[PATCH v9 net-next 00/15] AccECN protocol case handling series

[PATCH v9 net-next 00/15] AccECN protocol case handling series

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Hello,

Plesae find the v9 AccECN case handling patch series, which covers
several excpetional case handling of Accurate ECN spec (RFC9768),
adds new identifiers to be used by CC modules, adds ecn_delta into
rate_sample, and keeps the ACE counter for computation, etc.

This patch series is part of the full AccECN patch series, which is available at
https://github.com/L4STeam/linux-net-next/commits/upstream_l4steam/

Best regards,
Chia-Yu

---
v9:
- Add 1 patch for 2-bit tcpi_ecn_mode and 24-bit tcpi_options by reducing bits used by tcpi_accecn_fail_mode and tcpi_accecn_opt_seen in tcp_info (Neal Cardwell <ncardwell@google.com>)
- Add missing comma in patch #3 (Jakub Kicinski <kuba@kernel.org>)
- Update patch message of patch #15

v8:
- Add apcketdrill patch #14 into this series (Paolo Abeni <pabeni@redhat.com> & Jakub Kicinski <kuba@kernel.org>)
- Add helper function in patch #10 (Paolo Abeni <pabeni@redhat.com>)

v7:
- Update comments in #3 (Paolo Abeni <pabeni@redhat.com>)
- Update comments and use synack_type TCP_SYNACK_RETRANS and num_timeout in #9. (Paolo Abeni <pabeni@redhat.com>)

v6:
- Update comment in #3 to highlight RX path is only used for virtio-net (Paolo Abeni <pabeni@redhat.com>)
- Rename TCP_CONG_WANTS_ECT_1 to TCP_CONG_ECT_1_NEGOTIATION to distiguish from TCP_CONG_ECT_1_ESTABLISH (Paolo Abeni <pabeni@redhat.com>)
- Move TCP_CONG_ECT_1_ESTABLISH in #6 to latter patch series (Paolo Abeni <pabeni@redhat.com>)
- Add new synack_type instead of moving the increment of num_retran in #9 (Paolo Abeni <pabeni@redhat.com>)
- Use new synack_type TCP_SYNACK_RETRANS and num_retrans for SYN/ACK retx fallbackk for AccECN in #10 (Paolo Abeni <pabeni@redhat.com>)
- Do not cast const struct into non-const in #11, and set AccECN fail mode after tcp_rtx_synack() (Paolo Abeni <pabeni@redhat.com>)

v5:
- Move previous #11 in v4 in latter patch after discussion with RFC author.
- Add #3 to update the comments for SKB_GSO_TCP_ECN and SKB_GSO_TCP_ACCECN. (Parav Pandit <parav@nvidia.com>)
- Add gro self-test for TCP CWR flag in #4. (Eric Dumazet <edumazet@google.com>)
- Add fixes: tag into #7 (Paolo Abeni <pabeni@redhat.com>)
- Update commit message of #8 and if condition check (Paolo Abeni <pabeni@redhat.com>)
- Add empty line between variable declarations and code in #13 (Paolo Abeni <pabeni@redhat.com>)

v4:
- Add previous #13 in v2 back after dicussion with the RFC author.
- Add TCP_ACCECN_OPTION_PERSIST to tcp_ecn_option sysctl to ignore AccECN fallback policy on sending AccECN option.

v3:
- Add additional min() check if pkts_acked_ewma is not initialized in #1. (Paolo Abeni <pabeni@redhat.com>)
- Change TCP_CONG_WANTS_ECT_1 into individual flag add helper function INET_ECN_xmit_wants_ect_1() in #3. (Paolo Abeni <pabeni@redhat.com>)
- Add empty line between variable declarations and code in #4. (Paolo Abeni <pabeni@redhat.com>)
- Update commit message to fix old AccECN commits in #5. (Paolo Abeni <pabeni@redhat.com>)
- Remove unnecessary brackets in #10. (Paolo Abeni <pabeni@redhat.com>)
- Move patch #3 in v2 to a later Prague patch serise and remove patch #13 in v2. (Paolo Abeni <pabeni@redhat.com>)

---

Chia-Yu Chang (13):
  selftests/net: gro: add self-test for TCP CWR flag
  tcp: ECT_1_NEGOTIATION and NEEDS_ACCECN identifiers
  tcp: disable RFC3168 fallback identifier for CC modules
  tcp: accecn: handle unexpected AccECN negotiation feedback
  tcp: accecn: retransmit downgraded SYN in AccECN negotiation
  tcp: add TCP_SYNACK_RETRANS synack_type
  tcp: accecn: retransmit SYN/ACK without AccECN option or non-AccECN
    SYN/ACK
  tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion
  tcp: accecn: fallback outgoing half link to non-AccECN
  tcp: accecn: detect loss ACK w/ AccECN option and add
    TCP_ACCECN_OPTION_PERSIST
  tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info
  tcp: accecn: enable AccECN
  selftests/net: packetdrill: add TCP Accurate ECN cases

Ilpo Järvinen (2):
  tcp: try to avoid safer when ACKs are thinned
  gro: flushing when CWR is set negatively affects AccECN

