X-Git-Url: http://plrg.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FObject%2FArchive.cpp;h=99b0650c8b7e26efc0d7bd832f1cdc4f42cb0250;hb=b508a338d9d922a1ec3fbef698bd9fc6b5217ae0;hp=ba35e49faa67b26269d7de4c6742bdf6e13ca5e3;hpb=cdfe54f8a9e12b4d3a86bbc96e4a968f37ec4290;p=oota-llvm.git

diff --git a/lib/Object/Archive.cpp b/lib/Object/Archive.cpp
index ba35e49faa6..99b0650c8b7 100644
--- a/lib/Object/Archive.cpp
+++ b/lib/Object/Archive.cpp
@@ -17,9 +17,11 @@
 #include "llvm/ADT/Twine.h"
 #include "llvm/Support/Endian.h"
 #include "llvm/Support/MemoryBuffer.h"
+#include "llvm/Support/Path.h"
 
 using namespace llvm;
 using namespace object;
+using namespace llvm::support::endian;
 
 static const char *const Magic = "!<arch>\n";
 static const char *const ThinMagic = "!<thin>\n";
@@ -41,10 +43,10 @@ StringRef ArchiveMemberHeader::getName() const {
   return llvm::StringRef(Name, end);
 }
 
-uint32_t ArchiveMemberHeader::getSize() const {
+ErrorOr<uint32_t> ArchiveMemberHeader::getSize() const {
   uint32_t Ret;
   if (llvm::StringRef(Size, sizeof(Size)).rtrim(" ").getAsInteger(10, Ret))
-    llvm_unreachable("Size is not a decimal number.");
+    return object_error::parse_failed; // Size is not a decimal number.
   return Ret;
 }
 
@@ -80,22 +82,30 @@ unsigned ArchiveMemberHeader::getGID() const {
   return Ret;
 }
 
-Archive::Child::Child(const Archive *Parent, const char *Start)
+Archive::Child::Child(const Archive *Parent, StringRef Data,
+                      uint16_t StartOfFile)
+    : Parent(Parent), Data(Data), StartOfFile(StartOfFile) {}
+
+Archive::Child::Child(const Archive *Parent, const char *Start,
+                      std::error_code *EC)
     : Parent(Parent) {
   if (!Start)
     return;
 
-  const ArchiveMemberHeader *Header =
-      reinterpret_cast<const ArchiveMemberHeader *>(Start);
   uint64_t Size = sizeof(ArchiveMemberHeader);
-  if (!Parent->IsThin || Header->getName() == "/" || Header->getName() == "//")
-    Size += Header->getSize();
   Data = StringRef(Start, Size);
+  if (!isThinMember()) {
+    ErrorOr<uint64_t> MemberSize = getRawSize();
+    if ((*EC = MemberSize.getError()))
+      return;
+    Size += MemberSize.get();
+    Data = StringRef(Start, Size);
+  }
 
   // Setup StartOfFile and PaddingBytes.
   StartOfFile = sizeof(ArchiveMemberHeader);
   // Don't include attached name.
-  StringRef Name = Header->getName();
+  StringRef Name = getRawName();
   if (Name.startswith("#1/")) {
     uint64_t NameSize;
     if (Name.substr(3).rtrim(" ").getAsInteger(10, NameSize))
@@ -104,19 +114,49 @@ Archive::Child::Child(const Archive *Parent, const char *Start)
   }
 }
 
-uint64_t Archive::Child::getSize() const {
-  if (Parent->IsThin)
-    return getHeader()->getSize();
+ErrorOr<uint64_t> Archive::Child::getSize() const {
+  if (Parent->IsThin) {
+    ErrorOr<uint32_t> Size = getHeader()->getSize();
+    if (std::error_code EC = Size.getError())
+      return EC;
+    return Size.get();
+  }
   return Data.size() - StartOfFile;
 }
 
-uint64_t Archive::Child::getRawSize() const {
-  if (Parent->IsThin)
-    return getHeader()->getSize();
-  return Data.size();
+ErrorOr<uint64_t> Archive::Child::getRawSize() const {
+  ErrorOr<uint32_t> Size = getHeader()->getSize();
+  if (std::error_code EC = Size.getError())
+    return EC;
+  return Size.get();
+}
+
+bool Archive::Child::isThinMember() const {
+  StringRef Name = getHeader()->getName();
+  return Parent->IsThin && Name != "/" && Name != "//";
 }
 
