Add abilities to generate bitcode buffers from munged bitcode.

lib/Bitcode/NaCl/TestUtils/NaClBitcodeMungeWriter.cpp

Index: lib/Bitcode/NaCl/TestUtils/NaClBitcodeMungeWriter.cpp
diff --git a/lib/Bitcode/NaCl/TestUtils/NaClBitcodeMungeWriter.cpp b/lib/Bitcode/NaCl/TestUtils/NaClBitcodeMungeWriter.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..85da68ddaf3cbcb769ad2f29c2a59d8c67696384
--- /dev/null
+++ b/lib/Bitcode/NaCl/TestUtils/NaClBitcodeMungeWriter.cpp
@@ -0,0 +1,336 @@
+//===- NaClBitcodeMungeWriter.cpp - Write munged bitcode --------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// Implements method NaClMungedBitcode.write(), which writes out a munged
+// list of bitcode records using a bitstream writer.
+
+#include "llvm/Bitcode/NaCl/NaClBitcodeMungeUtils.h"
+
+#include "llvm/Bitcode/NaCl/NaClBitstreamWriter.h"
+#include "llvm/Bitcode/NaCl/NaClReaderWriter.h"
+
+using namespace llvm;
+
+namespace {
+
+// For debugging. When true, shows each PNaCl record that is
+// emitted to the bitcode file.
+static bool DebugEmitRecord = false;
+
+// State of bitcode writing.
+struct WriteState {
+  // The block ID associated with the block being written.
+  int WriteBlockID = -1;
+  // The SetBID for the blockinfo block.
+  int SetBID = -1;
+  // The stack of maximum abbreviation indices allowed by block enter record.
+  SmallVector<uint64_t, 3> AbbrevIndexLimitStack;
+  // The set of write flags to use.
+  const NaClMungedBitcode::WriteFlags &Flags;
+  // The results of the attempted write.
+  NaClMungedBitcode::WriteResults Results;
+
+  WriteState(const NaClMungedBitcode::WriteFlags &Flags) : Flags(Flags) {}
+
+  // Returns stream to print error message to.
+  raw_ostream &Error() {
+    Results.HasErrors = true;
+    return errs() << "Error (Block " << WriteBlockID << "): ";
+  }
+
+  // Returns stream to print error message to, assuming that
+  // the error message can be repaired if Flags.TryToRecover is true.
+  raw_ostream &RecoverableError() {
+    if (Flags.TryToRecover)
+      Results.HasRepairs = true;
+    return Error();
+  }
+
+  // Converts the abbreviation record to the corresponding abbreviation.
+  // Returns nullptr if unable to build abbreviation.
+  NaClBitCodeAbbrev *buildAbbrev(const NaClBitcodeAbbrevRecord &Record);
+
+  // Emits the given record to the bitcode file. Returns true if
+  // successful.
+  bool emitRecord(NaClBitstreamWriter &Writer,
+                  const NaClBitcodeAbbrevRecord &Record);
+
+  // Adds any missing end blocks to written bitcode.
+  void writeMissingEndBlocks(NaClBitstreamWriter &Writer) {
+    while (!AbbrevIndexLimitStack.empty()) {
+      Writer.ExitBlock();
+      AbbrevIndexLimitStack.pop_back();
+    }
+  }
+};
+
+bool WriteState::emitRecord(NaClBitstreamWriter &Writer,
+                            const NaClBitcodeAbbrevRecord &Record) {
+  size_t NumValues = Record.Values.size();
+  if (DebugEmitRecord) {
+    errs() << "Emit " << Record.Abbrev << ": <" << Record.Code;
+    for (size_t i = 0; i < NumValues; ++i) {
+      errs() << ", " << Record.Values[i];
+    }
+    errs() << ">\n";
+  }
+
+  switch (Record.Code) {
+  case naclbitc::BLK_CODE_ENTER: {
+    unsigned NumBits = naclbitc::DEFAULT_MAX_ABBREV;

[jvoung (off chromium), 2015/05/06 22:08:09]

Should this also be NaClBitsNeededForValue( naclbitc::DEFAULT_MAX_ABBREV) ?

If I'm reading this correctly, the initial value of NumBits happened to  never
be read, because the it'll luckily bail first.

