Add support for VP9 packetization/depacketization.

webrtc/modules/rtp_rtcp/source/rtp_format_vp9_unittest.cc

+/*
+ *  Copyright (c) 2015 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 <vector>
+
+#include "testing/gmock/include/gmock/gmock.h"
+#include "testing/gtest/include/gtest/gtest.h"
+#include "webrtc/modules/rtp_rtcp/source/rtp_format_vp9.h"
+#include "webrtc/typedefs.h"
+
+namespace webrtc {
+namespace {
+void VerifyHeader(const RTPVideoHeaderVP9& expected,
+                  const RTPVideoHeaderVP9& actual) {
+  EXPECT_EQ(expected.interLayerPredicted, actual.interLayerPredicted);
+  EXPECT_EQ(expected.interPicPredicted, actual.interPicPredicted);
+  EXPECT_EQ(expected.flexibleMode, actual.flexibleMode);
+  EXPECT_EQ(expected.beginningOfFrame, actual.beginningOfFrame);
+  EXPECT_EQ(expected.endOfFrame, actual.endOfFrame);
+  EXPECT_EQ(expected.ssDataAvailable, actual.ssDataAvailable);
+  EXPECT_EQ(expected.pictureId, actual.pictureId);
+  EXPECT_EQ(expected.maxPictureId, actual.maxPictureId);
+  EXPECT_EQ(expected.temporalIdx == kNoTemporalIdx ? 0 : expected.temporalIdx,
+      actual.temporalIdx);
+  EXPECT_EQ(expected.spatialIdx == kNoSpatialIdx ? 0 : expected.spatialIdx,
+      actual.spatialIdx);
+  EXPECT_EQ(expected.gofIdx, actual.gofIdx);
+  EXPECT_EQ(expected.tl0PicIdx, actual.tl0PicIdx);
+  EXPECT_EQ(expected.temporalUpSwitch, actual.temporalUpSwitch);
+
+  EXPECT_EQ(expected.numRefPics, actual.numRefPics);
+  for (uint8_t i = 0; i < expected.numRefPics; ++i) {
+    EXPECT_EQ(expected.pidDiff[i], actual.pidDiff[i]);
+  }
+  if (expected.ssDataAvailable) {
+    EXPECT_EQ(expected.spatialLayerResolutionPresent,
+              actual.spatialLayerResolutionPresent);
+    EXPECT_EQ(expected.numSpatialLayers, actual.numSpatialLayers);
+    if (expected.spatialLayerResolutionPresent) {
+      for (size_t i = 0; i < expected.numSpatialLayers; i++) {
+        EXPECT_EQ(expected.width[i], actual.width[i]);
+        EXPECT_EQ(expected.height[i], actual.height[i]);
+      }
+    }
+    EXPECT_EQ(expected.gof.numFramesInGof, actual.gof.numFramesInGof);
+    for (size_t i = 0; i < expected.gof.numFramesInGof; i++) {
+      EXPECT_EQ(expected.gof.temporalUpSwitch[i],
+                actual.gof.temporalUpSwitch[i]);
+      EXPECT_EQ(expected.gof.temporalIdx[i], actual.gof.temporalIdx[i]);
+      EXPECT_EQ(expected.gof.numRefPics[i], actual.gof.numRefPics[i]);
+      for (size_t j = 0; j < expected.gof.numRefPics[i]; j++) {
+        EXPECT_EQ(expected.gof.pidDiff[i][j], actual.gof.pidDiff[i][j]);
+      }
+    }
+  }
+}
+
+void VerifyPayload(const RtpDepacketizer::ParsedPayload& parsed,
+                   const uint8_t* payload,
+                   size_t payload_length) {
+  EXPECT_EQ(payload, parsed.payload);
+  EXPECT_EQ(payload_length, parsed.payload_length);
+  EXPECT_THAT(std::vector<uint8_t>(parsed.payload,
+                                   parsed.payload + parsed.payload_length),
+              ::testing::ElementsAreArray(payload, payload_length));
+}
+
+void ParseAndCheckPacket(const uint8_t* packet,
+                         const RTPVideoHeaderVP9& expected,
+                         size_t expected_hdr_length,
+                         size_t expected_length) {
+  rtc::scoped_ptr<RtpDepacketizer> depacketizer(new RtpDepacketizerVp9());
+  RtpDepacketizer::ParsedPayload parsed;
+  ASSERT_TRUE(depacketizer->Parse(&parsed, packet, expected_length));
+  // TODO(asapersson): Add when type is added.
