From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=170.10.129.124 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1792527207; b=zbvQf0DHxktONi2PKVcl+hg72z7j6EvtDxeaJ9HCT7/bHT79VtZW7mHX3dmdNJgxJ/Yx WDuMz51RGNizMliXgoC1hwJrszbrZsiHB+4b8jjMG9dgXoQPN3nMvQjk9jTkvSA6GTyi9 6LIqVSTKScYqWkM3daAKmDbdAdo/o8gla5tQne9PVupkwnYwyYZytmysa6lqBOMwTTA/s m1R9N3T+4vOhC2muYmZwKEBaqk066VC/eFDuYjMTy8rYVdbgIRNptIOg8XdXhuEPNzAzg yeoSwDhxiItvzS0J7Ctpkslh15szNKbfJPVYLB14ME4SQtIu8QeOyLnFjGPXbpQD3kw== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1792527207; h=DMARC-Filter:DKIM-Signature:Received:X-MC-Unique: X-Mimecast-MFC-AGG-ID:Received:Received:From:To:Cc:Subject:Date: Message-ID:In-Reply-To:References:MIME-Version: Content-Transfer-Encoding:X-Scanned-By; bh=iF/2g8ezmRgNpkHpQC85116K3p+NIblnBxlCe9zwHwA=; b=0cIZcGIxAD7XA2xM6UzTg4mxtLFj75kG/z9ZPJz3LRFb3hmVpkG41yu8NGGl7I94QusE AR2JgQ8SNqT3Q+sZhcsc/RpcaVy6iwalaqouTXOjd3cmGy/8AfQ6CTJUPsX9Q5nL/Ajue zjltfP6xSDPdNFLTfwVrN8XT0t8F2oUmunzyM8V1oaWJkRn5Cq4CmCuPnQXGpSl6bp+hp Q28cQUiURvwwSL8anvivisVAk/WtovvxLWNC4JwPkvMeJejHYlbTFyL3USplodGvgB0hG Yy+AjjBus0sw5RNS95dSLoYILW6pNgLdf44St1DlgHfzgIoQk9vgZJOUUz1ojUl1YCg== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b="I/WwyRq2"; arc=none smtp.remote-ip=170.10.129.124 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (prime256v1) server-digest SHA256) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 68301606A321 for ; Tue, 20 Oct 2026 20:13:27 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.2 smtp.subspace.kernel.org 68301606A321 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key, unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=I/WwyRq2 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1792527206; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=iF/2g8ezmRgNpkHpQC85116K3p+NIblnBxlCe9zwHwA=; b=I/WwyRq2g4nSm9gRQwMQPZjyaiphmTiZcIoB2fnj4o+7DYdrQYkZDZJPd8LlDA7IWUYlMi YfAmigF65nWaPaggpJrTE8ItQvUIL7n+RmwHOZFHsmcYijr5EM7iNzE0c/s4bnxlPmk8fw RkxUqbVHDpPUCmLaa0O7Pt7V6DXY8oI= Received: from mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-433-AG6VDaP-MuaY6qqQwcB3AA-1; Tue, 20 Oct 2026 20:13:21 +0000 X-MC-Unique: AG6VDaP-MuaY6qqQwcB3AA-1 X-Mimecast-MFC-AGG-ID: AG6VDaP-MuaY6qqQwcB3AA_1792527199 Received: from mx-prod-int-03.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-03.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.12]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id CB91C1955F6E; Tue, 20 Oct 2026 20:13:18 +0000 (UTC) Received: from warthog.com (unknown [10.44.48.29]) by mx-prod-int-03.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 008D7195608B; Tue, 20 Oct 2026 20:13:14 +0000 (UTC) From: David Howells To: netdev@vger.kernel.org Cc: David Howells , Marc Dionne , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org, Jeffrey Altman , stable@vger.kernel.org Subject: [PATCH net v2 2/3] rxrpc: Fix counting of in-flight transmitted packets Date: Tue, 20 Oct 2026 21:12:53 +0100 Message-ID: <20261020201258.290827-3-dhowells@redhat.com> In-Reply-To: <20261020201258.290827-1-dhowells@redhat.com> References: <20261020201258.290827-1-dhowells@redhat.com> X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Scanned-By: MIMEDefang 3.0 on 10.30.177.12 The RACK-TLP algorithm [RFC8985] makes use of the count of in-flight packets in decisions (denoted as FlightSize). This is currently based on an approximation derived from the sum of other values: tx_nr_sent - (acks_nr_sacks + tx_nr_lost) + tx_nr_resent This, however, turns out not to be very good here as the component counters aren't updated together and this may lead to an incorrect estimate. Fix counting of in-flight transmitted packets by tracking them directly rather than deriving the value: (1) On each rxrpc_txqueue struct, keep a bitmap of which packets (txbuf) from the set managed by that txqueue are considered in-flight. (2) A new counter, tx_in_flight, is then added to the rxrpc_call struct to keep track of the number of packets considered