 Documentation/networking/ip-sysctl.rst        |   4 +-
 .../networking/net_cachelines/tcp_sock.rst    |   1 +
 include/linux/tcp.h                           |   4 +-
 include/net/inet_ecn.h                        |  20 +++-
 include/net/tcp.h                             |  32 +++++-
 include/net/tcp_ecn.h                         | 108 ++++++++++++------
 include/uapi/linux/tcp.h                      |  26 ++++-
 net/ipv4/inet_connection_sock.c               |   3 +
 net/ipv4/sysctl_net_ipv4.c                    |   4 +-
 net/ipv4/tcp.c                                |  10 ++
 net/ipv4/tcp_cong.c                           |   5 +-
 net/ipv4/tcp_input.c                          |  37 +++++-
 net/ipv4/tcp_minisocks.c                      |  43 ++++---
 net/ipv4/tcp_offload.c                        |   3 +-
 net/ipv4/tcp_output.c                         |  32 ++++--
 net/ipv4/tcp_timer.c                          |   2 +
 tools/testing/selftests/drivers/net/gro.c     |  81 +++++++++----
 tools/testing/selftests/drivers/net/gro.py    |   3 +-
 .../tcp_accecn_2nd_data_as_first.pkt          |  24 ++++
 .../tcp_accecn_2nd_data_as_first_connect.pkt  |  30 +++++
 .../tcp_accecn_3rd_ack_after_synack_rxmt.pkt  |  19 +++
 ..._accecn_3rd_ack_ce_updates_received_ce.pkt |  18 +++
 .../tcp_accecn_3rd_ack_lost_data_ce.pkt       |  22 ++++
 .../net/packetdrill/tcp_accecn_3rd_dups.pkt   |  26 +++++
 .../tcp_accecn_acc_ecn_disabled.pkt           |  14 +++
 .../tcp_accecn_accecn_then_notecn_syn.pkt     |  28 +++++
 .../tcp_accecn_accecn_to_rfc3168.pkt          |  18 +++
 .../tcp_accecn_client_accecn_options_drop.pkt |  34 ++++++
 .../tcp_accecn_client_accecn_options_lost.pkt |  38 ++++++
 .../tcp_accecn_clientside_disabled.pkt        |  12 ++
 ...cecn_close_local_close_then_remote_fin.pkt |  25 ++++
 .../tcp_accecn_delivered_2ndlargeack.pkt      |  25 ++++
 ..._accecn_delivered_falseoverflow_detect.pkt |  31 +++++
 .../tcp_accecn_delivered_largeack.pkt         |  24 ++++
 .../tcp_accecn_delivered_largeack2.pkt        |  25 ++++
 .../tcp_accecn_delivered_maxack.pkt           |  25 ++++
 .../tcp_accecn_delivered_updates.pkt          |  70 ++++++++++++
 .../net/packetdrill/tcp_accecn_ecn3.pkt       |  12 ++
 .../tcp_accecn_ecn_field_updates_opt.pkt      |  35 ++++++
 .../packetdrill/tcp_accecn_ipflags_drop.pkt   |  14 +++
 .../tcp_accecn_listen_opt_drop.pkt            |  16 +++
 .../tcp_accecn_multiple_syn_ack_drop.pkt      |  28 +++++
 .../tcp_accecn_multiple_syn_drop.pkt          |  18 +++
 .../tcp_accecn_negotiation_bleach.pkt         |  23 ++++
 .../tcp_accecn_negotiation_connect.pkt        |  23 ++++
 .../tcp_accecn_negotiation_listen.pkt         |  26 +++++
 .../tcp_accecn_negotiation_noopt_connect.pkt  |  23 ++++
 .../tcp_accecn_negotiation_optenable.pkt      |  23 ++++
 .../tcp_accecn_no_ecn_after_accecn.pkt        |  20 ++++
 .../net/packetdrill/tcp_accecn_noopt.pkt      |  27 +++++
 .../net/packetdrill/tcp_accecn_noprogress.pkt |  27 +++++
 .../tcp_accecn_notecn_then_accecn_syn.pkt     |  28 +++++
 .../tcp_accecn_rfc3168_to_fallback.pkt        |  18 +++
 .../tcp_accecn_rfc3168_to_rfc3168.pkt         |  18 +++
 .../tcp_accecn_sack_space_grab.pkt            |  28 +++++
 .../tcp_accecn_sack_space_grab_with_ts.pkt    |  39 +++++++
 ...tcp_accecn_serverside_accecn_disabled1.pkt |  20 ++++
 ...tcp_accecn_serverside_accecn_disabled2.pkt |  20 ++++
 .../tcp_accecn_serverside_broken.pkt          |  19 +++
 .../tcp_accecn_serverside_ecn_disabled.pkt    |  19 +++
 .../tcp_accecn_serverside_only.pkt            |  18 +++
 ...n_syn_ace_flags_acked_after_retransmit.pkt |  18 +++
 .../tcp_accecn_syn_ace_flags_drop.pkt         |  16 +++
 ...n_ack_ace_flags_acked_after_retransmit.pkt |  27 +++++
 .../tcp_accecn_syn_ack_ace_flags_drop.pkt     |  27 +++++
 .../net/packetdrill/tcp_accecn_syn_ce.pkt     |  13 +++
 .../net/packetdrill/tcp_accecn_syn_ect0.pkt   |  13 +++
 .../net/packetdrill/tcp_accecn_syn_ect1.pkt   |  13 +++
 .../net/packetdrill/tcp_accecn_synack_ce.pkt  |  28 +++++
 ..._accecn_synack_ce_updates_delivered_ce.pkt |  22 ++++
 .../packetdrill/tcp_accecn_synack_ect0.pkt    |  24 ++++
 .../packetdrill/tcp_accecn_synack_ect1.pkt    |  24 ++++
 .../packetdrill/tcp_accecn_synack_rexmit.pkt  |  15 +++
 .../packetdrill/tcp_accecn_synack_rxmt.pkt    |  25 ++++
 .../packetdrill/tcp_accecn_tsnoprogress.pkt   |  26 +++++
 .../net/packetdrill/tcp_accecn_tsprogress.pkt |  25 ++++
 76 files changed, 1684 insertions(+), 100 deletions(-)
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_after_synack_rxmt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_ce_updates_received_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_lost_data_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_dups.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_acc_ecn_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_then_notecn_syn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_to_rfc3168.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_lost.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_clientside_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_close_local_close_then_remote_fin.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_2ndlargeack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_falseoverflow_detect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack2.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_maxack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_updates.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ecn3.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ecn_field_updates_opt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ipflags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_listen_opt_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_ack_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_bleach.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_listen.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_noopt_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_optenable.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_no_ecn_after_accecn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_noopt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_noprogress.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_notecn_then_accecn_syn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_fallback.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_rfc3168.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab_with_ts.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled2.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_broken.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_ecn_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_only.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_acked_after_retransmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_acked_after_retransmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect0.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce_updates_delivered_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect0.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rexmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rxmt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_tsnoprogress.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_tsprogress.pkt

-- 
2.34.1

[PATCH v9 net-next 01/15] tcp: try to avoid safer when ACKs are thinned

[chia-yu.chang]

From: Ilpo Järvinen <ij@kernel.org>

Add newly acked pkts EWMA. When ACK thinning occurs, select
between safer and unsafe cep delta in AccECN processing based
on it. If the packets ACKed per ACK tends to be large, don't
conservatively assume ACE field overflow.

This patch uses the existing 2-byte holes in the rx group for new
u16 variables withtout creating more holes. Below are the pahole
outcomes before and after this patch:

[BEFORE THIS PATCH]
struct tcp_sock {
    [...]
    u32                        delivered_ecn_bytes[3]; /*  2744    12 */
    /* XXX 4 bytes hole, try to pack */

    [...]
    __cacheline_group_end__tcp_sock_write_rx[0];       /*  2816     0 */

    [...]
    /* size: 3264, cachelines: 51, members: 177 */
}

[AFTER THIS PATCH]
struct tcp_sock {
    [...]
    u32                        delivered_ecn_bytes[3]; /*  2744    12 */
    u16                        pkts_acked_ewma;        /*  2756     2 */
    /* XXX 2 bytes hole, try to pack */

    [...]
    __cacheline_group_end__tcp_sock_write_rx[0];       /*  2816     0 */

    [...]
    /* size: 3264, cachelines: 51, members: 178 */
}

Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Co-developed-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v3:
- Add additional min() check if pkts_acked_ewma is not initialized.
---
 .../networking/net_cachelines/tcp_sock.rst    |  1 +
 include/linux/tcp.h                           |  1 +
 net/ipv4/tcp.c                                |  2 ++
 net/ipv4/tcp_input.c                          | 20 ++++++++++++++++++-
 4 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/Documentation/networking/net_cachelines/tcp_sock.rst b/Documentation/networking/net_cachelines/tcp_sock.rst
index 26f32dbcf6ec..563daea10d6c 100644
--- a/Documentation/networking/net_cachelines/tcp_sock.rst
+++ b/Documentation/networking/net_cachelines/tcp_sock.rst
@@ -105,6 +105,7 @@ u32                           received_ce             read_mostly         read_w
 u32[3]                        received_ecn_bytes      read_mostly         read_write
 u8:4                          received_ce_pending     read_mostly         read_write
 u32[3]                        delivered_ecn_bytes                         read_write
+u16                           pkts_acked_ewma                             read_write
 u8:2                          syn_ect_snt             write_mostly        read_write
 u8:2                          syn_ect_rcv             read_mostly         read_write
 u8:2                          accecn_minlen           write_mostly        read_write
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 20b8c6e21fef..683f38362977 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -345,6 +345,7 @@ struct tcp_sock {
 	u32	rate_interval_us;  /* saved rate sample: time elapsed */
 	u32	rcv_rtt_last_tsecr;
 	u32	delivered_ecn_bytes[3];
+	u16	pkts_acked_ewma;/* Pkts acked EWMA for AccECN cep heuristic */
 	u64	first_tx_mstamp;  /* start of window send phase */
 	u64	delivered_mstamp; /* time we reached "delivered" */
 	u64	bytes_acked;	/* RFC4898 tcpEStatsAppHCThruOctetsAcked
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index d5319ebe2452..37a6e0aa9176 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -3418,6 +3418,7 @@ int tcp_disconnect(struct sock *sk, int flags)
 	tcp_accecn_init_counters(tp);
 	tp->prev_ecnfield = 0;
 	tp->accecn_opt_tstamp = 0;
+	tp->pkts_acked_ewma = 0;
 	if (icsk->icsk_ca_initialized && icsk->icsk_ca_ops->release)
 		icsk->icsk_ca_ops->release(sk);
 	memset(icsk->icsk_ca_priv, 0, sizeof(icsk->icsk_ca_priv));
@@ -5191,6 +5192,7 @@ static void __init tcp_struct_check(void)
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rate_interval_us);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_rtt_last_tsecr);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, delivered_ecn_bytes);
+	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, pkts_acked_ewma);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, first_tx_mstamp);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, delivered_mstamp);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_acked);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 198f8a0d37be..8e95a4e302f4 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -488,6 +488,10 @@ static void tcp_count_delivered(struct tcp_sock *tp, u32 delivered,
 		tcp_count_delivered_ce(tp, delivered);
 }
 
