Add a libfuzzer for RtpHeaderParser.

webrtc/test/fuzzers/rtp_header_fuzzer.cc

+/*
+ *  Copyright (c) 2016 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include <bitset>
+
+#include "webrtc/modules/rtp_rtcp/source/rtp_packet_received.h"
+#include "webrtc/modules/rtp_rtcp/include/rtp_rtcp_defines.h"
+#include "webrtc/modules/rtp_rtcp/source/rtp_utility.h"
+
+namespace webrtc {
+void FuzzOneInput(const uint8_t* data, size_t size) {
+  // We decide which header extensions to register by reading one byte
+  // from the beginning of |data| and interpreting it as a bitmask over
+  // the RTPExtensionType enum.
+  if (size <= 1)
+    return;
+  size--;

[phoglund, 2016/06/15 08:20:25]

Why is size decreased here? Because you're stepping one byte into data?

It took me a fair amount of thinking to puzzle out what this code does, do maybe
it can be made a bit more readable. What about

if (size <= 1)
  return;

// We decide which header extensions to register by reading one byte
// from the beginning of |data| and interpreting it as a bitmask over
// the RTPExtensionType enum.
std::bitset<8> extensionMask(data[0]);
data++;
size--;

...

It's one line more of code but it's a lot easier to grok we're reading
the first byte and then discarding it. I don't have to reason about the
difference between prefix and suffix ++. Also the comment makes more
sense over that operation rather than the guard clause at the top.

[katrielc, 2016/06/15 09:05:30]

Done.

+  std::bitset<8> extensionMask((data++)[0]);
+
+  RtpPacketReceived::ExtensionManager extensions;
+  for (int i = 0; i < kNumberOfRtpHeaderExtensions; i++)
+    if (extensionMask[i])

[phoglund, 2016/06/15 08:20:25]

Nit: use { } since body is more than one line (even if one is a comment)

[katrielc, 2016/06/15 09:05:30]

Done.

+      // We use i as the ID; it's used in negotiation so not relevant.
+      extensions.Register(static_cast<RTPExtensionType>(i), i);
+
+  RTPHeader rtp_header;
+  RtpUtility::RtpHeaderParser rtp_parser(data, size);
+  rtp_parser.Parse(&rtp_header, &extensions);
+}
+}  // namespace webrtc