[Karl, 2015/05/07 20:09:18]

Good catch. Fixing!

+    WriteBlockID = -1;
+    if (Record.Abbrev != naclbitc::ENTER_SUBBLOCK) {
+      RecoverableError()
+          << "Uses illegal abbreviation index in enter block record: "
+          << Record << "\n";
+      if (!Flags.TryToRecover)
+        return false;
+    }
+    if (NumValues == 2) {
+      WriteBlockID = Record.Values[0];
+      NumBits = Record.Values[1];
+      if (NumBits > 32 || NumBits < 2) {
+        RecoverableError()
+            << "Bit size " << NumBits << " for record should be "
+            << (NumBits > 32 ? "<= 32" : ">= 2") << ": " << Record << "\n";
+        if (!Flags.TryToRecover)
+          return false;
+        NumBits = 32;
+      }
+    } else {
+      Error() << "Values for enter record should be of size 2, but found "
+              << NumValues << ": " << Record << "\n";
+      return false;
+    }
+    uint64_t MaxAbbrev = (static_cast<uint64_t>(1) << NumBits) - 1;
+    AbbrevIndexLimitStack.push_back(MaxAbbrev);
+    if (WriteBlockID == naclbitc::BLOCKINFO_BLOCK_ID) {
+      unsigned DefaultMaxBits =
+          NaClBitsNeededForValue(naclbitc::DEFAULT_MAX_ABBREV);
+      if (NumBits != DefaultMaxBits) {
+        RecoverableError()
+            << "Numbits entry for abbreviations record not "
+            << DefaultMaxBits << " but found " << NumBits <<
+            ": " << Record << "\n";
+        if (!Flags.TryToRecover)
+          return false;
+      }
+      Writer.EnterBlockInfoBlock();
+    } else {
+      NaClBitcodeSelectorAbbrev CurCodeLen(MaxAbbrev);
+      Writer.EnterSubblock(WriteBlockID, CurCodeLen);
+    }
+    break;
+  }
+  case naclbitc::BLK_CODE_EXIT:
+    if (Record.Abbrev != naclbitc::END_BLOCK) {
+      RecoverableError()
+          << "Uses illegal abbreviation index in exit block record: "
+          << Record << "\n";
+      if (!Flags.TryToRecover)
+        return false;
+    }
+    if (NumValues != 0) {
+      RecoverableError() << "Exit block should not have values: "
+                         << Record << "\n";
+      if (!Flags.TryToRecover)
+        return false;
+    }
+    if (!AbbrevIndexLimitStack.empty())
+      AbbrevIndexLimitStack.pop_back();
+    Writer.ExitBlock();
+    break;
+  case naclbitc::BLK_CODE_DEFINE_ABBREV: {
+    if (Record.Abbrev != naclbitc::DEFINE_ABBREV) {
+      RecoverableError()
+          << "Uses illegal abbreviation index in define abbreviation record: "
+          << Record << "\n";
+      if (!Flags.TryToRecover)
+        return false;
+    }
+    NaClBitCodeAbbrev *Abbrev = buildAbbrev(Record);
+    if (Abbrev == NULL)
+      return Flags.TryToRecover;
+    if (WriteBlockID == naclbitc::BLOCKINFO_BLOCK_ID) {
+      Writer.EmitBlockInfoAbbrev(SetBID, Abbrev);
+    } else {
+      Writer.EmitAbbrev(Abbrev);
+    }
+    break;
+  }
+  case naclbitc::BLK_CODE_HEADER:
+    // Note: There is no abbreviation index here. Ignore.
+    for (uint64_t Value : Record.Values)
+      Writer.Emit(Value, 8);
+    break;
+  default:
+    bool UsesDefaultAbbrev = Record.Abbrev == naclbitc::UNABBREV_RECORD;
+    if (!UsesDefaultAbbrev
+        && !Writer.isUserRecordAbbreviation(Record.Abbrev)) {
+      uint64_t BlockAbbrevIndexLimit = 0;
+      if (!AbbrevIndexLimitStack.empty())
+        BlockAbbrevIndexLimit = AbbrevIndexLimitStack.back();
+      // Can't generate abbreviation ID because not enough bits
+      // specified in block ID.
+      raw_ostream &ErrStrm = RecoverableError();
+      ErrStrm << "Uses illegal abbreviation index";
+      if (Record.Abbrev > BlockAbbrevIndexLimit)

[jvoung (off chromium), 2015/05/06 22:08:09]

When does this > check fail? If for some reason the AbbrevIndexLimitStack is
empty and the Limit isn't known?