+#define PKTS_ACKED_WEIGHT	6
+#define PKTS_ACKED_PREC		6
+#define ACK_COMP_THRESH		4
+
 /* Returns the ECN CE delta */
 static u32 __tcp_accecn_process(struct sock *sk, const struct sk_buff *skb,
 				u32 delivered_pkts, u32 delivered_bytes,
@@ -499,6 +503,7 @@ static u32 __tcp_accecn_process(struct sock *sk, const struct sk_buff *skb,
 	u32 delta, safe_delta, d_ceb;
 	bool opt_deltas_valid;
 	u32 corrected_ace;
+	u32 ewma;
 
 	/* Reordered ACK or uncertain due to lack of data to send and ts */
 	if (!(flag & (FLAG_FORWARD_PROGRESS | FLAG_TS_PROGRESS)))
@@ -507,6 +512,18 @@ static u32 __tcp_accecn_process(struct sock *sk, const struct sk_buff *skb,
 	opt_deltas_valid = tcp_accecn_process_option(tp, skb,
 						     delivered_bytes, flag);
 
+	if (delivered_pkts) {
+		if (!tp->pkts_acked_ewma) {
+			ewma = delivered_pkts << PKTS_ACKED_PREC;
+		} else {
+			ewma = tp->pkts_acked_ewma;
+			ewma = (((ewma << PKTS_ACKED_WEIGHT) - ewma) +
+				(delivered_pkts << PKTS_ACKED_PREC)) >>
+				PKTS_ACKED_WEIGHT;
+		}
+		tp->pkts_acked_ewma = min_t(u32, ewma, 0xFFFFU);
+	}
+
 	if (!(flag & FLAG_SLOWPATH)) {
 		/* AccECN counter might overflow on large ACKs */
 		if (delivered_pkts <= TCP_ACCECN_CEP_ACE_MASK)
@@ -555,7 +572,8 @@ static u32 __tcp_accecn_process(struct sock *sk, const struct sk_buff *skb,
 		if (d_ceb <
 		    safe_delta * tp->mss_cache >> TCP_ACCECN_SAFETY_SHIFT)
 			return delta;
-	}
+	} else if (tp->pkts_acked_ewma > (ACK_COMP_THRESH << PKTS_ACKED_PREC))
+		return delta;
 
 	return safe_delta;
 }
-- 
2.34.1

[PATCH v9 net-next 02/15] gro: flushing when CWR is set negatively affects AccECN

[chia-yu.chang]

From: Ilpo Järvinen <ij@kernel.org>

As AccECN may keep CWR bit asserted due to different
interpretation of the bit, flushing with GRO because of
CWR may effectively disable GRO until AccECN counter
field changes such that CWR-bit becomes 0.

There is no harm done from not immediately forwarding the
CWR'ed segment with RFC3168 ECN.

Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
---
 net/ipv4/tcp_offload.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c
index fdda18b1abda..9bd710c7bc95 100644
--- a/net/ipv4/tcp_offload.c
+++ b/net/ipv4/tcp_offload.c
@@ -303,8 +303,7 @@ struct sk_buff *tcp_gro_receive(struct list_head *head, struct sk_buff *skb,
 		goto out_check_final;
 
 	th2 = tcp_hdr(p);
-	flush = (__force int)(flags & TCP_FLAG_CWR);
-	flush |= (__force int)((flags ^ tcp_flag_word(th2)) &
+	flush = (__force int)((flags ^ tcp_flag_word(th2)) &
 		  ~(TCP_FLAG_FIN | TCP_FLAG_PSH));
 	flush |= (__force int)(th->ack_seq ^ th2->ack_seq);
 	for (i = sizeof(*th); i < thlen; i += 4)
-- 
2.34.1

[PATCH v9 net-next 03/15] selftests/net: gro: add self-test for TCP CWR flag

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Currently, GRO does not flush packets when the CWR bit is set.
A corresponding self-test is being added, in which the CWR flag
is set for two consecutive packets, but the first packet with the
CWR flag set will not be flushed immediately.

+===================+==========+===============+===========+
|     Packet id     | CWR flag |    Payload    | Flushing? |
+===================+==========+===============+===========+
|         0         |     0    |  PAYLOAD_LEN  |     0     |
|        ...        |     0    |  PAYLOAD_LEN  |     1     |
+-------------------+----------+---------------+-----------+
| NUM_PACKETS/2 - 1 |     1    |  payload_len  |     0     |
|   NUM_PACKETS/2   |     1    |  payload_len  |     1     |
+-------------------+----------+---------------+-----------+
|        ...        |     0    |  PAYLOAD_LEN  |     0     |
|   NUM_PACKETS     |     0    |  PAYLOAD_LEN  |     1     |
+===================+==========+===============+===========+

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v9:
- Add missing comma

v8:
- Rebase to the latest tools/testing/selftests/drivers/net/gro.c

v7:
- Update comments
---
 tools/testing/selftests/drivers/net/gro.c  | 81 ++++++++++++++++------
 tools/testing/selftests/drivers/net/gro.py |  3 +-
 2 files changed, 60 insertions(+), 24 deletions(-)

diff --git a/tools/testing/selftests/drivers/net/gro.c b/tools/testing/selftests/drivers/net/gro.c
index e76c618704cf..3c0745b68bfa 100644
--- a/tools/testing/selftests/drivers/net/gro.c
+++ b/tools/testing/selftests/drivers/net/gro.c
@@ -17,8 +17,8 @@
  *  Pure ACK does not coalesce.
  *
  * flags_*:
- *  No packets with PSH, SYN, URG, RST set will be coalesced.
- *   - flags_psh, flags_syn, flags_rst, flags_urg
+ *  No packets with PSH, SYN, URG, RST, CWR set will be coalesced.
+ *   - flags_psh, flags_syn, flags_rst, flags_urg, flags_cwr
  *
  * tcp_*:
  *  Packets with incorrect checksum, non-consecutive seqno and
@@ -360,32 +360,58 @@ static void create_packet(void *buf, int seq_offset, int ack_offset,
 	fill_datalinklayer(buf);
 }
 
-/* send one extra flag, not first and not last pkt */
-static void send_flags(int fd, struct sockaddr_ll *daddr, int psh, int syn,
-		       int rst, int urg)
+#ifndef TH_CWR
+#define TH_CWR 0x80
+#endif
+static void set_flags(struct tcphdr *tcph, int payload_len, int psh, int syn,
+		      int rst, int urg, int cwr)
 {
-	static char flag_buf[MAX_HDR_LEN + PAYLOAD_LEN];
-	static char buf[MAX_HDR_LEN + PAYLOAD_LEN];
-	int payload_len, pkt_size, flag, i;
-	struct tcphdr *tcph;
-
-	payload_len = PAYLOAD_LEN * psh;
-	pkt_size = total_hdr_len + payload_len;
-	flag = NUM_PACKETS / 2;
-
-	create_packet(flag_buf, flag * payload_len, 0, payload_len, 0);
-
-	tcph = (struct tcphdr *)(flag_buf + tcp_offset);
 	tcph->psh = psh;
 	tcph->syn = syn;
 	tcph->rst = rst;
 	tcph->urg = urg;
+	if (cwr)
+		tcph->th_flags |= TH_CWR;
+	else
+		tcph->th_flags &= ~TH_CWR;
 	tcph->check = 0;
 	tcph->check = tcp_checksum(tcph, payload_len);
+}
+
+/* send extra flags of the (NUM_PACKETS / 2) and (NUM_PACKETS / 2 - 1)
+ * pkts, not first and not last pkt
+ */
+static void send_flags(int fd, struct sockaddr_ll *daddr, int psh, int syn,
+		       int rst, int urg, int cwr)
+{
+	static char flag_buf[2][MAX_HDR_LEN + PAYLOAD_LEN];
+	static char buf[MAX_HDR_LEN + PAYLOAD_LEN];
+	int payload_len, pkt_size, i;
+	struct tcphdr *tcph;
+	int flag[2];
+
+	payload_len = PAYLOAD_LEN * (psh || cwr);
+	pkt_size = total_hdr_len + payload_len;
+	flag[0] = NUM_PACKETS / 2;
+	flag[1] = NUM_PACKETS / 2 - 1;
+
+	/* Create and configure packets with flags
+	 */
+	for (i = 0; i < 2; i++) {
+		if (flag[i] > 0) {
+			create_packet(flag_buf[i], flag[i] * payload_len, 0,
+				      payload_len, 0);
+			tcph = (struct tcphdr *)(flag_buf[i] + tcp_offset);
+			set_flags(tcph, payload_len, psh, syn, rst, urg, cwr);
+		}
+	}
 
 	for (i = 0; i < NUM_PACKETS + 1; i++) {
-		if (i == flag) {
-			write_packet(fd, flag_buf, pkt_size, daddr);
+		if (i == flag[0]) {
+			write_packet(fd, flag_buf[0], pkt_size, daddr);
+			continue;
+		} else if (i == flag[1] && cwr) {
+			write_packet(fd, flag_buf[1], pkt_size, daddr);
 			continue;
 		}
 		create_packet(buf, i * PAYLOAD_LEN, 0, PAYLOAD_LEN, 0);
@@ -1068,16 +1094,19 @@ static void gro_sender(void)
 
