red: do not generate packets which are > 1200 bytes
and do not generate redundancy for packets that are larger
than 1024 bytes which is the maximum size red can encode.
Bug: webrtc:11640
Change-Id: I211cb196eee2a0659f22a601a6dee4b7dd4e5116
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/178781
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31846}
diff --git a/modules/audio_coding/codecs/red/audio_encoder_copy_red.cc b/modules/audio_coding/codecs/red/audio_encoder_copy_red.cc
index 2bfd2c4..1432e31 100644
--- a/modules/audio_coding/codecs/red/audio_encoder_copy_red.cc
+++ b/modules/audio_coding/codecs/red/audio_encoder_copy_red.cc
@@ -19,7 +19,10 @@
#include "rtc_base/checks.h"
namespace webrtc {
-static const int kRedMaxPacketSize = 1 << 10;
+// RED packets must be less than 1024 bytes to fit the 10 bit block length.
+static constexpr const int kRedMaxPacketSize = 1 << 10;
+// The typical MTU is 1200 bytes.
+static constexpr const size_t kAudioMaxRtpPacketLen = 1200;
AudioEncoderCopyRed::Config::Config() = default;
AudioEncoderCopyRed::Config::Config(Config&&) = default;
@@ -27,6 +30,7 @@
AudioEncoderCopyRed::AudioEncoderCopyRed(Config&& config)
: speech_encoder_(std::move(config.speech_encoder)),
+ max_packet_length_(kAudioMaxRtpPacketLen),
red_payload_type_(config.payload_type) {
RTC_CHECK(speech_encoder_) << "Speech encoder not provided.";
}
@@ -57,12 +61,16 @@
return speech_encoder_->GetTargetBitrate();
}
-size_t AudioEncoderCopyRed::CalculateHeaderLength() const {
+size_t AudioEncoderCopyRed::CalculateHeaderLength(size_t encoded_bytes) const {
size_t header_size = 1;
- if (secondary_info_.encoded_bytes > 0) {
+ size_t bytes_available = max_packet_length_ - encoded_bytes;
+ if (secondary_info_.encoded_bytes > 0 &&
+ secondary_info_.encoded_bytes < bytes_available) {
header_size += 4;
+ bytes_available -= secondary_info_.encoded_bytes;
}
- if (tertiary_info_.encoded_bytes > 0) {
+ if (tertiary_info_.encoded_bytes > 0 &&
+ tertiary_info_.encoded_bytes < bytes_available) {
header_size += 4;
}
return header_size > 1 ? header_size : 0;
@@ -78,19 +86,22 @@
RTC_CHECK(info.redundant.empty()) << "Cannot use nested redundant encoders.";
RTC_DCHECK_EQ(primary_encoded.size(), info.encoded_bytes);
- if (info.encoded_bytes == 0) {
+ if (info.encoded_bytes == 0 || info.encoded_bytes > kRedMaxPacketSize) {
return info;
}
+ RTC_DCHECK_GT(max_packet_length_, info.encoded_bytes);
// Allocate room for RFC 2198 header if there is redundant data.
// Otherwise this will send the primary payload type without
// wrapping in RED.