Seems like you should know that the Abbrev > SomeLimit of some kind at this
point, according to the "Can't generate abbreviation ID because not enough
bits..." comment.

[Karl, 2015/05/07 20:09:18]

Decided to simplify this code and be more generic.

+        ErrStrm << ". Must not exceed " << BlockAbbrevIndexLimit;
+      ErrStrm << ": " << Record << "\n";
+      if (!Flags.TryToRecover && !Flags.SaveBadAbbrevIndices)

[jvoung (off chromium), 2015/05/06 22:08:09]

So this code is saying that SaveBadAbbrevIndices will have one type of error
recovery, while plain "TryToRecover" will have a different type of error
recovery?

[Karl, 2015/05/07 20:09:18]

Hopefully the code is now clearer since I simplified. If SaveBadAbbrevIndices is
true, we write out an incorrect but read testable bitcode file. If TryToRecover,
we default to the default abbreviation as a recovery.

[jvoung (off chromium), 2015/05/07 21:33:56]

Thanks -- I like the new version more.

+        return false;
+      if (Flags.SaveBadAbbrevIndices) {
+        // Note: If this point is reached, the abbreviation is bad and
+        // RecoverOnBadAbbrevID is true. Generate bad abbreviation index

[jvoung (off chromium), 2015/05/06 22:08:09]

there is no RecoverOnBadAbbrevID anymore

You also talk about "Can't generate abbreviation ID because not enough bits"
earlier, so "If this point is reached, the abbreviation is bad..." applies to
the earlier code too?

[Karl, 2015/05/07 20:09:18]

Again, I simplified this code so that we don't need to be so specific (and
complicated).

+        // so that the bitcode reader can be tested.
+        Writer.EmitCode(Record.Abbrev);
+        // Note: We need to close blocks or the bitcode Writer will terminate
+        // due to assertions.
+        writeMissingEndBlocks(Writer);
+        Results.SavedBadAbbrevIndices = true;

[jvoung (off chromium), 2015/05/06 22:08:09]

Perhaps "Save..." should be "Write..." ?

In some ways Save is a bit less clear to me. It could be saved to some
field/vector somewhere, while the destination of Write might deduced to be the
Writer.

[Karl, 2015/05/07 20:09:18]

Hopefully I have clarified that the Save Case (now renamed WriteBadAbbrevIndex)
generates an illegal bitcode file that can be used to test the bitcode reader on
such a case.