+  // EXPECT_EQ(kRtpVideoVp9, parsed.type.Video.codec);
+  VerifyHeader(expected, parsed.type.Video.codecHeader.VP9);
+  const size_t kExpectedPayloadLength = expected_length - expected_hdr_length;
+  VerifyPayload(parsed, packet + expected_hdr_length, kExpectedPayloadLength);
+}
+}  // namespace
+
+// Payload descriptor for flexible mode
+//        0 1 2 3 4 5 6 7
+//        +-+-+-+-+-+-+-+-+
+//        |I|P|L|F|B|E|V|-| (REQUIRED)
+//        +-+-+-+-+-+-+-+-+
+//   I:   |M| PICTURE ID  | (RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+
+//   M:   | EXTENDED PID  | (RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+
+//   L:   |  T  |U|  S  |D| (CONDITIONALLY RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+                             -|
+//   P,F: | P_DIFF    |X|N| (CONDITIONALLY RECOMMENDED)  .
+//        +-+-+-+-+-+-+-+-+                              . up to 3 times
+//   X:   |EXTENDED P_DIFF| (OPTIONAL)                   .
+//        +-+-+-+-+-+-+-+-+                             -|
+//   V:   | SS            |
+//        | ..            |
+//        +-+-+-+-+-+-+-+-+
+//
+// Payload descriptor for non-flexible mode
+//        0 1 2 3 4 5 6 7
+//        +-+-+-+-+-+-+-+-+
+//        |I|P|L|F|B|E|V|-| (REQUIRED)
+//        +-+-+-+-+-+-+-+-+
+//   I:   |M| PICTURE ID  | (RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+
+//   M:   | EXTENDED PID  | (RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+
+//   L:   |GOF_IDX|  S  |D| (CONDITIONALLY RECOMMENDED)
+//        +-+-+-+-+-+-+-+-+
+//        |   TL0PICIDX   | (CONDITIONALLY REQUIRED)
+//        +-+-+-+-+-+-+-+-+
+//   V:   | SS            |
+//        | ..            |
+//        +-+-+-+-+-+-+-+-+
+
+class RtpPacketizerVp9Test : public ::testing::Test {
+ protected:
+  RtpPacketizerVp9Test() {}
+  virtual void SetUp() {
+    expected_.InitRTPVideoHeaderVP9();
+    // Always input one layer frame at a time.
+    expected_.beginningOfFrame = true;
+    expected_.endOfFrame = true;
+  }
+
+  rtc::scoped_ptr<uint8_t[]> packet_;
+  rtc::scoped_ptr<uint8_t[]> payload_;
+  size_t payload_size_;
+  size_t payload_pos_;
+  RTPVideoHeaderVP9 expected_;
+  rtc::scoped_ptr<RtpPacketizerVp9> packetizer_;
+
+  void Init(size_t payload_size, size_t packet_size) {
+    payload_.reset(new uint8_t[payload_size]);
+    memset(payload_.get(), 7, payload_size);
+    payload_size_ = payload_size;
+    payload_pos_ = 0;
+    packetizer_.reset(new RtpPacketizerVp9(expected_, packet_size));
+    packetizer_->SetPayloadData(payload_.get(), payload_size_, NULL);
+
+    const int kMaxPayloadDescriptorLength = 100;
+    packet_.reset(new uint8_t[payload_size_ + kMaxPayloadDescriptorLength]);
+  }
+
+  void CheckPayload(const uint8_t* packet,
+                    size_t start_pos,
+                    size_t end_pos,
+                    bool last) {
+    for (size_t i = start_pos; i < end_pos; ++i) {
+      EXPECT_EQ(packet[i], payload_[payload_pos_++]);
+    }
+    EXPECT_EQ(last, payload_pos_ == payload_size_);
+  }
+
+  void CreateParseAndCheckPackets(const size_t* expected_hdr_sizes,
+                                  const size_t* expected_sizes,
