/* * Copyright (c) 2021 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 "net/dcsctp/tx/retransmission_queue.h" #include #include #include #include #include #include #include #include #include #include "absl/algorithm/container.h" #include "absl/strings/string_view.h" #include "absl/types/optional.h" #include "api/array_view.h" #include "net/dcsctp/common/math.h" #include "net/dcsctp/common/sequence_numbers.h" #include "net/dcsctp/common/str_join.h" #include "net/dcsctp/packet/chunk/data_chunk.h" #include "net/dcsctp/packet/chunk/forward_tsn_chunk.h" #include "net/dcsctp/packet/chunk/forward_tsn_common.h" #include "net/dcsctp/packet/chunk/idata_chunk.h" #include "net/dcsctp/packet/chunk/iforward_tsn_chunk.h" #include "net/dcsctp/packet/chunk/sack_chunk.h" #include "net/dcsctp/packet/data.h" #include "net/dcsctp/public/dcsctp_options.h" #include "net/dcsctp/public/types.h" #include "net/dcsctp/timer/timer.h" #include "net/dcsctp/tx/outstanding_data.h" #include "net/dcsctp/tx/send_queue.h" #include "rtc_base/checks.h" #include "rtc_base/logging.h" #include "rtc_base/strings/string_builder.h" namespace dcsctp { namespace { // Allow sending only slightly less than an MTU, to account for headers. constexpr float kMinBytesRequiredToSendFactor = 0.9; } // namespace RetransmissionQueue::RetransmissionQueue( absl::string_view log_prefix, TSN my_initial_tsn, size_t a_rwnd, SendQueue& send_queue, std::function on_new_rtt, std::function on_clear_retransmission_counter, Timer& t3_rtx, const DcSctpOptions& options, bool supports_partial_reliability, bool use_message_interleaving, const DcSctpSocketHandoverState* handover_state) : options_(options), min_bytes_required_to_send_(options.mtu * kMinBytesRequiredToSendFactor), partial_reliability_(supports_partial_reliability), log_prefix_(std::string(log_prefix) + "tx: "), data_chunk_header_size_(use_message_interleaving ? IDataChunk::kHeaderSize : DataChunk::kHeaderSize), on_new_rtt_(std::move(on_new_rtt)), on_clear_retransmission_counter_( std::move(on_clear_retransmission_counter)), t3_rtx_(t3_rtx), cwnd_(handover_state ? handover_state->tx.cwnd : options_.cwnd_mtus_initial * options_.mtu), rwnd_(handover_state ? handover_state->tx.rwnd : a_rwnd), // https://tools.ietf.org/html/rfc4960#section-7.2.1 // "The initial value of ssthresh MAY be arbitrarily high (for // example, implementations MAY use the size of the receiver advertised // window)."" ssthresh_(handover_state ? handover_state->tx.ssthresh : rwnd_), partial_bytes_acked_( handover_state ? handover_state->tx.partial_bytes_acked : 0), send_queue_(send_queue), outstanding_data_( data_chunk_header_size_, tsn_unwrapper_.Unwrap(handover_state ? TSN(handover_state->tx.next_tsn) : my_initial_tsn), tsn_unwrapper_.Unwrap(handover_state ? TSN(handover_state->tx.next_tsn - 1) : TSN(*my_initial_tsn - 1)), [this](IsUnordered unordered, StreamID stream_id, MID message_id) { return send_queue_.Discard(unordered, stream_id, message_id); }) {} bool RetransmissionQueue::IsConsistent() const { return true; } // Returns how large a chunk will be, serialized, carrying the data size_t RetransmissionQueue::GetSerializedChunkSize(const Data& data) const { return RoundUpTo4(data_chunk_header_size_ + data.size()); } void RetransmissionQueue::MaybeExitFastRecovery( UnwrappedTSN