 	/* flags sub-tests */
 	} else if (strcmp(testname, "flags_psh") == 0) {
-		send_flags(txfd, &daddr, 1, 0, 0, 0);
+		send_flags(txfd, &daddr, 1, 0, 0, 0, 0);
 		write_packet(txfd, fin_pkt, total_hdr_len, &daddr);
 	} else if (strcmp(testname, "flags_syn") == 0) {
-		send_flags(txfd, &daddr, 0, 1, 0, 0);
+		send_flags(txfd, &daddr, 0, 1, 0, 0, 0);
 		write_packet(txfd, fin_pkt, total_hdr_len, &daddr);
 	} else if (strcmp(testname, "flags_rst") == 0) {
-		send_flags(txfd, &daddr, 0, 0, 1, 0);
+		send_flags(txfd, &daddr, 0, 0, 1, 0, 0);
 		write_packet(txfd, fin_pkt, total_hdr_len, &daddr);
 	} else if (strcmp(testname, "flags_urg") == 0) {
-		send_flags(txfd, &daddr, 0, 0, 0, 1);
+		send_flags(txfd, &daddr, 0, 0, 0, 1, 0);
+		write_packet(txfd, fin_pkt, total_hdr_len, &daddr);
+	} else if (strcmp(testname, "flags_cwr") == 0) {
+		send_flags(txfd, &daddr, 0, 0, 0, 0, 1);
 		write_packet(txfd, fin_pkt, total_hdr_len, &daddr);
 
 	/* tcp sub-tests */
@@ -1239,6 +1268,12 @@ static void gro_receiver(void)
 		correct_payload[2] = PAYLOAD_LEN * 2;
 		printf("urg flag ends coalescing: ");
 		check_recv_pkts(rxfd, correct_payload, 3);
+	} else if (strcmp(testname, "flags_cwr") == 0) {
+		correct_payload[0] = PAYLOAD_LEN;
+		correct_payload[1] = PAYLOAD_LEN * 2;
+		correct_payload[2] = PAYLOAD_LEN * 2;
+		printf("cwr flag ends coalescing: ");
+		check_recv_pkts(rxfd, correct_payload, 3);
 
 	/* tcp sub-tests */
 	} else if (strcmp(testname, "tcp_csum") == 0) {
diff --git a/tools/testing/selftests/drivers/net/gro.py b/tools/testing/selftests/drivers/net/gro.py
index 1bb8af571456..cbc1b19dbc91 100755
--- a/tools/testing/selftests/drivers/net/gro.py
+++ b/tools/testing/selftests/drivers/net/gro.py
@@ -17,6 +17,7 @@ Test cases:
   - flags_syn: Packets with SYN flag don't coalesce
   - flags_rst: Packets with RST flag don't coalesce
   - flags_urg: Packets with URG flag don't coalesce
+  - flags_cwr: Packets with CWR flag don't coalesce
   - tcp_csum: Packets with incorrect checksum don't coalesce
   - tcp_seq: Packets with non-consecutive seqno don't coalesce
   - tcp_ts: Packets with different timestamp options don't coalesce
@@ -191,7 +192,7 @@ def _gro_variants():
     common_tests = [
         "data_same", "data_lrg_sml", "data_sml_lrg",
         "ack",
-        "flags_psh", "flags_syn", "flags_rst", "flags_urg",
+        "flags_psh", "flags_syn", "flags_rst", "flags_urg", "flags_cwr",
         "tcp_csum", "tcp_seq", "tcp_ts", "tcp_opt",
         "ip_ecn", "ip_tos",
         "large_max", "large_rem",
-- 
2.34.1

[PATCH v9 net-next 04/15] tcp: ECT_1_NEGOTIATION and NEEDS_ACCECN identifiers

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Two CA module flags are added in this patch related to AccECN negotiation.
First, a new CA module flag (TCP_CONG_NEEDS_ACCECN) defines that the CA
expects to negotiate AccECN functionality using the ECE, CWR and AE flags
in the TCP header.

Second, during ECN negotiation, ECT(0) in the IP header is used. This patch
enables CA to control whether ECT(0) or ECT(1) should be used on a per-segment
basis. A new flag (TCP_CONG_ECT_1_NEGOTIATION) defines the expected ECT value
in the IP header by the CA when not-yet initialized for the connection.

The detailed AccECN negotiaotn during the 3WHS can be found in the AccECN spec:
  https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt

Co-developed-by: Olivier Tilmans <olivier.tilmans@nokia.com>
Signed-off-by: Olivier Tilmans <olivier.tilmans@nokia.com>
Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v6:
- Rename TCP_CONG_WANTS_ECT_1 to TCP_CONG_ECT_1_NEGOTIATION to distinguish
  it from TCP_CONG_ECT_1_ESTABLISH.
- Move TCP_CONG_ECT_1_ESTABLISH to latter TCP Prague patch series.

v3:
- Change TCP_CONG_WANTS_ECT_1 into individual flag.
- Add helper function INET_ECN_xmit_wants_ect_1().
---
 include/net/inet_ecn.h | 20 +++++++++++++++++---
 include/net/tcp.h      | 21 ++++++++++++++++++++-
 include/net/tcp_ecn.h  | 13 ++++++++++---
 net/ipv4/tcp_cong.c    |  5 +++--
 net/ipv4/tcp_input.c   |  3 ++-
 5 files changed, 52 insertions(+), 10 deletions(-)

diff --git a/include/net/inet_ecn.h b/include/net/inet_ecn.h
index ea32393464a2..827b87a95dab 100644
--- a/include/net/inet_ecn.h
+++ b/include/net/inet_ecn.h
@@ -51,11 +51,25 @@ static inline __u8 INET_ECN_encapsulate(__u8 outer, __u8 inner)
 	return outer;
 }
 
+/* Apply either ECT(0) or ECT(1) */
+static inline void __INET_ECN_xmit(struct sock *sk, bool use_ect_1)
+{
+	__u8 ect = use_ect_1 ? INET_ECN_ECT_1 : INET_ECN_ECT_0;
+
+	/* Mask the complete byte in case the connection alternates between
+	 * ECT(0) and ECT(1).
+	 */
+	inet_sk(sk)->tos &= ~INET_ECN_MASK;
+	inet_sk(sk)->tos |= ect;
+	if (inet6_sk(sk)) {
+		inet6_sk(sk)->tclass &= ~INET_ECN_MASK;
+		inet6_sk(sk)->tclass |= ect;
+	}
+}
+
 static inline void INET_ECN_xmit(struct sock *sk)
 {
-	inet_sk(sk)->tos |= INET_ECN_ECT_0;
-	if (inet6_sk(sk) != NULL)
-		inet6_sk(sk)->tclass |= INET_ECN_ECT_0;
+	__INET_ECN_xmit(sk, false);
 }
 
 static inline void INET_ECN_dontxmit(struct sock *sk)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 15f9b20f851f..41c781b6fff7 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1203,7 +1203,12 @@ enum tcp_ca_ack_event_flags {
 #define TCP_CONG_NON_RESTRICTED		BIT(0)
 /* Requires ECN/ECT set on all packets */
 #define TCP_CONG_NEEDS_ECN		BIT(1)
-#define TCP_CONG_MASK	(TCP_CONG_NON_RESTRICTED | TCP_CONG_NEEDS_ECN)
+/* Require successfully negotiated AccECN capability */
+#define TCP_CONG_NEEDS_ACCECN		BIT(2)
+/* Use ECT(1) instead of ECT(0) while the CA is uninitialized */
+#define TCP_CONG_ECT_1_NEGOTIATION	BIT(3)
+#define TCP_CONG_MASK  (TCP_CONG_NON_RESTRICTED | TCP_CONG_NEEDS_ECN | \
+			TCP_CONG_NEEDS_ACCECN | TCP_CONG_ECT_1_NEGOTIATION)
 
 union tcp_cc_info;
 