-Archive::Child Archive::Child::getNext() const {
+ErrorOr<StringRef> Archive::Child::getBuffer() const {
+  if (!isThinMember()) {
+    ErrorOr<uint32_t> Size = getSize();
+    if (std::error_code EC = Size.getError())
+      return EC;
+    return StringRef(Data.data() + StartOfFile, Size.get());
+  }
+  ErrorOr<StringRef> Name = getName();
+  if (std::error_code EC = Name.getError())
+    return EC;
+  SmallString<128> FullName = sys::path::parent_path(
+      Parent->getMemoryBufferRef().getBufferIdentifier());
+  sys::path::append(FullName, *Name);
+  ErrorOr<std::unique_ptr<MemoryBuffer>> Buf = MemoryBuffer::getFile(FullName);
+  if (std::error_code EC = Buf.getError())
+    return EC;
+  Parent->ThinBuffers.push_back(std::move(*Buf));
+  return Parent->ThinBuffers.back()->getBuffer();
+}
+
+ErrorOr<Archive::Child> Archive::Child::getNext() const {
   size_t SpaceToSkip = Data.size();
   // If it's odd, add 1 to make it even.
   if (SpaceToSkip & 1)
@@ -124,11 +164,19 @@ Archive::Child Archive::Child::getNext() const {
 
   const char *NextLoc = Data.data() + SpaceToSkip;
 
+  // Check to see if this is at the end of the archive.
+  if (NextLoc == Parent->Data.getBufferEnd())
+    return Child(Parent, nullptr, nullptr);
+
   // Check to see if this is past the end of the archive.
-  if (NextLoc >= Parent->Data.getBufferEnd())
-    return Child(Parent, nullptr);
+  if (NextLoc > Parent->Data.getBufferEnd())
+    return object_error::parse_failed;
 
-  return Child(Parent, NextLoc);
+  std::error_code EC;
+  Child Ret(Parent, NextLoc, &EC);
+  if (EC)
+    return EC;
+  return Ret;
 }
 
 uint64_t Archive::Child::getChildOffset() const {
@@ -151,22 +199,16 @@ ErrorOr<StringRef> Archive::Child::getName() const {
     std::size_t offset;
     if (name.substr(1).rtrim(" ").getAsInteger(10, offset))
       llvm_unreachable("Long name offset is not an integer");
-    const char *addr = Parent->StringTable->Data.begin()
-                       + sizeof(ArchiveMemberHeader)
-                       + offset;
+
     // Verify it.
-    if (Parent->StringTable == Parent->child_end()
-        || addr < (Parent->StringTable->Data.begin()
-                   + sizeof(ArchiveMemberHeader))
-        || addr > (Parent->StringTable->Data.begin()
-                   + sizeof(ArchiveMemberHeader)
-                   + Parent->StringTable->getSize()))
+    if (offset >= Parent->StringTable.size())
       return object_error::parse_failed;
+    const char *addr = Parent->StringTable.begin() + offset;
 
-    // GNU long file names end with a /.
-    if (Parent->kind() == K_GNU) {
-      StringRef::size_type End = StringRef(addr).find('/');
-      return StringRef(addr, End);
+    // GNU long file names end with a "/\n".
+    if (Parent->kind() == K_GNU || Parent->kind() == K_MIPS64) {
+      StringRef::size_type End = StringRef(addr).find('\n');
+      return StringRef(addr, End - 1);
     }
     return StringRef(addr);
   } else if (name.startswith("#1/")) {
@@ -187,7 +229,10 @@ ErrorOr<MemoryBufferRef> Archive::Child::getMemoryBufferRef() const {
   if (std::error_code EC = NameOrErr.getError())
     return EC;
   StringRef Name = NameOrErr.get();
-  return MemoryBufferRef(getBuffer(), Name);
+  ErrorOr<StringRef> Buf = getBuffer();
+  if (std::error_code EC = Buf.getError())
+    return EC;
+  return MemoryBufferRef(*Buf, Name);
 }
 