+                                  size_t expected_num_packets) {
+    ASSERT_TRUE(packetizer_.get() != NULL);
+    size_t length = 0;
+    bool last = false;
+    if (expected_num_packets == 0) {
+      EXPECT_FALSE(packetizer_->NextPacket(packet_.get(), &length, &last));
+      return;
+    }
+    for (size_t i = 0; i < expected_num_packets; ++i) {
+      EXPECT_TRUE(packetizer_->NextPacket(packet_.get(), &length, &last));
+      EXPECT_EQ(expected_sizes[i], length);
+      RTPVideoHeaderVP9 hdr = expected_;
+      hdr.beginningOfFrame = (i == 0);
+      hdr.endOfFrame = last;
+      ParseAndCheckPacket(packet_.get(), hdr, expected_hdr_sizes[i], length);
+      CheckPayload(packet_.get(), expected_hdr_sizes[i], length, last);
+    }
+    EXPECT_TRUE(last);
+  }
+};
+
+TEST_F(RtpPacketizerVp9Test, TestEqualSizedMode_OnePacket) {
+  const size_t kFrameSize = 25;
+  const size_t kPacketSize = 26;
+  Init(kFrameSize, kPacketSize);
+
+  // One packet:
+  // I:0, P:0, L:0, F:0, B:1, E:1, V:0  (1hdr + 25 payload)
+  const size_t kExpectedHdrSizes[] = {1};
+  const size_t kExpectedSizes[] = {26};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestEqualSizedMode_TwoPackets) {
+  const size_t kFrameSize = 27;
+  const size_t kPacketSize = 27;
+  Init(kFrameSize, kPacketSize);
+
+  // Two packets:
+  // I:0, P:0, L:0, F:0, B:1, E:0, V:0  (1hdr + 14 payload)
+  // I:0, P:0, L:0, F:0, B:0, E:1, V:0  (1hdr + 13 payload)
+  const size_t kExpectedHdrSizes[] = {1, 1};
+  const size_t kExpectedSizes[] = {15, 14};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestTooShortBufferToFitPayload) {
+  const size_t kFrameSize = 1;
+  const size_t kPacketSize = 1;
+  Init(kFrameSize, kPacketSize);  // 1hdr + 1 payload
+
+  const size_t kExpectedNum = 0;
+  CreateParseAndCheckPackets(NULL, NULL, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestOneBytePictureId) {
+  const size_t kFrameSize = 30;
+  const size_t kPacketSize = 12;
+
+  expected_.pictureId = kMaxOneBytePictureId;   // 2 byte payload descriptor
+  expected_.maxPictureId = kMaxOneBytePictureId;
+  Init(kFrameSize, kPacketSize);
+
+  // Three packets:
+  // I:1, P:0, L:0, F:0, B:1, E:0, V:0 (2hdr + 10 payload)
+  // I:1, P:0, L:0, F:0, B:0, E:0, V:0 (2hdr + 10 payload)
+  // I:1, P:0, L:0, F:0, B:0, E:1, V:0 (2hdr + 10 payload)
+  const size_t kExpectedHdrSizes[] = {2, 2, 2};
+  const size_t kExpectedSizes[] = {12, 12, 12};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestTwoBytePictureId) {
+  const size_t kFrameSize = 31;
+  const size_t kPacketSize = 13;
+
+  expected_.pictureId = kMaxTwoBytePictureId;  // 3 byte payload descriptor
+  Init(kFrameSize, kPacketSize);
+
+  // Four packets:
+  // I:1, P:0, L:0, F:0, B:1, E:0, V:0 (3hdr + 8 payload)
+  // I:1, P:0, L:0, F:0, B:0, E:0, V:0 (3hdr + 8 payload)
+  // I:1, P:0, L:0, F:0, B:0, E:0, V:0 (3hdr + 8 payload)
+  // I:1, P:0, L:0, F:0, B:0, E:1, V:0 (3hdr + 7 payload)
+  const size_t kExpectedHdrSizes[] = {3, 3, 3, 3};
+  const size_t kExpectedSizes[] = {11, 11, 11, 10};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestLayerInfoWithNonFlexibleMode) {
+  const size_t kFrameSize = 30;
+  const size_t kPacketSize = 25;
+