@@ -1344,6 +1349,20 @@ static inline bool tcp_ca_needs_ecn(const struct sock *sk)
 	return icsk->icsk_ca_ops->flags & TCP_CONG_NEEDS_ECN;
 }
 
+static inline bool tcp_ca_needs_accecn(const struct sock *sk)
+{
+	const struct inet_connection_sock *icsk = inet_csk(sk);
+
+	return icsk->icsk_ca_ops->flags & TCP_CONG_NEEDS_ACCECN;
+}
+
+static inline bool tcp_ca_ect_1_negotiation(const struct sock *sk)
+{
+	const struct inet_connection_sock *icsk = inet_csk(sk);
+
+	return icsk->icsk_ca_ops->flags & TCP_CONG_ECT_1_NEGOTIATION;
+}
+
 static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index f13e5cd2b1ac..fdde1c342b35 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -31,6 +31,12 @@ enum tcp_accecn_option {
 	TCP_ACCECN_OPTION_FULL = 2,
 };
 
+/* Apply either ECT(0) or ECT(1) based on TCP_CONG_ECT_1_NEGOTIATION flag */
+static inline void INET_ECN_xmit_ect_1_negotiation(struct sock *sk)
+{
+	__INET_ECN_xmit(sk, tcp_ca_ect_1_negotiation(sk));
+}
+
 static inline void tcp_ecn_queue_cwr(struct tcp_sock *tp)
 {
 	/* Do not set CWR if in AccECN mode! */
@@ -561,7 +567,7 @@ static inline void tcp_ecn_send_synack(struct sock *sk, struct sk_buff *skb)
 		TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_ECE;
 	else if (tcp_ca_needs_ecn(sk) ||
 		 tcp_bpf_ca_needs_ecn(sk))
-		INET_ECN_xmit(sk);
+		INET_ECN_xmit_ect_1_negotiation(sk);
 
 	if (tp->ecn_flags & TCP_ECN_MODE_ACCECN) {
 		TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_ACE;
@@ -579,7 +585,8 @@ static inline void tcp_ecn_send_syn(struct sock *sk, struct sk_buff *skb)
 	bool use_ecn, use_accecn;
 	u8 tcp_ecn = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn);
 
-	use_accecn = tcp_ecn == TCP_ECN_IN_ACCECN_OUT_ACCECN;
+	use_accecn = tcp_ecn == TCP_ECN_IN_ACCECN_OUT_ACCECN ||
+		     tcp_ca_needs_accecn(sk);
 	use_ecn = tcp_ecn == TCP_ECN_IN_ECN_OUT_ECN ||
 		  tcp_ecn == TCP_ECN_IN_ACCECN_OUT_ECN ||
 		  tcp_ca_needs_ecn(sk) || bpf_needs_ecn || use_accecn;
@@ -595,7 +602,7 @@ static inline void tcp_ecn_send_syn(struct sock *sk, struct sk_buff *skb)
 
 	if (use_ecn) {
 		if (tcp_ca_needs_ecn(sk) || bpf_needs_ecn)
-			INET_ECN_xmit(sk);
+			INET_ECN_xmit_ect_1_negotiation(sk);
 
 		TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_ECE | TCPHDR_CWR;
 		if (use_accecn) {
diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
index df758adbb445..e9f6c77e0631 100644
--- a/net/ipv4/tcp_cong.c
+++ b/net/ipv4/tcp_cong.c
@@ -16,6 +16,7 @@
 #include <linux/gfp.h>
 #include <linux/jhash.h>
 #include <net/tcp.h>
+#include <net/tcp_ecn.h>
 #include <trace/events/tcp.h>
 
 static DEFINE_SPINLOCK(tcp_cong_list_lock);
@@ -227,7 +228,7 @@ void tcp_assign_congestion_control(struct sock *sk)
 
 	memset(icsk->icsk_ca_priv, 0, sizeof(icsk->icsk_ca_priv));
 	if (ca->flags & TCP_CONG_NEEDS_ECN)
-		INET_ECN_xmit(sk);
+		INET_ECN_xmit_ect_1_negotiation(sk);
 	else
 		INET_ECN_dontxmit(sk);
 }
@@ -257,7 +258,7 @@ static void tcp_reinit_congestion_control(struct sock *sk,
 	memset(icsk->icsk_ca_priv, 0, sizeof(icsk->icsk_ca_priv));
 
 	if (ca->flags & TCP_CONG_NEEDS_ECN)
-		INET_ECN_xmit(sk);
+		INET_ECN_xmit_ect_1_negotiation(sk);
 	else
 		INET_ECN_dontxmit(sk);
 
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 8e95a4e302f4..ccbab5569680 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -7266,7 +7266,8 @@ static void tcp_ecn_create_request(struct request_sock *req,
 	u32 ecn_ok_dst;
 
 	if (tcp_accecn_syn_requested(th) &&
-	    READ_ONCE(net->ipv4.sysctl_tcp_ecn) >= 3) {
+	    (READ_ONCE(net->ipv4.sysctl_tcp_ecn) >= 3 ||
+	     tcp_ca_needs_accecn(listen_sk))) {
 		inet_rsk(req)->ecn_ok = 1;
 		tcp_rsk(req)->accecn_ok = 1;
 		tcp_rsk(req)->syn_ect_rcv = TCP_SKB_CB(skb)->ip_dsfield &
-- 
2.34.1

[PATCH v9 net-next 05/15] tcp: disable RFC3168 fallback identifier for CC modules

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

When AccECN is not successfully negociated for a TCP flow, it defaults
fallback to classic ECN (RFC3168). However, L4S service will fallback
to non-ECN.

This patch enables congestion control module to control whether it
should not fallback to classic ECN after unsuccessful AccECN negotiation.
A new CA module flag (TCP_CONG_NO_FALLBACK_RFC3168) identifies this
behavior expected by the CA.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v3:
- Add empty line between variable declarations and code.
---
 include/net/tcp.h        | 12 +++++++++++-
 include/net/tcp_ecn.h    | 11 ++++++++---
 net/ipv4/tcp_input.c     |  2 +-
 net/ipv4/tcp_minisocks.c |  7 ++++---
 4 files changed, 24 insertions(+), 8 deletions(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 41c781b6fff7..426571272688 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1207,8 +1207,11 @@ enum tcp_ca_ack_event_flags {
 #define TCP_CONG_NEEDS_ACCECN		BIT(2)
 /* Use ECT(1) instead of ECT(0) while the CA is uninitialized */
 #define TCP_CONG_ECT_1_NEGOTIATION	BIT(3)
+/* Cannot fallback to RFC3168 during AccECN negotiation */
+#define TCP_CONG_NO_FALLBACK_RFC3168	BIT(4)
 #define TCP_CONG_MASK  (TCP_CONG_NON_RESTRICTED | TCP_CONG_NEEDS_ECN | \
-			TCP_CONG_NEEDS_ACCECN | TCP_CONG_ECT_1_NEGOTIATION)
+			TCP_CONG_NEEDS_ACCECN | TCP_CONG_ECT_1_NEGOTIATION | \
+			TCP_CONG_NO_FALLBACK_RFC3168)
 
 union tcp_cc_info;
 
@@ -1363,6 +1366,13 @@ static inline bool tcp_ca_ect_1_negotiation(const struct sock *sk)
 	return icsk->icsk_ca_ops->flags & TCP_CONG_ECT_1_NEGOTIATION;
 }
 