 ErrorOr<std::unique_ptr<Binary>>
@@ -207,8 +252,13 @@ ErrorOr<std::unique_ptr<Archive>> Archive::create(MemoryBufferRef Source) {
   return std::move(Ret);
 }
 
+void Archive::setFirstRegular(const Child &C) {
+  FirstRegularData = C.Data;
+  FirstRegularStartOfFile = C.StartOfFile;
+}
+
 Archive::Archive(MemoryBufferRef Source, std::error_code &ec)
-    : Binary(Binary::ID_Archive, Source), SymbolTable(child_end()) {
+    : Binary(Binary::ID_Archive, Source) {
   StringRef Buffer = Data.getBuffer();
   // Check for sufficient magic.
   if (Buffer.startswith(ThinMagic)) {
@@ -221,15 +271,26 @@ Archive::Archive(MemoryBufferRef Source, std::error_code &ec)
   }
 
   // Get the special members.
-  child_iterator i = child_begin(false);
-  child_iterator e = child_end();
+  child_iterator I = child_begin(false);
+  if ((ec = I->getError()))
+    return;
+  child_iterator E = child_end();
 
-  if (i == e) {
-    ec = object_error::success;
+  if (I == E) {
+    ec = std::error_code();
     return;
   }
+  const Child *C = &**I;
 
-  StringRef Name = i->getRawName();
+  auto Increment = [&]() {
+    ++I;
+    if ((ec = I->getError()))
+      return true;
+    C = &**I;
+    return false;
+  };
+
+  StringRef Name = C->getRawName();
 
   // Below is the pattern that is used to figure out the archive format
   // GNU archive format
@@ -252,53 +313,74 @@ Archive::Archive(MemoryBufferRef Source, std::error_code &ec)
 
   if (Name == "__.SYMDEF") {
     Format = K_BSD;
-    SymbolTable = i;
-    ++i;
-    FirstRegular = i;
-    ec = object_error::success;
+    // We know that the symbol table is not an external file, so we just assert
+    // there is no error.
+    SymbolTable = *C->getBuffer();
+    if (Increment())
+      return;
+    setFirstRegular(*C);
+
+    ec = std::error_code();
     return;
   }
 
   if (Name.startswith("#1/")) {
     Format = K_BSD;
     // We know this is BSD, so getName will work since there is no string table.
-    ErrorOr<StringRef> NameOrErr = i->getName();
+    ErrorOr<StringRef> NameOrErr = C->getName();
     ec = NameOrErr.getError();
     if (ec)
       return;
     Name = NameOrErr.get();
     if (Name == "__.SYMDEF SORTED" || Name == "__.SYMDEF") {
-      SymbolTable = i;
-      ++i;
+      // We know that the symbol table is not an external file, so we just
+      // assert there is no error.
+      SymbolTable = *C->getBuffer();
+      if (Increment())
+        return;
     }
-    FirstRegular = i;
+    setFirstRegular(*C);
     return;
   }
 
-  if (Name == "/") {
-    SymbolTable = i;
+  // MIPS 64-bit ELF archives use a special format of a symbol table.
+  // This format is marked by `ar_name` field equals to "/SYM64/".
+  // For detailed description see page 96 in the following document:
+  // http://techpubs.sgi.com/library/manuals/4000/007-4658-001/pdf/007-4658-001.pdf
 
-    ++i;
-    if (i == e) {
-      ec = object_error::parse_failed;
+  bool has64SymTable = false;
+  if (Name == "/" || Name == "/SYM64/") {
+    // We know that the symbol table is not an external file, so we just assert
+    // there is no error.
+    SymbolTable = *C->getBuffer();
+    if (Name == "/SYM64/")
+      has64SymTable = true;
+
+    if (Increment())
+      return;
+    if (I == E) {
+      ec = std::error_code();
       return;
     }
-    Name = i->getRawName();
+    Name = C->getRawName();
   }
 
   if (Name == "//") {
-    Format = K_GNU;
-    StringTable = i;
-    ++i;
-    FirstRegular = i;
-    ec = object_error::success;
+    Format = has64SymTable ? K_MIPS64 : K_GNU;
+    // The string table is never an external member, so we just assert on the
+    // ErrorOr.
+    StringTable = *C->getBuffer();
+    if (Increment())
+      return;
+    setFirstRegular(*C);
+    ec = std::error_code();
     return;
   }
 
   if (Name[0] != '/') {
-    Format = K_GNU;
-    FirstRegular = i;
-    ec = object_error::success;
+    Format = has64SymTable ? K_MIPS64 : K_GNU;
+    setFirstRegular(*C);
+    ec = std::error_code();
     return;
   }
 
