTIONS:     ]
+MULTISECTIONS:     Reserved1: 0x0
+MULTISECTIONS:     Reserved2: 0x0
+MULTISECTIONS:   }
+MULTISECTIONS:   Section {
+MULTISECTIONS:     Index: 2
+MULTISECTIONS:     Name: __eh_frame (5F 5F 65 68 5F 66 72 61 6D 65 00 00 00 00 00 00)
+MULTISECTIONS:     Segment: __TEXT (5F 5F 54 45 58 54 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Address: 0x100000FC0
+MULTISECTIONS:     Size: 0x40
+MULTISECTIONS:     Offset: 4032
+MULTISECTIONS:     Alignment: 3
+MULTISECTIONS:     RelocationOffset: 0x0
+MULTISECTIONS:     RelocationCount: 0
+MULTISECTIONS:     Type: 0x0
+MULTISECTIONS:     Attributes [ (0x0)
+MULTISECTIONS:     ]
+MULTISECTIONS:     Reserved1: 0x0
+MULTISECTIONS:     Reserved2: 0x0
+MULTISECTIONS:   }
+MULTISECTIONS: ]
+
+MULTISECTIONS: Format: Mach-O 32-bit i386
+MULTISECTIONS: Arch: i386
+MULTISECTIONS: AddressSize: 32bit
+MULTISECTIONS: Sections [
+MULTISECTIONS:   Section {
+MULTISECTIONS:     Index: 0
+MULTISECTIONS:     Name: __text (5F 5F 74 65 78 74 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Segment: __TEXT (5F 5F 54 45 58 54 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Address: 0x1FA0
+MULTISECTIONS:     Size: 0x15
+MULTISECTIONS:     Offset: 4000
+MULTISECTIONS:     Alignment: 4
+MULTISECTIONS:     RelocationOffset: 0x0
+MULTISECTIONS:     RelocationCount: 0
+MULTISECTIONS:     Type: 0x0
+MULTISECTIONS:     Attributes [ (0x800004)
+MULTISECTIONS:       PureInstructions (0x800000)
+MULTISECTIONS:       SomeInstructions (0x4)
+MULTISECTIONS:     ]
+MULTISECTIONS:     Reserved1: 0x0
+MULTISECTIONS:     Reserved2: 0x0
+MULTISECTIONS:   }
+MULTISECTIONS:   Section {
+MULTISECTIONS:     Index: 1
+MULTISECTIONS:     Name: __unwind_info (5F 5F 75 6E 77 69 6E 64 5F 69 6E 66 6F 00 00 00)
+MULTISECTIONS:     Segment: __TEXT (5F 5F 54 45 58 54 00 00 00 00 00 00 00 00 00 00)
+MULTISECTIONS:     Address: 0x1FB5
+MULTISECTIONS:     Size: 0x48
+MULTISECTIONS:     Offset: 4021
+MULTISECTIONS:     Alignment: 0
+MULTISECTIONS:     RelocationOffset: 0x0
+MULTISECTIONS:     RelocationCount: 0
+MULTISECTIONS:     Type: 0x0
+MULTISECTIONS:     Attributes [ (0x0)
+MULTISECTIONS:     ]
+MULTISECTIONS:     Reserved1: 0x0
+MULTISECTIONS:     Reserved2: 0x0
+MULTISECTIONS:   }
+MULTISECTIONS: ]

tools/llvm-readobj/llvm-readobj.cpp

 //===----------------------------------------------------------------------===//
 
 #include "llvm-readobj.h"
 #include "Error.h"
 #include "ObjDumper.h"
 #include "StreamWriter.h"
 #include "llvm/Object/Archive.h"
 #include "llvm/Object/ELFObjectFile.h"
+#include "llvm/Object/MachOUniversal.h"
 #include "llvm/Object/ObjectFile.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/DataTypes.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/FileSystem.h"
 #include "llvm/Support/ManagedStatic.h"
 #include "llvm/Support/PrettyStackTrace.h"
for (Archive::child_iterator ArcI = Arc->child_begin(),
 
     if (ObjectFile *Obj = dyn_cast<ObjectFile>(&*ChildOrErr.get()))
       dumpObject(Obj);
     else
       reportError(Arc->getFileName(), readobj_error::unrecognized_file_format);
   }
 }
 
+/// @brief Dumps each object file in \a MachO Universal Binary;
+static void dumpMachOUniversalBinary(const MachOUniversalBinary *UBinary) {
+  for (const MachOUniversalBinary::ObjectForArch &Obj : UBinary->objects()) {
+    ErrorOr<std::unique_ptr<MachOObjectFile>> ObjOrErr = Obj.getAsObjectFile();
+    if (std::error_code EC = ObjOrErr.getError()) {
+      reportError(UBinary->getFileName(), EC.message());
+      continue;
+    }
+
+    if (MachOObjectFile *MachOObj = ObjOrErr.get().get())
+      dumpObject(MachOObj);
+  }
+}
 
 /// @brief Opens \a File and dumps it.
 static void dumpInput(StringRef File) {
   // If file isn't stdin, check that it exists.
   if (File != "-" && !sys::fs::exists(File)) {
     reportError(File, readobj_error::file_not_found);
     return;
   }
 
   // Attempt to open the binary.
   ErrorOr<OwningBinary<Binary>> BinaryOrErr = createBinary(File);
   if (std::error_code EC = BinaryOrErr.getError()) {
     reportError(File, EC);
     return;
   }
   Binary &Binary = *BinaryOrErr.get().getBinary();
 
   if (Archive *Arc = dyn_cast<Archive>(&Binary))
     dumpArchive(Arc);
+  else if (MachOUniversalBinary *UBinary = dyn_cast<MachOUniversalBinary>(&Binary))
+    dumpMachOUniversalBinary(UBinary);
   else if (ObjectFile *Obj = dyn_cast<ObjectFile>(&Binary))
     dumpObject(Obj);
   else
     reportError(File, readobj_error::unrecognized_file_format);
 }
 
 
 int main(int argc, const char *argv[]) {