- const size_t header_length_bytes = CalculateHeaderLength();
+ const size_t header_length_bytes = CalculateHeaderLength(info.encoded_bytes);
encoded->SetSize(header_length_bytes);
size_t header_offset = 0;
+ size_t bytes_available = max_packet_length_ - info.encoded_bytes;
if (tertiary_info_.encoded_bytes > 0 &&
- tertiary_info_.encoded_bytes < kRedMaxPacketSize) {
+ tertiary_info_.encoded_bytes + secondary_info_.encoded_bytes <
+ bytes_available) {
encoded->AppendData(tertiary_encoded_);
const uint32_t timestamp_delta =
@@ -101,10 +112,11 @@
(timestamp_delta << 2) | (tertiary_info_.encoded_bytes >> 8));
encoded->data()[header_offset + 3] = tertiary_info_.encoded_bytes & 0xff;
header_offset += 4;
+ bytes_available -= tertiary_info_.encoded_bytes;
}
if (secondary_info_.encoded_bytes > 0 &&
- secondary_info_.encoded_bytes < kRedMaxPacketSize) {
+ secondary_info_.encoded_bytes < bytes_available) {
encoded->AppendData(secondary_encoded_);
const uint32_t timestamp_delta =
@@ -115,6 +127,7 @@
(timestamp_delta << 2) | (secondary_info_.encoded_bytes >> 8));
encoded->data()[header_offset + 3] = secondary_info_.encoded_bytes & 0xff;
header_offset += 4;
+ bytes_available -= secondary_info_.encoded_bytes;
}
encoded->AppendData(primary_encoded);
@@ -200,4 +213,9 @@
return speech_encoder_->GetFrameLengthRange();
}
+void AudioEncoderCopyRed::OnReceivedOverhead(size_t overhead_bytes_per_packet) {
+ max_packet_length_ = kAudioMaxRtpPacketLen - overhead_bytes_per_packet;
+ return speech_encoder_->OnReceivedOverhead(overhead_bytes_per_packet);
+}
+
} // namespace webrtc
diff --git a/modules/audio_coding/codecs/red/audio_encoder_copy_red.h b/modules/audio_coding/codecs/red/audio_encoder_copy_red.h
index 4d7fc40..9806772 100644
--- a/modules/audio_coding/codecs/red/audio_encoder_copy_red.h
+++ b/modules/audio_coding/codecs/red/audio_encoder_copy_red.h
@@ -62,6 +62,7 @@
void OnReceivedUplinkBandwidth(
int target_audio_bitrate_bps,
absl::optional<int64_t> bwe_period_ms) override;
+ void OnReceivedOverhead(size_t overhead_bytes_per_packet) override;
absl::optional<std::pair<TimeDelta, TimeDelta>> GetFrameLengthRange()
const override;
@@ -71,13 +72,16 @@
rtc::Buffer* encoded) override;
private:
- size_t CalculateHeaderLength() const;
+ size_t CalculateHeaderLength(size_t encoded_bytes) const;
+
std::unique_ptr<AudioEncoder> speech_encoder_;
+ size_t max_packet_length_;
int red_payload_type_;
rtc::Buffer secondary_encoded_;
EncodedInfoLeaf secondary_info_;
rtc::Buffer tertiary_encoded_;
EncodedInfoLeaf tertiary_info_;
+
RTC_DISALLOW_COPY_AND_ASSIGN(AudioEncoderCopyRed);
};
diff --git a/modules/audio_coding/codecs/red/audio_encoder_copy_red_unittest.cc b/modules/audio_coding/codecs/red/audio_encoder_copy_red_unittest.cc
index fbc0b8a..3352799 100644
--- a/modules/audio_coding/codecs/red/audio_encoder_copy_red_unittest.cc
+++ b/modules/audio_coding/codecs/red/audio_encoder_copy_red_unittest.cc
@@ -355,6 +355,32 @@
EXPECT_EQ(encoded_[8], primary_payload_type);
}
+TEST_F(AudioEncoderCopyRedTest, RespectsPayloadMTU) {
+ const int primary_payload_type = red_payload_type_ + 1;
+ AudioEncoder::EncodedInfo info;
+ info.encoded_bytes = 600;
+ info.encoded_timestamp = timestamp_;
+ info.payload_type = primary_payload_type;
+
+ EXPECT_CALL(*mock_encoder_, EncodeImpl(_, _, _))
+ .WillOnce(Invoke(MockAudioEncoder::FakeEncoding(info)));
+ Encode();
+ info.encoded_timestamp = timestamp_; // update timestamp.
+ info.encoded_bytes = 500;
+ EXPECT_CALL(*mock_encoder_, EncodeImpl(_, _, _))
+ .WillOnce(Invoke(MockAudioEncoder::FakeEncoding(info)));
+ Encode(); // Second call will produce a redundant encoding.
+
+ EXPECT_EQ(encoded_.size(), 5u + 600u + 500u);
+
+ info.encoded_timestamp = timestamp_; // update timestamp.
+ info.encoded_bytes = 400;
+ EXPECT_CALL(*mock_encoder_, EncodeImpl(_, _, _))
+ .WillOnce(Invoke(MockAudioEncoder::FakeEncoding(info)));
+ Encode(); // Third call will drop the oldest packet.
+ EXPECT_EQ(encoded_.size(), 5u + 500u + 400u);
+}
+
#if GTEST_HAS_DEATH_TEST && !defined(WEBRTC_ANDROID)
// This test fixture tests various error conditions that makes the