@@ -308,24 +390,31 @@ Archive::Archive(MemoryBufferRef Source, std::error_code &ec)
   }
 
   Format = K_COFF;
-  SymbolTable = i;
+  // We know that the symbol table is not an external file, so we just assert
+  // there is no error.
+  SymbolTable = *C->getBuffer();
 
-  ++i;
-  if (i == e) {
-    FirstRegular = i;
-    ec = object_error::success;
+  if (Increment())
+    return;
+
+  if (I == E) {
+    setFirstRegular(*C);
+    ec = std::error_code();
     return;
   }
 
-  Name = i->getRawName();
+  Name = C->getRawName();
 
   if (Name == "//") {
-    StringTable = i;
-    ++i;
+    // The string table is never an external member, so we just assert on the
+    // ErrorOr.
+    StringTable = *C->getBuffer();
+    if (Increment())
+      return;
   }
 
-  FirstRegular = i;
-  ec = object_error::success;
+  setFirstRegular(*C);
+  ec = std::error_code();
 }
 
 Archive::child_iterator Archive::child_begin(bool SkipInternal) const {
@@ -333,28 +422,36 @@ Archive::child_iterator Archive::child_begin(bool SkipInternal) const {
     return child_end();
 
   if (SkipInternal)
-    return FirstRegular;
+    return Child(this, FirstRegularData, FirstRegularStartOfFile);
 
   const char *Loc = Data.getBufferStart() + strlen(Magic);
-  Child c(this, Loc);
-  return c;
+  std::error_code EC;
+  Child c(this, Loc, &EC);
+  if (EC)
+    return child_iterator(EC);
+  return child_iterator(c);
 }
 
 Archive::child_iterator Archive::child_end() const {
-  return Child(this, nullptr);
+  return Child(this, nullptr, nullptr);
 }
 
 StringRef Archive::Symbol::getName() const {
-  return Parent->SymbolTable->getBuffer().begin() + StringIndex;
+  return Parent->getSymbolTable().begin() + StringIndex;
 }
 
-ErrorOr<Archive::child_iterator> Archive::Symbol::getMember() const {
-  const char *Buf = Parent->SymbolTable->getBuffer().begin();
-  const char *Offsets = Buf + 4;
+ErrorOr<Archive::Child> Archive::Symbol::getMember() const {
+  const char *Buf = Parent->getSymbolTable().begin();
+  const char *Offsets = Buf;
+  if (Parent->kind() == K_MIPS64)
+    Offsets += sizeof(uint64_t);
+  else
+    Offsets += sizeof(uint32_t);
   uint32_t Offset = 0;
   if (Parent->kind() == K_GNU) {
-    Offset = *(reinterpret_cast<const support::ubig32_t*>(Offsets)
-               + SymbolIndex);
+    Offset = read32be(Offsets + SymbolIndex * 4);
+  } else if (Parent->kind() == K_MIPS64) {
+    Offset = read64be(Offsets + SymbolIndex * 8);
   } else if (Parent->kind() == K_BSD) {
     // The SymbolIndex is an index into the ranlib structs that start at
     // Offsets (the first uint32_t is the number of bytes of the ranlib
@@ -362,41 +459,37 @@ ErrorOr<Archive::child_iterator> Archive::Symbol::getMember() const {
     // being a string table offset and the second being the offset into
     // the archive of the member that defines the symbol.  Which is what
     // is needed here.
-    Offset = *(reinterpret_cast<const support::ulittle32_t *>(Offsets) +
-               (SymbolIndex * 2) + 1);
+    Offset = read32le(Offsets + SymbolIndex * 8 + 4);
   } else {
-    uint32_t MemberCount = *reinterpret_cast<const support::ulittle32_t*>(Buf);
-    
     // Skip offsets.
-    Buf += sizeof(support::ulittle32_t)
-           + (MemberCount * sizeof(support::ulittle32_t));
-
-    uint32_t SymbolCount = *reinterpret_cast<const support::ulittle32_t*>(Buf);
+    uint32_t MemberCount = read32le(Buf);
+    Buf += MemberCount * 4 + 4;
 