cumulative_tsn_ack) { // https://tools.ietf.org/html/rfc4960#section-7.2.4 // "When a SACK acknowledges all TSNs up to and including this [fast // recovery] exit point, Fast Recovery is exited." if (fast_recovery_exit_tsn_.has_value() && cumulative_tsn_ack >= *fast_recovery_exit_tsn_) { RTC_DLOG(LS_VERBOSE) << log_prefix_ << "exit_point=" << *fast_recovery_exit_tsn_->Wrap() << " reached - exiting fast recovery"; fast_recovery_exit_tsn_ = absl::nullopt; } } void RetransmissionQueue::HandleIncreasedCumulativeTsnAck( size_t outstanding_bytes, size_t total_bytes_acked) { // Allow some margin for classifying as fully utilized, due to e.g. that too // small packets (less than kMinimumFragmentedPayload) are not sent + // overhead. bool is_fully_utilized = outstanding_bytes + options_.mtu >= cwnd_; size_t old_cwnd = cwnd_; if (phase() == CongestionAlgorithmPhase::kSlowStart) { if (is_fully_utilized && !is_in_fast_recovery()) { // https://tools.ietf.org/html/rfc4960#section-7.2.1 // "Only when these three conditions are met can the cwnd be // increased; otherwise, the cwnd MUST not be increased. If these // conditions are met, then cwnd MUST be increased by, at most, the // lesser of 1) the total size of the previously outstanding DATA // chunk(s) acknowledged, and 2) the destination's path MTU." cwnd_ += std::min(total_bytes_acked, options_.mtu); RTC_DLOG(LS_VERBOSE) << log_prefix_ << "SS increase cwnd=" << cwnd_ << " (" << old_cwnd << ")"; } } else if (phase() == CongestionAlgorithmPhase::kCongestionAvoidance) { // https://tools.ietf.org/html/rfc4960#section-7.2.2 // "Whenever cwnd is greater than ssthresh, upon each SACK arrival // that advances the Cumulative TSN Ack Point, increase // partial_bytes_acked by the total number of bytes of all new chunks // acknowledged in that SACK including chunks acknowledged by the new // Cumulative TSN Ack and by Gap Ack Blocks." size_t old_pba = partial_bytes_acked_; partial_bytes_acked_ += total_bytes_acked; if (partial_bytes_acked_ >= cwnd_ && is_fully_utilized) { // https://tools.ietf.org/html/rfc4960#section-7.2.2 // "When partial_bytes_acked is equal to or greater than cwnd and // before the arrival of the SACK the sender had cwnd or more bytes of // data outstanding (i.e., before arrival of the SACK, flightsize was // greater than or equal to cwnd), increase cwnd by MTU, and reset // partial_bytes_acked to (partial_bytes_acked - cwnd)." // Errata: https://datatracker.ietf.org/doc/html/rfc8540#section-3.12 partial_bytes_acked_ -= cwnd_; cwnd_ += options_.mtu; RTC_DLOG(LS_VERBOSE) << log_prefix_ << "CA increase cwnd=" << cwnd_ << " (" << old_cwnd << ") ssthresh=" << ssthresh_ << ", pba=" << partial_bytes_acked_ << " (" << old_pba << ")"; } else { RTC_DLOG(LS_VERBOSE) << log_prefix_ << "CA unchanged cwnd=" << cwnd_ << " (" << old_cwnd << ") ssthresh=" << ssthresh_ << ", pba=" << partial_bytes_acked_ << " (" << old_pba << ")"; } } } void RetransmissionQueue::HandlePacketLoss(UnwrappedTSN highest_tsn_acked) { if (!is_in_fast_recovery()) { // https://tools.ietf.org/html/rfc4960#section-7.2.4 // "If not in Fast Recovery, adjust the ssthresh and cwnd of the // destination address(es) to which the missing DATA chunks were last // sent, according to the formula described in Section 7.2.3." size_t old_cwnd = cwnd_; size_t old_pba = partial_bytes_acked_; ssthresh_ = std::max(cwnd_ / 2, options_.cwnd_mtus_min * options_.mtu); cwnd_ = ssthresh_; partial_bytes_acked_ = 0; RTC_DLOG(LS_VERBOSE) << log_prefix_ << "packet loss detected (not fast recovery). cwnd=" << cwnd_ << " (" << old_cwnd << "), ssthresh=" << ssthresh_ << ", pba=" << partial_bytes_acked_ << " (" << old_pba << ")"; // https://tools.ietf.org/html/rfc4960#section-7.2.4 // "If not in Fast Recovery, enter Fast Recovery and mark the highest // outstanding TSN as the Fast Recovery exit point." fast_recovery_exit_tsn_ = outstanding_data_.highest_outstanding_tsn(); RTC_DLOG(LS_VERBOSE) << log_prefix_ << "fast recovery initiated with exit_point=" << *fast_recovery_exit_tsn_->Wrap(); } else { // https://tools.ietf.org/html/rfc4960#section-7.2.4 // "While in Fast Recovery, the ssthresh and cwnd SHOULD NOT change for // any destinations due to a subsequent Fast Recovery event (i.e., one // SHOULD NOT reduce the cwnd further due to a subsequent Fast Retransmit)." RTC_DLOG(LS_VERBOSE) << log_prefix_ << "packet loss detected (fast recovery). No changes."; } } void RetransmissionQueue::UpdateReceiverWindow(uint32_t a_rwnd) { rwnd_ = outstanding_data_.outstanding_bytes() >= a_rwnd ? 0 : a_rwnd - outstanding_data_.outstanding_bytes(); } void RetransmissionQueue::StartT3RtxTimerIfOutstandingData() { // Note: Can't use `outstanding_bytes()` as that one doesn't count chunks to // be retransmitted. if (outstanding_data_.empty()) { // https://tools.ietf.org/html/rfc4960#section-6.3.2 // "Whenever all outstanding data sent to an address have been // acknowledged, turn off the T3-rtx timer of that address. // Note: Already stopped in `StopT3RtxTimerOnIncreasedCumulativeTsnAck`." } else { // https://tools.ietf.org/html/rfc4960#section-6.3.2 // "Whenever a SACK is received that acknowledges the DATA chunk // with the earliest outstanding TSN for that address, restart the T3-rtx // timer for that address with its current RTO (if there is still // outstanding data on that address)." // "Whenever a SACK is received missing a TSN that was previously // acknowledged via a Gap Ack Block, start the T3-rtx for the destination // address to which the DATA chunk was originally transmitted if it is not // already running." if (!t3_rtx_.is_running()) { t3_rtx_.Start(); } } } bool RetransmissionQueue::IsSackValid(const SackChunk& sack) const { // https://tools.ietf.org/html/rfc4960#section-6.2.1 // "If Cumulative TSN Ack is less than the Cumulative TSN Ack Point, // then drop the SACK. Since Cumulative TSN Ack is monotonically increasing, // a SACK whose Cumulative TSN Ack is less than the Cumulative TSN Ack Point // indicates an out-of- order SACK." // // Note: Important not to drop SACKs with identical TSN to that previously // received, as the gap ack blocks or dup tsn fields may have changed. UnwrappedTSN cumulative_tsn_ack = tsn_unwrapper_.PeekUnwrap(sack.cumulative_tsn_ack()); if (cumulative_tsn_ack < outstanding_data_.last_cumulative_tsn_ack()) { // https://tools.ietf.org/html/rfc4960#section-6.2.1 // "If Cumulative TSN Ack is less than the Cumulative TSN Ack Point, // then drop the SACK. Since Cumulative TSN Ack is monotonically // increasing, a SACK whose Cumulative TSN Ack is less than the Cumulative // TSN Ack Point indicates an out-of- order SACK." return false; } else if (cumulative_tsn_ack > outstanding_data_.highest_outstanding_tsn()) { return false; } return true; } bool RetransmissionQueue::HandleSack(TimeMs now, const SackChunk& sack) { if (!IsSackValid(sack)) { return false; } UnwrappedTSN old_last_cumulative_tsn_ack = outstanding_data_.last_cumulative_tsn_ack(); size_t old_outstanding_bytes = outstanding_data_.outstanding_bytes(); size_t old_rwnd = rwnd_; UnwrappedTSN cumulative_tsn_ack = tsn_unwrapper_.Unwrap(sack.cumulative_tsn_ack()); if (sack.gap_ack_blocks().empty()) { UpdateRTT(now, cumulative_tsn_ack); } OutstandingData::AckInfo ack_info = outstanding_data_.HandleSack( cumulative_tsn_ack, sack.gap_ack_blocks(), is_in_fast_retransmit_); // Update of outstanding_data_ is now done. Congestion control remains. UpdateReceiverWindow(sack.a_rwnd()); RTC_DLOG(LS_VERBOSE) << log_prefix_ << "Received SACK, cum_tsn_ack=" << *cumulative_tsn_ack.Wrap() << " (" << *old_last_cumulative_tsn_ack.Wrap() << "), outstanding_bytes=" << outstanding_data_.outstanding_bytes() << " (" << old_outstanding_bytes << "), rwnd=" << rwnd_ << " (" << old_rwnd << ")"; MaybeExitFastRecovery(cumulative_tsn_ack); if (cumulative_tsn_ack > old_last_cumulative_tsn_ack) { // https://tools.ietf.org/html/rfc4960#section-6.3.2 // "Whenever a SACK is received that acknowledges the DATA chunk // with the earliest outstanding TSN for that address, restart the T3-rtx // timer for that address with its current RTO (if there is still // outstanding data on that address)." // Note: It may be started again in a bit further down. t3_rtx_.Stop(); HandleIncreasedCumulativeTsnAck(old_outstanding_bytes, ack_info.bytes_acked); } if (ack_info.has_packet_loss) { is_in_fast_retransmit_ = true; HandlePacketLoss(ack_info.highest_tsn_acked); } // https://tools.ietf.org/html/rfc4960#section-8.2 // "When an outstanding TSN is acknowledged [...] the endpoint shall clear // the error counter ..." if (ack_info.bytes_acked > 0) { on_clear_retransmission_counter_(); } StartT3RtxTimerIfOutstandingData(); RTC_DCHECK(IsConsistent()); return true; } void RetransmissionQueue::UpdateRTT(TimeMs now, UnwrappedTSN cumulative_tsn_ack) { // RTT updating is flawed in SCTP, as explained in e.g. Pedersen J, Griwodz C, // Halvorsen P (2006) Considerations of SCTP retransmission delays for thin // streams. // Due to delayed acknowledgement, the SACK may be sent much later which // increases the calculated RTT. // TODO(boivie): Consider occasionally sending DATA chunks with I-bit set and // use only those packets for measurement. absl::optional rtt = outstanding_data_.MeasureRTT(now, cumulative_tsn_ack); if (rtt.has_value()) { on_new_rtt_(*rtt); } } void RetransmissionQueue::HandleT3RtxTimerExpiry() { size_t old_cwnd = cwnd_; size_t old_outstanding_bytes = outstanding_bytes(); // https://tools.ietf.org/html/rfc4960#section-6.3.3 // "For the destination address for which the timer expires, adjust // its ssthresh with rules defined in Section 7.2.3 and set the cwnd <- MTU." ssthresh_ = std::max(cwnd_ / 2, 4 * options_.mtu); cwnd_ = 1 * options_.mtu; // Errata: https://datatracker.ietf.org/doc/html/rfc8540#section-3.11 partial_bytes_acked_ = 0; // https://tools.ietf.org/html/rfc4960#section-6.3.3 // "For the destination address for which the timer expires, set RTO // <- RTO * 2 ("back off the timer"). The maximum value discussed in rule C7 // above (RTO.max) may be used to provide an upper bound to this doubling // operation." // Already done by the Timer implementation. // https://tools.ietf.org/html/rfc4960#section-6.3.3 // "Determine how many of the earliest (i.e., lowest TSN) outstanding // DATA chunks for the address for which the T3-rtx has expired will fit into // a single packet" // https://tools.ietf.org/html/rfc4960#section-6.3.3 // "Note: Any DATA chunks that were sent to the address for which the // T3-rtx timer expired but did not fit in one MTU (rule E3 above) should be // marked for retransmission and sent as soon as cwnd allows (normally, when a // SACK arrives)." outstanding_data_.NackAll(); // https://tools.ietf.org/html/rfc4960#section-6.3.3 // "Start the retransmission timer T3-rtx on the destination address // to which the retransmission is sent, if rule R1 above indicates to do so." // Already done by the Timer implementation. RTC_DLOG(LS_INFO) << log_prefix_ << "t3-rtx expired. new cwnd=" << cwnd_ << " (" << old_cwnd << "), ssthresh=" << ssthresh_ << ", outstanding_bytes " << outstanding_bytes() << " (" << old_outstanding_bytes << ")"; RTC_DCHECK(IsConsistent()); } std::vector> RetransmissionQueue::GetChunksToSend( TimeMs now, size_t bytes_remaining_in_packet) { // Chunks are always padded to even divisible by four. RTC_DCHECK(IsDivisibleBy4(bytes_remaining_in_packet)); std::vector> to_be_sent; size_t old_outstanding_bytes = outstanding_bytes(); size_t old_rwnd = rwnd_; if (is_in_fast_retransmit_) { // https://tools.ietf.org/html/rfc4960#section-7.2.4 // "Determine how many of the earliest (i.e., lowest TSN) DATA chunks // marked for retransmission will fit into a single packet ... Retransmit // those K DATA chunks in a single packet. When a Fast Retransmit is being // performed, the sender SHOULD ignore the value of cwnd and SHOULD NOT // delay retransmission for this single packet." is_in_fast_retransmit_ = false; to_be_sent = outstanding_data_.GetChunksToBeRetransmitted(bytes_remaining_in_packet); size_t to_be_sent_bytes = absl::c_accumulate( to_be_sent, 0, [&](size_t r, const std::pair& d) { return r + GetSerializedChunkSize(d.second); }); RTC_DLOG(LS_VERBOSE) << log_prefix_ << "fast-retransmit: sending " << to_be_sent.size() << " chunks, " << to_be_sent_bytes << " bytes"; } else { // Normal sending. Calculate the bandwidth budget (how many bytes that is // allowed to be sent), and fill that up first with chunks that are // scheduled to be retransmitted. If there is still budget, send new chunks // (which will have their TSN assigned here.) size_t max_bytes = RoundDownTo4(std::min(max_bytes_to_send(), bytes_remaining_in_packet)); to_be_sent = outstanding_data_.GetChunksToBeRetransmitted(max_bytes); max_bytes -= absl::c_accumulate( to_be_sent, 0, [&](size_t r, const std::pair& d) { return r + GetSerializedChunkSize(d.second); }); while (max_bytes > data_chunk_header_size_) { RTC_DCHECK(IsDivisibleBy4(max_bytes)); absl::optional chunk_opt = send_queue_.Produce(now, max_bytes - data_chunk_header_size_); if (!chunk_opt.has_value()) { break; } size_t chunk_size = GetSerializedChunkSize(chunk_opt->data); max_bytes -= chunk_size; rwnd_ -= chunk_size; absl::optional tsn = outstanding_data_.Insert( chunk_opt->data, partial_reliability_ ? chunk_opt->max_retransmissions : MaxRetransmits::NoLimit(), now, partial_reliability_ ? chunk_opt->expires_at : TimeMs::InfiniteFuture()); if (tsn.has_value()) { to_be_sent.emplace_back(tsn->Wrap(), std::move(chunk_opt->data)); } } } if (!to_be_sent.empty()) { // https://tools.ietf.org/html/rfc4960#section-6.3.2 // "Every time a DATA chunk is sent to any address (including a // retransmission), if the T3-rtx timer of that address is not running, // start it running so that it will expire after the RTO of that address." if (!t3_rtx_.is_running()) { t3_rtx_.Start(); } RTC_DLOG(LS_VERBOSE) << log_prefix_ << "Sending TSN " << StrJoin(to_be_sent, ",", [&](rtc::StringBuilder& sb, const std::pair& c) { sb << *c.first; }) << " - " << absl::c_accumulate( to_be_sent, 0, [&](size_t r, const std::pair& d) { return r + GetSerializedChunkSize(d.second); }) << " bytes. outstanding_bytes=" << outstanding_bytes() << " (" << old_outstanding_bytes << "), cwnd=" << cwnd_ << ", rwnd=" << rwnd_ << " (" << old_rwnd << ")"; } RTC_DCHECK(IsConsistent()); return to_be_sent; } bool RetransmissionQueue::can_send_data() const { return cwnd_ < options_.avoid_fragmentation_cwnd_mtus * options_.mtu || max_bytes_to_send() >= min_bytes_required_to_send_; } bool RetransmissionQueue::ShouldSendForwardTsn(TimeMs now) { if (!partial_reliability_) { return false; } outstanding_data_.ExpireOutstandingChunks(now); bool ret = outstanding_data_.ShouldSendForwardTsn(); RTC_DCHECK(IsConsistent()); return ret; } size_t RetransmissionQueue::max_bytes_to_send() const { size_t left = outstanding_bytes() >= cwnd_ ? 0 : cwnd_ - outstanding_bytes(); if (outstanding_bytes() == 0) { // https://datatracker.ietf.org/doc/html/rfc4960#section-6.1 // ... However, regardless of the value of rwnd (including if it is 0), the // data sender can always have one DATA chunk in flight to the receiver if // allowed by cwnd (see rule B, below). return left; } return std::min(rwnd(), left); } void RetransmissionQueue::PrepareResetStreams( rtc::ArrayView streams) { // TODO(boivie): These calls are now only affecting the send queue. The // packet buffer can also change behavior - for example draining the chunk // producer and eagerly assign TSNs so that an "Outgoing SSN Reset Request" // can be sent quickly, with a known `sender_last_assigned_tsn`. send_queue_.PrepareResetStreams(streams); } bool RetransmissionQueue::CanResetStreams() const { return send_queue_.CanResetStreams(); } void RetransmissionQueue::CommitResetStreams() { send_queue_.CommitResetStreams(); } void RetransmissionQueue::RollbackResetStreams() { send_queue_.RollbackResetStreams(); } HandoverReadinessStatus RetransmissionQueue::GetHandoverReadiness() const { HandoverReadinessStatus status; if (!outstanding_data_.empty()) { status.Add(HandoverUnreadinessReason::kRetransmissionQueueOutstandingData); } if (fast_recovery_exit_tsn_.has_value()) { status.Add(HandoverUnreadinessReason::kRetransmissionQueueFastRecovery); } if (outstanding_data_.has_data_to_be_retransmitted()) { status.Add(HandoverUnreadinessReason::kRetransmissionQueueNotEmpty); } return status; } void RetransmissionQueue::AddHandoverState(DcSctpSocketHandoverState& state) { state.tx.next_tsn = next_tsn().value(); state.tx.rwnd = rwnd_; state.tx.cwnd = cwnd_; state.tx.ssthresh = ssthresh_; state.tx.partial_bytes_acked = partial_bytes_acked_; } } // namespace dcsctp