+  expected_.gofIdx = 3;
+  expected_.spatialIdx = 2;
+  expected_.interLayerPredicted = true;  // D
+  expected_.tl0PicIdx = 117;
+  Init(kFrameSize, kPacketSize);
+
+  // Two packets:
+  //    | I:0, P:0, L:1, F:0, B:1, E:0, V:0 | (3hdr + 15 payload)
+  // L: | GOF_IDX:3, S:2, D:1 | TL0PICIDX:117 |
+  //    | I:0, P:0, L:1, F:0, B:0, E:1, V:0 | (3hdr + 15 payload)
+  // L: | GOF_IDX:3, S:2, D:1 | TL0PICIDX:117 |
+  const size_t kExpectedHdrSizes[] = {3, 3};
+  const size_t kExpectedSizes[] = {18, 18};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestLayerInfoWithFlexibleMode) {
+  const size_t kFrameSize = 21;
+  const size_t kPacketSize = 23;
+
+  expected_.flexibleMode = true;
+  expected_.temporalIdx = 3;
+  expected_.temporalUpSwitch = true;  // U
+  expected_.spatialIdx = 2;
+  expected_.interLayerPredicted = false;  // D
+  Init(kFrameSize, kPacketSize);
+
+  // One packet:
+  // I:0, P:0, L:1, F:1, B:1, E:1, V:0 (2hdr + 21 payload)
+  // L:   T:3, U:1, S:2, D:0
+  const size_t kExpectedHdrSizes[] = {2};
+  const size_t kExpectedSizes[] = {23};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestRefIdx) {
+  const size_t kFrameSize = 16;
+  const size_t kPacketSize = 22;
+
+  expected_.interPicPredicted = true;  // P
+  expected_.flexibleMode = true;       // F
+  expected_.pictureId = 100;
+  expected_.numRefPics = 2;
+  expected_.pidDiff[0] = 3;
+  expected_.pidDiff[1] = 1171;
+  Init(kFrameSize, kPacketSize);
+
+  // Two packets:
+  // I:1, P:1, L:0, F:1, B:1, E:1, V:0 (6hdr + 16 payload)
+  // I:   100 (2 bytes)
+  // P,F: P_DIFF:3, X:0, N:1
+  //      P_DIFF:1171, X:1, N:0 (2 bytes)
+  const size_t kExpectedHdrSizes[] = {6};
+  const size_t kExpectedSizes[] = {22};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestRefIdxFailsWithoutPictureId) {
+  const size_t kFrameSize = 16;
+  const size_t kPacketSize = 22;
+
+  expected_.interPicPredicted = true;
+  expected_.flexibleMode = true;
+  expected_.numRefPics = 1;
+  expected_.pidDiff[0] = 3;
+  Init(kFrameSize, kPacketSize);
+
+  const size_t kExpectedNum = 0;
+  CreateParseAndCheckPackets(NULL, NULL, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestSsDataWithoutSpatialResolutionPresent) {
+  const size_t kFrameSize = 21;
+  const size_t kPacketSize = 25;
+
+  expected_.ssDataAvailable = true;
+  expected_.numSpatialLayers = 1;
+  expected_.spatialLayerResolutionPresent = false;
+  expected_.gof.numFramesInGof = 1;
+  expected_.gof.temporalIdx[0] = 0;
+  expected_.gof.temporalUpSwitch[0] = true;
+  expected_.gof.numRefPics[0] = 1;
+  expected_.gof.pidDiff[0][0] = 4;
+  Init(kFrameSize, kPacketSize);
+
+  // One packet:
+  // I:0, P:0, L:0, F:0, B:1, E:1, V:1 (4hdr + 21 payload)
+  // N_S:0, Y:0, N_G:0
+  // T:0, U:1, R:1 | P_DIFF[0][0]:4
+  const size_t kExpectedHdrSizes[] = {4};
+  const size_t kExpectedSizes[] = {25};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+TEST_F(RtpPacketizerVp9Test, TestSsData) {
+  const size_t kFrameSize = 21;
+  const size_t kPacketSize = 39;
+
+  expected_.ssDataAvailable = true;
+  expected_.numSpatialLayers = 2;
+  expected_.spatialLayerResolutionPresent = true;
+  expected_.width[0] = 640;