+    uint32_t SymbolCount = read32le(Buf);
     if (SymbolIndex >= SymbolCount)
       return object_error::parse_failed;
 
     // Skip SymbolCount to get to the indices table.
-    const char *Indices = Buf + sizeof(support::ulittle32_t);
+    const char *Indices = Buf + 4;
 
     // Get the index of the offset in the file member offset table for this
     // symbol.
-    uint16_t OffsetIndex =
-      *(reinterpret_cast<const support::ulittle16_t*>(Indices)
-        + SymbolIndex);
+    uint16_t OffsetIndex = read16le(Indices + SymbolIndex * 2);
     // Subtract 1 since OffsetIndex is 1 based.
     --OffsetIndex;
 
     if (OffsetIndex >= MemberCount)
       return object_error::parse_failed;
 
-    Offset = *(reinterpret_cast<const support::ulittle32_t*>(Offsets)
-               + OffsetIndex);
+    Offset = read32le(Offsets + OffsetIndex * 4);
   }
 
   const char *Loc = Parent->getData().begin() + Offset;
-  child_iterator Iter(Child(Parent, Loc));
-  return Iter;
+  std::error_code EC;
+  Child C(Parent, Loc, &EC);
+  if (EC)
+    return EC;
+  return C;
 }
 
 Archive::Symbol Archive::Symbol::getNext() const {
@@ -415,10 +508,9 @@ Archive::Symbol Archive::Symbol::getNext() const {
     // and the second being the offset into the archive of the member that
     // define the symbol. After that the next uint32_t is the byte count of
     // the string table followed by the string table.
-    const char *Buf = Parent->SymbolTable->getBuffer().begin();
+    const char *Buf = Parent->getSymbolTable().begin();
     uint32_t RanlibCount = 0;
-    RanlibCount = (*reinterpret_cast<const support::ulittle32_t *>(Buf)) /
-                  (sizeof(uint32_t) * 2);
+    RanlibCount = read32le(Buf) / 8;
     // If t.SymbolIndex + 1 will be past the count of symbols (the RanlibCount)
     // don't change the t.StringIndex as we don't want to reference a ranlib
     // past RanlibCount.
@@ -426,17 +518,14 @@ Archive::Symbol Archive::Symbol::getNext() const {
       const char *Ranlibs = Buf + 4;
       uint32_t CurRanStrx = 0;
       uint32_t NextRanStrx = 0;
-      CurRanStrx = *(reinterpret_cast<const support::ulittle32_t *>(Ranlibs) +
-                     (t.SymbolIndex * 2));
-      NextRanStrx = *(reinterpret_cast<const support::ulittle32_t *>(Ranlibs) +
-                      ((t.SymbolIndex + 1) * 2));
+      CurRanStrx = read32le(Ranlibs + t.SymbolIndex * 8);
+      NextRanStrx = read32le(Ranlibs + (t.SymbolIndex + 1) * 8);
       t.StringIndex -= CurRanStrx;
       t.StringIndex += NextRanStrx;
     }
   } else {
     // Go to one past next null.
-    t.StringIndex =
-        Parent->SymbolTable->getBuffer().find('\0', t.StringIndex) + 1;
+    t.StringIndex = Parent->getSymbolTable().find('\0', t.StringIndex) + 1;
   }
   ++t.SymbolIndex;
   return t;
@@ -446,11 +535,14 @@ Archive::symbol_iterator Archive::symbol_begin() const {
   if (!hasSymbolTable())
     return symbol_iterator(Symbol(this, 0, 0));
 