+        return false;
+      }
+      UsesDefaultAbbrev = true;
+    }
+    if (WriteBlockID == naclbitc::BLOCKINFO_BLOCK_ID
+        && Record.Code == naclbitc::BLOCKINFO_CODE_SETBID) {
+      // Note: SetBID records are handled by Writer->EmitBlockInfoAbbrev,
+      // based on the SetBID value. Don't bother to generate SetBID record here.
+      // Rather just set SetBID and let call to Writer->EmitBlockInfoAbbrev
+      // generate the SetBID record.
+      if (NumValues != 1) {
+        Error() << "SetBID record expects 1 value but found "
+                << NumValues << ": " << Record << "\n";
+        return false;
+      }
+      SetBID = Record.Values[0];
+      return true;
+    }
+    if (UsesDefaultAbbrev)
+      Writer.EmitRecord(Record.Code, Record.Values);
+    else
+      Writer.EmitRecord(Record.Code, Record.Values, Record.Abbrev);
+  }
+  return true;
+}
+
+static NaClBitCodeAbbrev *deleteAbbrev(NaClBitCodeAbbrev *Abbrev) {
+  Abbrev->dropRef();
+  return nullptr;
+}
+
+NaClBitCodeAbbrev *WriteState::buildAbbrev(
+    const NaClBitcodeAbbrevRecord &Record) {
+  // Note: Recover by removing abbreviation.
+  NaClBitCodeAbbrev *Abbrev = new NaClBitCodeAbbrev();
+  size_t Index = 0;
+  size_t NumValues = Record.Values.size();
+  if (NumValues == 0) {
+    RecoverableError() << "Empty abbreviation record not allowed: "
+                       << Record << "\n";
+    return deleteAbbrev(Abbrev);
+  }
+  size_t NumAbbreviations = Record.Values[Index++];
+  if (NumAbbreviations == 0) {
+    RecoverableError() << "Abbreviation must contain at least one operator: "
+                       << Record << "\n";
+    return deleteAbbrev(Abbrev);
+  }
+  for (size_t Count = 0; Count < NumAbbreviations; ++Count) {
+    if (Index >= NumValues) {
+      RecoverableError()
+          << "Malformed abbreviation found. Expects " << NumAbbreviations
+          << " operands but ound " << Count << ": " << Record << "\n";
+      return deleteAbbrev(Abbrev);
+    }
+    switch (Record.Values[Index++]) {
+    case 1:
+      if (Index >= NumValues) {
+        RecoverableError() << "Malformed literal abbreviation: "
+                           << Record << "\n";
+        return deleteAbbrev(Abbrev);
+      }
+      Abbrev->Add(NaClBitCodeAbbrevOp(Record.Values[Index++]));
+      break;
+    case 0: {
+      if (Index >= NumValues) {
+        RecoverableError() << "Malformed abbreviation found: "
+                           << Record << "\n";
+        return deleteAbbrev(Abbrev);
+      }
+      unsigned Kind = Record.Values[Index++];
+      switch (Kind) {
+      case NaClBitCodeAbbrevOp::Fixed:
+        if (Index >= NumValues) {
+          RecoverableError() << "Malformed fixed abbreviation found: "
+                             << Record << "\n";
+          return deleteAbbrev(Abbrev);
+        }
+        Abbrev->Add(NaClBitCodeAbbrevOp(NaClBitCodeAbbrevOp::Fixed,
+                                        Record.Values[Index++]));
+        break;
+      case NaClBitCodeAbbrevOp::VBR:
+        if (Index >= NumValues) {
+          RecoverableError() << "Malformed vbr abbreviation found: "
+                             << Record << "\n";
+          return deleteAbbrev(Abbrev);
+        }
+        Abbrev->Add(NaClBitCodeAbbrevOp(NaClBitCodeAbbrevOp::VBR,
+                                        Record.Values[Index++]));
+        break;
+      case NaClBitCodeAbbrevOp::Array:
+        if (Index >= NumValues) {
+          RecoverableError() << "Malformed array abbreviation found: "
+                             << Record << "\n";
+          return deleteAbbrev(Abbrev);
+        }
+        Abbrev->Add(NaClBitCodeAbbrevOp(NaClBitCodeAbbrevOp::Array));
+        break;
+      case NaClBitCodeAbbrevOp::Char6:
+        Abbrev->Add(NaClBitCodeAbbrevOp(NaClBitCodeAbbrevOp::Char6));
+        break;
+      default:
+        RecoverableError() << "Unknown abbreviation kind " << Kind
+                           << ": " << Record << "\n";
+        return deleteAbbrev(Abbrev);
+      }
+      break;
+    }
+    default:
+      RecoverableError() << "Error: Bad literal flag " << Record.Values[Index]
+                         << ": " << Record << "\n";
+      return deleteAbbrev(Abbrev);
+    }
+  }
+  return Abbrev;
+}
+
+} // end of anonymous namespace.
+
+NaClMungedBitcode::WriteResults NaClMungedBitcode::writeMaybeRepair(
+    SmallVectorImpl<char> &Buffer, bool AddHeader,
+    const WriteFlags &Flags) const {
+  NaClBitstreamWriter Writer(Buffer);
+  WriteState State(Flags);
+  if (AddHeader) {
+    NaClWriteHeader(Writer, true);
+  }
+  for (const NaClBitcodeAbbrevRecord &Record : *this) {
+    if (!State.emitRecord(Writer, Record))
+      break;
+  }
+  if (!State.AbbrevIndexLimitStack.empty()) {
+    State.RecoverableError()
+        << "Bitcode missing " << State.AbbrevIndexLimitStack.size()
+        << " close blocks.\n";
+    if (Flags.TryToRecover)
+      State.writeMissingEndBlocks(Writer);
+  }
+  return State.Results;
+}