+  expected_.width[1] = 1280;
+  expected_.height[0] = 360;
+  expected_.height[1] = 720;
+  expected_.gof.numFramesInGof = 3;
+  expected_.gof.temporalIdx[0] = 0;
+  expected_.gof.temporalIdx[1] = 1;
+  expected_.gof.temporalIdx[2] = 2;
+  expected_.gof.temporalUpSwitch[0] = true;
+  expected_.gof.temporalUpSwitch[1] = true;
+  expected_.gof.temporalUpSwitch[2] = false;
+  expected_.gof.numRefPics[0] = 0;
+  expected_.gof.numRefPics[1] = 3;
+  expected_.gof.numRefPics[2] = 2;
+  expected_.gof.pidDiff[1][0] = 5;
+  expected_.gof.pidDiff[1][1] = 6;
+  expected_.gof.pidDiff[1][2] = 7;
+  expected_.gof.pidDiff[2][0] = 8;
+  expected_.gof.pidDiff[2][1] = 9;
+  Init(kFrameSize, kPacketSize);
+
+  // One packet:
+  // I:0, P:0, L:0, F:0, B:1, E:1, V:1 (18hdr + 21 payload)
+  // N_S:1, Y:1, N_G:2
+  // WIDTH:640   // 2 bytes
+  // HEIGHT:360  // 2 bytes
+  // WIDTH:1280  // 2 bytes
+  // HEIGHT:720  // 2 bytes
+  // T:0, U:1, R:0
+  // T:1, U:1, R:3 | P_DIFF[1][0]:5 | P_DIFF[1][1]:6 | P_DIFF[1][2]:7
+  // T:2, U:0, R:2 | P_DIFF[2][0]:8 | P_DIFF[2][0]:9
+  const size_t kExpectedHdrSizes[] = {18};
+  const size_t kExpectedSizes[] = {39};
+  const size_t kExpectedNum = GTEST_ARRAY_SIZE_(kExpectedSizes);
+  CreateParseAndCheckPackets(kExpectedHdrSizes, kExpectedSizes, kExpectedNum);
+}
+
+
+class RtpDepacketizerVp9Test : public ::testing::Test {
+ protected:
+  RtpDepacketizerVp9Test()
+      : depacketizer_(new RtpDepacketizerVp9()) {}
+
+  virtual void SetUp() {
+    expected_.InitRTPVideoHeaderVP9();
+  }
+
+  RTPVideoHeaderVP9 expected_;
+  rtc::scoped_ptr<RtpDepacketizer> depacketizer_;
+};
+
+TEST_F(RtpDepacketizerVp9Test, ParseBasicHeader) {
+  const uint8_t kHeaderLength = 1;
+  uint8_t packet[4] = {0};
+  packet[0] = 0x0C;  // I:0 P:0 L:0 F:0 B:1 E:1 V:0 R:0
+  expected_.beginningOfFrame = true;
+  expected_.endOfFrame = true;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseOneBytePictureId) {
+  const uint8_t kHeaderLength = 2;
+  uint8_t packet[10] = {0};
+  packet[0] = 0x80;  // I:1 P:0 L:0 F:0 B:0 E:0 V:0 R:0
+  packet[1] = kMaxOneBytePictureId;
+
+  expected_.pictureId = kMaxOneBytePictureId;
+  expected_.maxPictureId = kMaxOneBytePictureId;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseTwoBytePictureId) {
+  const uint8_t kHeaderLength = 3;
+  uint8_t packet[10] = {0};
+  packet[0] = 0x80;  // I:1 P:0 L:0 F:0 B:0 E:0 V:0 R:0
+  packet[1] = 0x80 | ((kMaxTwoBytePictureId >> 8) & 0x7F);
+  packet[2] = kMaxTwoBytePictureId & 0xFF;
+
+  expected_.pictureId = kMaxTwoBytePictureId;
+  expected_.maxPictureId = kMaxTwoBytePictureId;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseLayerInfoWithNonFlexibleMode) {
+  const uint8_t kHeaderLength = 3;
+  const uint8_t kGofIdx = 7;
+  const uint8_t kSpatialIdx = 1;
+  const uint8_t kDbit = 1;
+  const uint8_t kTl0PicIdx = 17;
+  uint8_t packet[13] = {0};
+  packet[0] = 0x20;  // I:0 P:0 L:1 F:0 B:0 E:0 V:0 R:0
+  packet[1] = (kGofIdx << 4) | (kSpatialIdx << 1) | kDbit;  // GOF_IDX:7 S:1 D:1
+  packet[2] = kTl0PicIdx;                                   // TL0PICIDX:17
+
+  expected_.gofIdx = kGofIdx;
+  expected_.spatialIdx = kSpatialIdx;
+  expected_.interLayerPredicted = kDbit ? true : false;