+static inline bool tcp_ca_no_fallback_rfc3168(const struct sock *sk)
+{
+	const struct inet_connection_sock *icsk = inet_csk(sk);
+
+	return icsk->icsk_ca_ops->flags & TCP_CONG_NO_FALLBACK_RFC3168;
+}
+
 static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index fdde1c342b35..2e1637edf1d3 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -507,7 +507,9 @@ static inline void tcp_ecn_rcv_synack(struct sock *sk, const struct sk_buff *skb
 		 * | ECN    | AccECN | 0   0   1  | Classic ECN |
 		 * +========+========+============+=============+
 		 */
-		if (tcp_ecn_mode_pending(tp))
+		if (tcp_ca_no_fallback_rfc3168(sk))
+			tcp_ecn_mode_set(tp, TCP_ECN_DISABLED);
+		else if (tcp_ecn_mode_pending(tp))
 			/* Downgrade from AccECN, or requested initially */
 			tcp_ecn_mode_set(tp, TCP_ECN_MODE_RFC3168);
 		break;
@@ -531,9 +533,11 @@ static inline void tcp_ecn_rcv_synack(struct sock *sk, const struct sk_buff *skb
 	}
 }
 
-static inline void tcp_ecn_rcv_syn(struct tcp_sock *tp, const struct tcphdr *th,
+static inline void tcp_ecn_rcv_syn(struct sock *sk, const struct tcphdr *th,
 				   const struct sk_buff *skb)
 {
+	struct tcp_sock *tp = tcp_sk(sk);
+
 	if (tcp_ecn_mode_pending(tp)) {
 		if (!tcp_accecn_syn_requested(th)) {
 			/* Downgrade to classic ECN feedback */
@@ -545,7 +549,8 @@ static inline void tcp_ecn_rcv_syn(struct tcp_sock *tp, const struct tcphdr *th,
 			tcp_ecn_mode_set(tp, TCP_ECN_MODE_ACCECN);
 		}
 	}
-	if (tcp_ecn_mode_rfc3168(tp) && (!th->ece || !th->cwr))
+	if (tcp_ecn_mode_rfc3168(tp) &&
+	    (!th->ece || !th->cwr || tcp_ca_no_fallback_rfc3168(sk)))
 		tcp_ecn_mode_set(tp, TCP_ECN_DISABLED);
 }
 
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index ccbab5569680..e5c9cf586437 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -6861,7 +6861,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb,
 		tp->snd_wl1    = TCP_SKB_CB(skb)->seq;
 		tp->max_window = tp->snd_wnd;
 
-		tcp_ecn_rcv_syn(tp, th, skb);
+		tcp_ecn_rcv_syn(sk, th, skb);
 
 		tcp_mtup_init(sk);
 		tcp_sync_mss(sk, icsk->icsk_pmtu_cookie);
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index bd5462154f97..9776c921d1bb 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -485,9 +485,10 @@ static void tcp_ecn_openreq_child(struct sock *sk,
 		tp->accecn_opt_demand = 1;
 		tcp_ecn_received_counters_payload(sk, skb);
 	} else {
-		tcp_ecn_mode_set(tp, inet_rsk(req)->ecn_ok ?
-				     TCP_ECN_MODE_RFC3168 :
-				     TCP_ECN_DISABLED);
+		if (inet_rsk(req)->ecn_ok && !tcp_ca_no_fallback_rfc3168(sk))
+			tcp_ecn_mode_set(tp, TCP_ECN_MODE_RFC3168);
+		else
+			tcp_ecn_mode_set(tp, TCP_ECN_DISABLED);
 	}
 }
 
-- 
2.34.1

[PATCH v9 net-next 06/15] tcp: accecn: handle unexpected AccECN negotiation feedback

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

According to Section 3.1.2 of AccECN spec (RFC9768), if a TCP Client
has sent a SYN requesting AccECN feedback with (AE,CWR,ECE) = (1,1,1)
then receives a SYN/ACK with the currently reserved combination
(AE,CWR,ECE) = (1,0,1) but it does not have logic specific to such a
combination, the Client MUST enable AccECN mode as if the SYN/ACK
confirmed that the Server supported AccECN and as if it fed back that
the IP-ECN field on the SYN had arrived unchanged.

Fixes: 3cae34274c79 ("tcp: accecn: AccECN negotiation").
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v5:
- Add "Fixes" tag.

v3:
- Update commit message to fix old AccECN commits.
---
 include/net/tcp_ecn.h | 44 ++++++++++++++++++++++++++++++-------------
 1 file changed, 31 insertions(+), 13 deletions(-)

diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index 2e1637edf1d3..a709fb1756eb 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -473,6 +473,26 @@ static inline u8 tcp_accecn_option_init(const struct sk_buff *skb,
 	return TCP_ACCECN_OPT_COUNTER_SEEN;
 }
 
+static inline void tcp_ecn_rcv_synack_accecn(struct sock *sk,
+					     const struct sk_buff *skb, u8 dsf)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+
+	tcp_ecn_mode_set(tp, TCP_ECN_MODE_ACCECN);
+	tp->syn_ect_rcv = dsf & INET_ECN_MASK;
+	/* Demand Accurate ECN option in response to the SYN on the SYN/ACK
+	 * and the TCP server will try to send one more packet with an AccECN
+	 * Option at a later point during the connection.
+	 */
+	if (tp->rx_opt.accecn &&
+	    tp->saw_accecn_opt < TCP_ACCECN_OPT_COUNTER_SEEN) {
+		u8 saw_opt = tcp_accecn_option_init(skb, tp->rx_opt.accecn);
+
+		tcp_accecn_saw_opt_fail_recv(tp, saw_opt);
+		tp->accecn_opt_demand = 2;
+	}
+}
+
 /* See Table 2 of the AccECN draft */
 static inline void tcp_ecn_rcv_synack(struct sock *sk, const struct sk_buff *skb,
 				      const struct tcphdr *th, u8 ip_dsfield)
@@ -495,13 +515,11 @@ static inline void tcp_ecn_rcv_synack(struct sock *sk, const struct sk_buff *skb
 		tcp_ecn_mode_set(tp, TCP_ECN_DISABLED);
 		break;
 	case 0x1:
-	case 0x5:
 		/* +========+========+============+=============+
 		 * | A      | B      |  SYN/ACK   |  Feedback   |
 		 * |        |        |    B->A    |  Mode of A  |
 		 * |        |        | AE CWR ECE |             |
 		 * +========+========+============+=============+
-		 * | AccECN | Nonce  | 1   0   1  | (Reserved)  |
 		 * | AccECN | ECN    | 0   0   1  | Classic ECN |
 		 * | Nonce  | AccECN | 0   0   1  | Classic ECN |
 		 * | ECN    | AccECN | 0   0   1  | Classic ECN |
@@ -509,20 +527,20 @@ static inline void tcp_ecn_rcv_synack(struct sock *sk, const struct sk_buff *skb
 		 */
 		if (tcp_ca_no_fallback_rfc3168(sk))
 			tcp_ecn_mode_set(tp, TCP_ECN_DISABLED);
-		else if (tcp_ecn_mode_pending(tp))
-			/* Downgrade from AccECN, or requested initially */
+		else
 			tcp_ecn_mode_set(tp, TCP_ECN_MODE_RFC3168);
 		break;
-	default:
-		tcp_ecn_mode_set(tp, TCP_ECN_MODE_ACCECN);
-		tp->syn_ect_rcv = ip_dsfield & INET_ECN_MASK;
-		if (tp->rx_opt.accecn &&
-		    tp->saw_accecn_opt < TCP_ACCECN_OPT_COUNTER_SEEN) {
-			u8 saw_opt = tcp_accecn_option_init(skb, tp->rx_opt.accecn);
-
-			tcp_accecn_saw_opt_fail_recv(tp, saw_opt);
-			tp->accecn_opt_demand = 2;
+	case 0x5:
+		if (tcp_ecn_mode_pending(tp)) {
+			tcp_ecn_rcv_synack_accecn(sk, skb, ip_dsfield);
+			if (INET_ECN_is_ce(ip_dsfield)) {
+				tp->received_ce++;
+				tp->received_ce_pending++;
+			}
 		}
+		break;
+	default:
+		tcp_ecn_rcv_synack_accecn(sk, skb, ip_dsfield);
 		if (INET_ECN_is_ce(ip_dsfield) &&
 		    tcp_accecn_validate_syn_feedback(sk, ace,
 						     tp->syn_ect_snt)) {
-- 
2.34.1

[PATCH v9 net-next 07/15] tcp: accecn: retransmit downgraded SYN in AccECN negotiation

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Based on AccECN spec (RFC9768), if the sender of an AccECN SYN
(the TCP Client) times out before receiving the SYN/ACK, it SHOULD
attempt to negotiate the use of AccECN at least one more time by
continuing to set all three TCP ECN flags (AE,CWR,ECE) = (1,1,1) on
the first retransmitted SYN (using the usual retransmission time-outs).