-  const char *buf = SymbolTable->getBuffer().begin();
+  const char *buf = getSymbolTable().begin();
   if (kind() == K_GNU) {
     uint32_t symbol_count = 0;
-    symbol_count = *reinterpret_cast<const support::ubig32_t*>(buf);
+    symbol_count = read32be(buf);
     buf += sizeof(uint32_t) + (symbol_count * (sizeof(uint32_t)));
+  } else if (kind() == K_MIPS64) {
+    uint64_t symbol_count = read64be(buf);
+    buf += sizeof(uint64_t) + (symbol_count * (sizeof(uint64_t)));
   } else if (kind() == K_BSD) {
     // The __.SYMDEF or "__.SYMDEF SORTED" member starts with a uint32_t
     // which is the number of bytes of ranlib structs that follow.  The ranlib
@@ -459,11 +551,10 @@ Archive::symbol_iterator Archive::symbol_begin() const {
     // define the symbol. After that the next uint32_t is the byte count of
     // the string table followed by the string table.
     uint32_t ranlib_count = 0;
-    ranlib_count = (*reinterpret_cast<const support::ulittle32_t *>(buf)) /
-                   (sizeof(uint32_t) * 2);
+    ranlib_count = read32le(buf) / 8;
     const char *ranlibs = buf + 4;
     uint32_t ran_strx = 0;
-    ran_strx = *(reinterpret_cast<const support::ulittle32_t *>(ranlibs));
+    ran_strx = read32le(ranlibs);
     buf += sizeof(uint32_t) + (ranlib_count * (2 * (sizeof(uint32_t))));
     // Skip the byte count of the string table.
     buf += sizeof(uint32_t);
@@ -471,34 +562,33 @@ Archive::symbol_iterator Archive::symbol_begin() const {
   } else {
     uint32_t member_count = 0;
     uint32_t symbol_count = 0;
-    member_count = *reinterpret_cast<const support::ulittle32_t*>(buf);
+    member_count = read32le(buf);
     buf += 4 + (member_count * 4); // Skip offsets.
-    symbol_count = *reinterpret_cast<const support::ulittle32_t*>(buf);
+    symbol_count = read32le(buf);
     buf += 4 + (symbol_count * 2); // Skip indices.
   }
-  uint32_t string_start_offset = buf - SymbolTable->getBuffer().begin();
+  uint32_t string_start_offset = buf - getSymbolTable().begin();
   return symbol_iterator(Symbol(this, 0, string_start_offset));
 }
 
 Archive::symbol_iterator Archive::symbol_end() const {
-  if (!hasSymbolTable())
-    return symbol_iterator(Symbol(this, 0, 0));
+  return symbol_iterator(Symbol(this, getNumberOfSymbols(), 0));
+}
 
-  const char *buf = SymbolTable->getBuffer().begin();
-  uint32_t symbol_count = 0;
-  if (kind() == K_GNU) {
-    symbol_count = *reinterpret_cast<const support::ubig32_t*>(buf);
-  } else if (kind() == K_BSD) {
-    symbol_count = (*reinterpret_cast<const support::ulittle32_t *>(buf)) /
-                   (sizeof(uint32_t) * 2);
-  } else {
-    uint32_t member_count = 0;
-    member_count = *reinterpret_cast<const support::ulittle32_t*>(buf);
-    buf += 4 + (member_count * 4); // Skip offsets.
-    symbol_count = *reinterpret_cast<const support::ulittle32_t*>(buf);
-  }
-  return symbol_iterator(
-    Symbol(this, symbol_count, 0));
+uint32_t Archive::getNumberOfSymbols() const {
+  if (!hasSymbolTable())
+    return 0;
+  const char *buf = getSymbolTable().begin();
+  if (kind() == K_GNU)
+    return read32be(buf);
+  if (kind() == K_MIPS64)
+    return read64be(buf);
+  if (kind() == K_BSD)
+    return read32le(buf) / 8;
+  uint32_t member_count = 0;
+  member_count = read32le(buf);
+  buf += 4 + (member_count * 4); // Skip offsets.
+  return read32le(buf);
 }
 
 Archive::child_iterator Archive::findSym(StringRef name) const {
@@ -518,6 +608,4 @@ Archive::child_iterator Archive::findSym(StringRef name) const {
   return child_end();
 }
 
-bool Archive::hasSymbolTable() const {
-  return SymbolTable != child_end();
-}
+bool Archive::hasSymbolTable() const { return !SymbolTable.empty(); }