+  expected_.tl0PicIdx = kTl0PicIdx;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseLayerInfoWithFlexibleMode) {
+  const uint8_t kHeaderLength = 2;
+  const uint8_t kTemporalIdx = 2;
+  const uint8_t kUbit = 1;
+  const uint8_t kSpatialIdx = 0;
+  const uint8_t kDbit = 0;
+  uint8_t packet[13] = {0};
+  packet[0] = 0x38;  // I:0 P:0 L:1 F:1 B:1 E:0 V:0 R:0
+  packet[1] = (kTemporalIdx << 5) | (kUbit << 4) | (kSpatialIdx << 1) | kDbit;
+
+  // I:0 P:0 L:1 F:1 B:1 E:0 V:0
+  // L:   T:2 U:1 S:0 D:0
+  expected_.beginningOfFrame = true;
+  expected_.flexibleMode = true;
+  expected_.temporalIdx = kTemporalIdx;
+  expected_.temporalUpSwitch = kUbit ? true : false;
+  expected_.spatialIdx = kSpatialIdx;
+  expected_.interLayerPredicted = kDbit ? true : false;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseRefIdx) {
+  const uint8_t kHeaderLength = 7;
+  const int16_t kPictureId = 17;
+  const int16_t kPdiff1 = 17;
+  const int16_t kPdiff2 = 18;
+  const int16_t kExtPdiff3 = 2171;
+  uint8_t packet[13] = {0};
+  packet[0] = 0xD8;  // I:1 P:1 L:0 F:1 B:1 E:0 V:0 R:0
+  packet[1] = 0x80 | ((kPictureId >> 8) & 0x7F);  // Two byte pictureID.
+  packet[2] = kPictureId;
+  packet[3] = (kPdiff1 << 2) | (0 << 1) | 1;            // P_DIFF X:0 N:1
+  packet[4] = (kPdiff2 << 2) | (0 << 1) | 1;            // P_DIFF X:0 N:1
+  packet[5] = ((kExtPdiff3 >> 8) << 2) | (1 << 1) | 0;  // P_DIFF X:1 N:0
+  packet[6] = kExtPdiff3 & 0xff;                        // EXTENDED P_DIFF
+
+  // I:1 P:1 L:0 F:1 B:1 E:0 V:0
+  // I:    PICTURE ID:17
+  // I:
+  // P,F:  P_DIFF:17 X:0 N:1   => refPictureId = 17 - 17 = 0
+  // P,F:  P_DIFF:18 X:0 N:1   => refPictureId = 0x7FFF + 1 + 17 - 18 = 0x7FFF
+  // P,F:  P_DIFF:2171 X:1 N:0 => refPictureId = 0x7FFF + 1 + 17 - 2171 = 30614
+  expected_.beginningOfFrame = true;
+  expected_.interPicPredicted = true;
+  expected_.flexibleMode = true;
+  expected_.pictureId = kPictureId;
+  expected_.numRefPics = 3;
+  expected_.pidDiff[0] = kPdiff1;
+  expected_.pidDiff[1] = kPdiff2;
+  expected_.pidDiff[2] = kExtPdiff3;
+  expected_.refPictureId[0] = 0;
+  expected_.refPictureId[1] = 0x7FFF;
+  expected_.refPictureId[2] = 30614;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseRefIdxFailsWithNoPictureId) {
+  const int16_t kPdiff = 3;
+  uint8_t packet[13] = {0};
+  packet[0] = 0x58;                          // I:0 P:1 L:0 F:1 B:1 E:0 V:0 R:0
+  packet[1] = (kPdiff << 2) | (0 << 1) | 0;  // P,F:  P_DIFF:3 X:0 N:0
+
+  RtpDepacketizer::ParsedPayload parsed;
+  EXPECT_FALSE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseRefIdxFailsWithTooManyRefPics) {
+  const int16_t kPdiff = 3;
+  uint8_t packet[13] = {0};
+  packet[0] = 0xD8;                          // I:1 P:1 L:0 F:1 B:1 E:0 V:0 R:0
+  packet[1] = kMaxOneBytePictureId;          // I:    PICTURE ID:127
+  packet[2] = (kPdiff << 2) | (0 << 1) | 1;  // P,F:  P_DIFF:3 X:0 N:1
+  packet[3] = (kPdiff << 2) | (0 << 1) | 1;  // P,F:  P_DIFF:3 X:0 N:1
+  packet[4] = (kPdiff << 2) | (0 << 1) | 1;  // P,F:  P_DIFF:3 X:0 N:1
+  packet[5] = (kPdiff << 2) | (0 << 1) | 0;  // P,F:  P_DIFF:3 X:0 N:0
+
+  RtpDepacketizer::ParsedPayload parsed;