If this first retransmission also fails to be acknowledged, in
deployment scenarios where AccECN path traversal might be problematic,
the TCP Client SHOULD send subsequent retransmissions of the SYN with
the three TCP-ECN flags cleared (AE,CWR,ECE) = (0,0,0).

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v5:
- Update commit message and the if condition statement.
---
 net/ipv4/tcp_output.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 256b669e8d3b..d5d695a501f8 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -3606,12 +3606,15 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs)
 			tcp_retrans_try_collapse(sk, skb, avail_wnd);
 	}
 
-	/* RFC3168, section 6.1.1.1. ECN fallback
-	 * As AccECN uses the same SYN flags (+ AE), this check covers both
-	 * cases.
-	 */
-	if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN_ECN) == TCPHDR_SYN_ECN)
-		tcp_ecn_clear_syn(sk, skb);
+	if (!tcp_ecn_mode_pending(tp) || icsk->icsk_retransmits > 1) {
+		/* RFC3168, section 6.1.1.1. ECN fallback
+		 * As AccECN uses the same SYN flags (+ AE), this check
+		 * covers both cases.
+		 */
+		if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN_ECN) ==
+		    TCPHDR_SYN_ECN)
+			tcp_ecn_clear_syn(sk, skb);
+	}
 
 	/* Update global and local TCP statistics. */
 	segs = tcp_skb_pcount(skb);
-- 
2.34.1

[PATCH v9 net-next 08/15] tcp: add TCP_SYNACK_RETRANS synack_type

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Before this patch, retransmitted SYN/ACK did not have a specific synack_type;
however, the upcoming patch needs to distinguish between retransmitted and
non-retransmitted SYN/ACK for AccECN negotiation to transmit the fallback
SYN/ACK during AccECN negotiation. Therefore, this patch introduces a new
synack_type (TCP_SYNACK_RETRANS).

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v6:
- Add new synack_type instead of moving the increment of num_retran.
---
 include/net/tcp.h     | 1 +
 net/ipv4/tcp_output.c | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 426571272688..96c3b27de0f5 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -541,6 +541,7 @@ enum tcp_synack_type {
 	TCP_SYNACK_NORMAL,
 	TCP_SYNACK_FASTOPEN,
 	TCP_SYNACK_COOKIE,
+	TCP_SYNACK_RETRANS,
 };
 struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,
 				struct request_sock *req,
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index d5d695a501f8..329e7e461c52 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -3956,6 +3956,7 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,
 
 	switch (synack_type) {
 	case TCP_SYNACK_NORMAL:
+	case TCP_SYNACK_RETRANS:
 		skb_set_owner_edemux(skb, req_to_sk(req));
 		break;
 	case TCP_SYNACK_COOKIE:
@@ -4641,7 +4642,7 @@ int tcp_rtx_synack(const struct sock *sk, struct request_sock *req)
 	/* Paired with WRITE_ONCE() in sock_setsockopt() */
 	if (READ_ONCE(sk->sk_txrehash) == SOCK_TXREHASH_ENABLED)
 		WRITE_ONCE(tcp_rsk(req)->txhash, net_tx_rndhash());
-	res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_NORMAL,
+	res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_RETRANS,
 				  NULL);
 	if (!res) {
 		TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS);
-- 
2.34.1

[PATCH v9 net-next 09/15] tcp: accecn: retransmit SYN/ACK without AccECN option or non-AccECN SYN/ACK

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

For Accurate ECN, the first SYN/ACK sent by the TCP server shall set the
ACE flag (see Table 1 of RFC9768) and the AccECN option to complete the
capability negotiation. However, if the TCP server needs to retransmit such
a SYN/ACK (for example, because it did not receive an ACK acknowledging its
SYN/ACK, or received a second SYN requesting AccECN support), the TCP server
retransmits the SYN/ACK without the AccECN option. This is because the
SYN/ACK may be lost due to congestion, or a middlebox may block the AccECN
option. Furthermore, if this retransmission also times out, to expedite
connection establishment, the TCP server should retransmit the SYN/ACK with
(AE,CWR,ECE) = (0,0,0) and without the AccECN option, while maintaining
AccECN feedback mode.

This complies with Section 3.2.3.2.2 of the AccECN specification (RFC9768).

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

---
v7:
- Update comments and use synack_type TCP_SYNACK_RETRANS and num_timeout.

v6:
- Use new synack_type TCP_SYNACK_RETRANS and num_retrans.
---
 include/net/tcp_ecn.h | 20 +++++++++++++++-----
 net/ipv4/tcp_output.c |  4 ++--
 2 files changed, 17 insertions(+), 7 deletions(-)

diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index a709fb1756eb..796c613b5ef3 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -649,12 +649,22 @@ static inline void tcp_ecn_clear_syn(struct sock *sk, struct sk_buff *skb)
 }
 
 static inline void
-tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th)
+tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th,
+		    enum tcp_synack_type synack_type)
 {
-	if (tcp_rsk(req)->accecn_ok)
-		tcp_accecn_echo_syn_ect(th, tcp_rsk(req)->syn_ect_rcv);
-	else if (inet_rsk(req)->ecn_ok)
-		th->ece = 1;
+	/* Accurate ECN shall retransmit SYN/ACK with ACE=0 if the
+	 * previously retransmitted SYN/ACK also times out.
+	 */
+	if (!req->num_timeout || synack_type != TCP_SYNACK_RETRANS) {
+		if (tcp_rsk(req)->accecn_ok)
+			tcp_accecn_echo_syn_ect(th, tcp_rsk(req)->syn_ect_rcv);
+		else if (inet_rsk(req)->ecn_ok)
+			th->ece = 1;
+	} else if (tcp_rsk(req)->accecn_ok) {
+		th->ae  = 0;
+		th->cwr = 0;
+		th->ece = 0;
+	}
 }
 
 static inline bool tcp_accecn_option_beacon_check(const struct sock *sk)
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 329e7e461c52..8536ad08a668 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1106,7 +1106,7 @@ static unsigned int tcp_synack_options(const struct sock *sk,
 
 	if (treq->accecn_ok &&
 	    READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn_option) &&
-	    req->num_timeout < 1 && remaining >= TCPOLEN_ACCECN_BASE) {
+	    synack_type != TCP_SYNACK_RETRANS && remaining >= TCPOLEN_ACCECN_BASE) {
 		opts->use_synack_ecn_bytes = 1;
 		remaining -= tcp_options_fit_accecn(opts, 0, remaining);
 	}
@@ -4039,7 +4039,7 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,
 	memset(th, 0, sizeof(struct tcphdr));
 	th->syn = 1;
 	th->ack = 1;
-	tcp_ecn_make_synack(req, th);
+	tcp_ecn_make_synack(req, th, synack_type);
 	th->source = htons(ireq->ir_num);
 	th->dest = ireq->ir_rmt_port;
 	skb->mark = ireq->ir_mark;
-- 
2.34.1

[PATCH v9 net-next 10/15] tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Based on specification:
  https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt

Based on Section 3.1.5 of AccECN spec (RFC9768), a TCP Server in
AccECN mode MUST NOT set ECT on any packet for the rest of the connection,
if it has received or sent at least one valid SYN or Acceptable SYN/ACK
with (AE,CWR,ECE) = (0,0,0) during the handshake.