to be in flight. (3) Set a packet's bit when the packet is transmitted and if the bit wasn't already set, increment tx_in_flight. (4) Clear a packet's bit when an ACK is received for the packet or if the packet is declared lost - and if the bit wasn't already cleared, decrement tx_in_flight. The value in tx_in_flight is then used instead of the approximation. The tx_nr_sent and tx_nr_resent counters are then removed and the rxrpc_rack tracepoint updated. Fixes: 7c482665931b ("rxrpc: Implement RACK/TLP to deal with transmission stalls [RFC8985]") Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260925191520.2206700-1-dhowells%40redhat.com Signed-off-by: David Howells cc: Marc Dionne cc: Jeffrey Altman cc: Eric Dumazet cc: "David S. Miller" cc: Jakub Kicinski cc: Paolo Abeni cc: Simon Horman cc: linux-afs@lists.infradead.org cc: stable@vger.kernel.org --- include/trace/events/rxrpc.h | 11 ++++------- net/rxrpc/ar-internal.h | 18 ++---------------- net/rxrpc/input.c | 8 ++++---- net/rxrpc/input_rack.c | 19 +++++++++++-------- net/rxrpc/output.c | 5 +++-- 5 files changed, 24 insertions(+), 37 deletions(-) diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 56dc9b614071..67a5612cdc55 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -2584,9 +2584,8 @@ TRACE_EVENT(rxrpc_rack, __field(rxrpc_serial_t, ack_serial) __field(rxrpc_seq_t, seq) __field(enum rxrpc_rack_timer_mode, mode) - __field(unsigned short, nr_sent) + __field(unsigned short, in_flight) __field(unsigned short, nr_lost) - __field(unsigned short, nr_resent) __field(unsigned short, nr_sacked) __field(ktime_t, timo) ), @@ -2596,18 +2595,16 @@ TRACE_EVENT(rxrpc_rack, __entry->ack_serial = call->rx_serial; __entry->seq = call->rack_end_seq; __entry->mode = call->rack_timer_mode; - __entry->nr_sent = call->tx_nr_sent; + __entry->in_flight = call->tx_in_flight; __entry->nr_lost = call->tx_nr_lost; - __entry->nr_resent = call->tx_nr_resent; __entry->nr_sacked = call->acks_nr_sacks; __entry->timo = timo; ), - TP_printk("c=%08x r=%08x q=%08x %s slrs=%u,%u,%u,%u t=%lld", + TP_printk("c=%08x r=%08x q=%08x %s fls=%u,%u,%u t=%lld", __entry->call, __entry->ack_serial, __entry->seq, __print_symbolic(__entry->mode, rxrpc_rack_timer_modes), - __entry->nr_sent, __entry->nr_lost, - __entry->nr_resent, __entry->nr_sacked, + __entry->in_flight, __entry->nr_lost, __entry->nr_sacked, ktime_to_us(__entry->timo)) ); diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index f0c8cbcc5dd5..d0e18eb6301e 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -782,9 +782,8 @@ struct rxrpc_call { rxrpc_seq_t tx_top; /* Highest Tx slot allocated. */ rxrpc_serial_t tx_last_serial; /* Serial of last DATA transmitted */ u16 tx_backoff; /* Delay to insert due to Tx failure (ms) */ - u16 tx_nr_sent; /* Number of packets sent, but unacked */ + u16 tx_in_flight; /* Number of packets considered "in flight" */ u16 tx_nr_lost; /* Number of packets marked lost */ - u16 tx_nr_resent; /* Number of packets resent, but unacked */ u16 tx_winsize; /* Maximum size of Tx window */ #define RXRPC_TX_MAX_WINDOW 128 u8 tx_jumbo_max; /* Maximum subpkts peer will accept */ @@ -980,6 +979,7 @@ struct rxrpc_txqueue { unsigned long segment_acked; /* Bit-per-buf: Set if ACK'd */ unsigned long segment_lost; /* Bit-per-buf: Set if declared lost */ unsigned long segment_retransmitted; /* Bit-per-buf: Set if retransmitted */ + unsigned long in_flight; /* Bit-per-buf: Set if "in flight" */ unsigned long rtt_samples; /* Bit-per-buf: Set if available for RTT */ unsigned long ever_retransmitted; /* Bit-per-buf: Set if ever retransmitted */ @@ -1601,20 +1601,6 @@ static inline unsigned int rxrpc_tx_window_space(const struct rxrpc_call *call) return max(winsize - transmitted, 0); } -static inline unsigned int rxrpc_left_out(const struct rxrpc_call *call) -{ - return call->acks_nr_sacks + call->tx_nr_lost; -} - -/* - * Calculate the number of transmitted DATA packets assumed to be in flight - * [approx RFC6675]. - */ -static inline unsigned int rxrpc_tx_in_flight(const struct rxrpc_call *call) -{ - return call->tx_nr_sent - rxrpc_left_out(call) + call->tx_nr_resent; -} - /* * debug tracing */ diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 73cafe6bfa9f..f497eba59cf6 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -33,7 +33,7 @@ static void rxrpc_congestion_management(struct rxrpc_call *call, struct rxrpc_ack_summary *summary) { summary->change = rxrpc_cong_no_change; - summary->in_flight = rxrpc_tx_in_flight(call); + summary->in_flight = call->tx_in_flight; if (test_and_clear_bit(RXRPC_CALL_RETRANS_TIMEOUT, &call->flags)) { summary->retrans_timeo = true; @@ -288,11 +288,11 @@ static bool rxrpc_rotate_tx_window(struct rxrpc_call *call, rxrpc_seq_t to, trace_rxrpc_rotate(call, tq, summary, seq, rxrpc_rotate_trace_snak); } - call->tx_nr_sent--; + if (__test_and_clear_bit(ix, &tq->in_flight)) + call->tx_in_flight--; if (__test_and_clear_bit(ix, &tq->segment_lost)) call->tx_nr_lost--; - if (__test_and_clear_bit(ix, &tq->segment_retransmitted)) - call->tx_nr_resent--; + __clear_bit(ix, &tq->segment_retransmitted); __clear_bit(ix, &tq->ever_retransmitted); rxrpc_put_txbuf(tq->bufs[ix], rxrpc_txbuf_put_rotated); diff --git a/net/rxrpc/input_rack.c b/net/rxrpc/input_rack.c index 072e1d9ecb2e..e43ce032a261 100644 --- a/net/rxrpc/input_rack.c +++ b/net/rxrpc/input_rack.c @@ -23,12 +23,13 @@ static bool rxrpc_rack_sent_after(ktime_t t1, rxrpc_seq_t seq1, static void rxrpc_rack_mark_lost(struct rxrpc_call *call, struct rxrpc_txqueue *tq, unsigned int ix) { - if (__test_and_set_bit(ix, &tq->segment_lost)) { - if (__test_and_clear_bit(ix, &tq->segment_retransmitted)) - call->tx_nr_resent--; - } else { + if (__test_and_set_bit(ix, &tq->segment_lost)) + __clear_bit(ix, &tq->segment_retransmitted); + else call->tx_nr_lost++; - } + + if (__test_and_clear_bit(ix, &tq->in_flight)) + call->tx_in_flight--; tq->segment_xmit_ts[ix] = UINT_MAX; } @@ -136,6 +137,8 @@ void rxrpc_input_rack(struct rxrpc_call *call, unsigned int ix = __ffs(new_acks); __clear_bit(ix, &new_acks); + if (__test_and_clear_bit(ix, &tq->in_flight)) + call->tx_in_flight--; rxrpc_input_rack_one(call, summary, tq, ix); } @@ -289,7 +292,7 @@ static void rxrpc_rack_mark_losses_on_rto(struct rxrpc_call *call) */ ktime_t rxrpc_tlp_calc_pto(struct rxrpc_call *call, ktime_t now) { - unsigned int flight_size = rxrpc_tx_in_flight(call); + unsigned int flight_size = call->tx_in_flight; ktime_t rto_at = ktime_add(call->tx_last_sent, rxrpc_get_rto_backoff(call, false)); ktime_t pto; @@ -313,7 +316,7 @@ ktime_t rxrpc_tlp_calc_pto(struct rxrpc_call *call, ktime_t now) */ void rxrpc_tlp_send_probe(struct rxrpc_call *call) { - unsigned int in_flight = rxrpc_tx_in_flight(call); + unsigned int in_flight = call->tx_in_flight; if (after_eq(call->acks_hard_ack, call->tx_transmitted)) return; /* Everything we transmitted has been acked. */ @@ -333,7 +336,7 @@ void rxrpc_tlp_send_probe(struct rxrpc_call *call) call->tlp_serial = call->tx_last_serial; call->tlp_seq = call->tx_transmitted; trace_rxrpc_tlp_probe(call, rxrpc_tlp_probe_trace_transmit_new); - in_flight = rxrpc_tx_in_flight(call); + in_flight = call->tx_in_flight; } else { /* Retransmit the highest-sequence DATA sent */ call->tx_last_serial = 0; diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c index 88cad087f13b..3ccd9a27643a 100644 --- a/net/rxrpc/output.c +++ b/net/rxrpc/output.c @@ -571,15 +571,16 @@ static size_t rxrpc_prepare_data_packet(struct rxrpc_call *call, _debug("prep[%u] tq=%x q=%x", i, tq->qbase, seq); + if (!__test_and_set_bit(ix, &tq->in_flight)) + call->tx_in_flight++; + /* Record (re-)transmission for RACK [RFC8985 6.1]. */ if (__test_and_clear_bit(ix, &tq->segment_lost)) call->tx_nr_lost--; if (req->retrans) { __set_bit(ix, &tq->ever_retransmitted); __set_bit(ix, &tq->segment_retransmitted); - call->tx_nr_resent++; } else { - call->tx_nr_sent++; start_tlp = true; } tq->segment_xmit_ts[ix] = xmit_ts;