+  EXPECT_FALSE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseSsData) {
+  const uint8_t kHeaderLength = 5;
+  const uint8_t kYbit = 0;
+  const size_t kNs = 2;
+  const size_t kNg = 2;
+  uint8_t packet[23] = {0};
+  packet[0] = 0x0A;  // I:0 P:0 L:0 F:0 B:1 E:0 V:1 R:0
+  packet[1] = ((kNs - 1) << 5) | (kYbit << 4) | (kNg - 1);  // N_S Y N_G
+  packet[2] = (0 << 5) | (1 << 4) | (0 << 2) | 0;           // T:0 U:1 R:0 -
+  packet[3] = (2 << 5) | (0 << 4) | (1 << 2) | 0;           // T:2 U:0 R:1 -
+  packet[4] = 33;
+
+  expected_.beginningOfFrame = true;
+  expected_.ssDataAvailable = true;
+  expected_.numSpatialLayers = kNs;
+  expected_.spatialLayerResolutionPresent = kYbit ? true : false;
+  expected_.gof.numFramesInGof = kNg;
+  expected_.gof.temporalIdx[0] = 0;
+  expected_.gof.temporalIdx[1] = 2;
+  expected_.gof.temporalUpSwitch[0] = true;
+  expected_.gof.temporalUpSwitch[1] = false;
+  expected_.gof.numRefPics[0] = 0;
+  expected_.gof.numRefPics[1] = 1;
+  expected_.gof.pidDiff[1][0] = 33;
+  ParseAndCheckPacket(packet, expected_, kHeaderLength, sizeof(packet));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseFirstPacketInKeyFrame) {
+  uint8_t packet[2] = {0};
+  packet[0] = 0x08;  // I:0 P:0 L:0 F:0 B:1 E:0 V:0 R:0
+
+  RtpDepacketizer::ParsedPayload parsed;
+  ASSERT_TRUE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+  EXPECT_EQ(kVideoFrameKey, parsed.frame_type);
+  EXPECT_TRUE(parsed.type.Video.isFirstPacket);
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseLastPacketInDeltaFrame) {
+  uint8_t packet[2] = {0};
+  packet[0] = 0x44;  // I:0 P:1 L:0 F:0 B:0 E:1 V:0 R:0
+
+  RtpDepacketizer::ParsedPayload parsed;
+  ASSERT_TRUE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+  EXPECT_EQ(kVideoFrameDelta, parsed.frame_type);
+  EXPECT_FALSE(parsed.type.Video.isFirstPacket);
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseResolution) {
+  const uint16_t kWidth[2] = {640, 1280};
+  const uint16_t kHeight[2] = {360, 720};
+  uint8_t packet[20] = {0};
+  packet[0] = 0x0A;  // I:0 P:0 L:0 F:0 B:1 E:0 V:1 R:0
+  packet[1] = (1 << 5) | (1 << 4) | 0;  // N_S:1 Y:1 N_G:0
+  packet[2] = kWidth[0] >> 8;
+  packet[3] = kWidth[0] & 0xFF;
+  packet[4] = kHeight[0] >> 8;
+  packet[5] = kHeight[0] & 0xFF;
+  packet[6] = kWidth[1] >> 8;
+  packet[7] = kWidth[1] & 0xFF;
+  packet[8] = kHeight[1] >> 8;
+  packet[9] = kHeight[1] & 0xFF;
+  packet[10] = 0;  // T:0 U:0 R:0 -
+
+  RtpDepacketizer::ParsedPayload parsed;
+  ASSERT_TRUE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+  EXPECT_EQ(kWidth[0], parsed.type.Video.width);
+  EXPECT_EQ(kHeight[0], parsed.type.Video.height);
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseFailsForNoPayloadLength) {
+  uint8_t packet[1] = {0};
+  RtpDepacketizer::ParsedPayload parsed;
+  EXPECT_FALSE(depacketizer_->Parse(&parsed, packet, 0));
+}
+
+TEST_F(RtpDepacketizerVp9Test, ParseFailsForTooShortBufferToFitPayload) {
+  const uint8_t kHeaderLength = 1;
+  uint8_t packet[kHeaderLength] = {0};
+  RtpDepacketizer::ParsedPayload parsed;
+  EXPECT_FALSE(depacketizer_->Parse(&parsed, packet, sizeof(packet)));
+}
+
+}  // namespace webrtc