In addition, a host in AccECN mode that is feeding back the IP-ECN
field on a SYN or SYN/ACK MUST feed back the IP-ECN field on the
latest valid SYN or acceptable SYN/ACK to arrive.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

---
v8:
- Add new helper function tcp_accecn_ace_fail_send_set_retrans()

v6:
- Do not cast const struct request_sock into struct request_sock
- Set tcp_accecn_fail_mode after calling tcp_rtx_synack().
---
 include/net/tcp_ecn.h           |  7 +++++++
 net/ipv4/inet_connection_sock.c |  3 +++
 net/ipv4/tcp_input.c            |  2 ++
 net/ipv4/tcp_minisocks.c        | 36 ++++++++++++++++++++++++---------
 net/ipv4/tcp_output.c           |  3 ++-
 net/ipv4/tcp_timer.c            |  2 ++
 6 files changed, 42 insertions(+), 11 deletions(-)

diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index 796c613b5ef3..f5e1f6b1bec3 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -97,6 +97,13 @@ static inline void tcp_accecn_fail_mode_set(struct tcp_sock *tp, u8 mode)
 	tp->accecn_fail_mode |= mode;
 }
 
+static inline void tcp_accecn_ace_fail_send_set_retrans(struct request_sock *req,
+							struct tcp_sock *tp)
+{
+	if (req->num_retrans > 1 && tcp_rsk(req)->accecn_ok)
+		tcp_accecn_fail_mode_set(tp, TCP_ACCECN_ACE_FAIL_SEND);
+}
+
 #define TCP_ACCECN_OPT_NOT_SEEN		0x0
 #define TCP_ACCECN_OPT_EMPTY_SEEN	0x1
 #define TCP_ACCECN_OPT_COUNTER_SEEN	0x2
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 97d57c52b9ad..9d16cb9c3db4 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -20,6 +20,7 @@
 #include <net/tcp_states.h>
 #include <net/xfrm.h>
 #include <net/tcp.h>
+#include <net/tcp_ecn.h>
 #include <net/sock_reuseport.h>
 #include <net/addrconf.h>
 
@@ -1103,6 +1104,8 @@ static void reqsk_timer_handler(struct timer_list *t)
 	    (!resend ||
 	     !tcp_rtx_synack(sk_listener, req) ||
 	     inet_rsk(req)->acked)) {
+		tcp_accecn_ace_fail_send_set_retrans(req,
+						     tcp_sk(sk_listener));
 		if (req->num_timeout++ == 0)
 			atomic_dec(&queue->young);
 		mod_timer(&req->rsk_timer, jiffies + tcp_reqsk_timeout(req));
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index e5c9cf586437..db361daebff8 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -6240,6 +6240,8 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb,
 	if (th->syn) {
 		if (tcp_ecn_mode_accecn(tp)) {
 			accecn_reflector = true;
+			tp->syn_ect_rcv = TCP_SKB_CB(skb)->ip_dsfield &
+					  INET_ECN_MASK;
 			if (tp->rx_opt.accecn &&
 			    tp->saw_accecn_opt < TCP_ACCECN_OPT_COUNTER_SEEN) {
 				u8 saw_opt = tcp_accecn_option_init(skb, tp->rx_opt.accecn);
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 9776c921d1bb..889c4307b35f 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -749,16 +749,32 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 		 */
 		if (!tcp_oow_rate_limited(sock_net(sk), skb,
 					  LINUX_MIB_TCPACKSKIPPEDSYNRECV,
-					  &tcp_rsk(req)->last_oow_ack_time) &&
-
-		    !tcp_rtx_synack(sk, req)) {
-			unsigned long expires = jiffies;
-
-			expires += tcp_reqsk_timeout(req);
-			if (!fastopen)
-				mod_timer_pending(&req->rsk_timer, expires);
-			else
-				req->rsk_timer.expires = expires;
+					  &tcp_rsk(req)->last_oow_ack_time)) {
+			if (tcp_rsk(req)->accecn_ok) {
+				u8 ect_rcv = TCP_SKB_CB(skb)->ip_dsfield &
+					     INET_ECN_MASK;
+
+				tcp_rsk(req)->syn_ect_rcv = ect_rcv;
+				if (tcp_accecn_ace(tcp_hdr(skb)) == 0x0) {
+					u8 fail_mode = TCP_ACCECN_ACE_FAIL_RECV;
+
+					tcp_accecn_fail_mode_set(tcp_sk(sk),
+								 fail_mode);
+				}
+			}
+			if (!tcp_rtx_synack(sk, req)) {
+				unsigned long expires = jiffies;
+
+				tcp_accecn_ace_fail_send_set_retrans(req,
+								     tcp_sk(sk));
+
+				expires += tcp_reqsk_timeout(req);
+				if (!fastopen)
+					mod_timer_pending(&req->rsk_timer,
+							  expires);
+				else
+					req->rsk_timer.expires = expires;
+			}
 		}
 		return NULL;
 	}
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 8536ad08a668..042e7e9b13cc 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -334,7 +334,8 @@ static void tcp_ecn_send(struct sock *sk, struct sk_buff *skb,
 		return;
 
 	if (tcp_ecn_mode_accecn(tp)) {
-		if (!tcp_accecn_ace_fail_recv(tp))
+		if (!tcp_accecn_ace_fail_recv(tp) &&
+		    !tcp_accecn_ace_fail_send(tp))
 			INET_ECN_xmit(sk);
 		tcp_accecn_set_ace(tp, skb, th);
 		skb_shinfo(skb)->gso_type |= SKB_GSO_TCP_ACCECN;
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 160080c9021d..a07ec1e883f1 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -22,6 +22,7 @@
 #include <linux/module.h>
 #include <linux/gfp.h>
 #include <net/tcp.h>
+#include <net/tcp_ecn.h>
 #include <net/rstreason.h>
 
 static u32 tcp_clamp_rto_to_user_timeout(const struct sock *sk)
@@ -479,6 +480,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req)
 	 * it's not good to give up too easily.
 	 */
 	tcp_rtx_synack(sk, req);
+	tcp_accecn_ace_fail_send_set_retrans(req, tcp_sk(sk));
 	req->num_timeout++;
 	tcp_update_rto_stats(sk);
 	if (!tp->retrans_stamp)
-- 
2.34.1

[PATCH v9 net-next 11/15] tcp: accecn: fallback outgoing half link to non-AccECN

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

According to Section 3.2.2.1 of AccECN spec (RFC9768), if the Server
is in AccECN mode and in SYN-RCVD state, and if it receives a value of
zero on a pure ACK with SYN=0 and no SACK blocks, for the rest of the
connection the Server MUST NOT set ECT on outgoing packets and MUST
NOT respond to AccECN feedback. Nonetheless, as a Data Receiver it
MUST NOT disable AccECN feedback.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v3:
- Remove unnecessary brackets.
---
 include/net/tcp_ecn.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index f5e1f6b1bec3..bf7d3f9f22c7 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -182,7 +182,9 @@ static inline void tcp_accecn_third_ack(struct sock *sk,
 	switch (ace) {
 	case 0x0:
 		/* Invalid value */
-		tcp_accecn_fail_mode_set(tp, TCP_ACCECN_ACE_FAIL_RECV);
+		if (!TCP_SKB_CB(skb)->sacked)
+			tcp_accecn_fail_mode_set(tp, TCP_ACCECN_ACE_FAIL_RECV |
+						     TCP_ACCECN_OPT_FAIL_RECV);
 		break;
 	case 0x7:
 	case 0x5:
-- 
2.34.1

[PATCH v9 net-next 12/15] tcp: accecn: detect loss ACK w/ AccECN option and add TCP_ACCECN_OPTION_PERSIST

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Detect spurious retransmission of a previously sent ACK carrying the
AccECN option after the second retransmission. Since this might be caused
by the middlebox dropping ACK with options it does not recognize, disable
the sending of the AccECN option in all subsequent ACKs. This patch
follows Section 3.2.3.2.2 of AccECN spec (RFC9768).

Also, a new AccECN option sending mode is added to tcp_ecn_option sysctl:
(TCP_ECN_OPTION_PERSIST), which ignores the AccECN fallback policy and
persistently sends AccECN option once it fits into TCP option space.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>

---
v5:
- Add empty line between variable declarations and code
---
 Documentation/networking/ip-sysctl.rst |  4 +++-
 include/linux/tcp.h                    |  3 ++-
 include/net/tcp_ecn.h                  |  2 ++
 net/ipv4/sysctl_net_ipv4.c             |  2 +-
 net/ipv4/tcp_input.c                   | 10 ++++++++++
 net/ipv4/tcp_output.c                  |  7 ++++++-
 6 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst
index bc9a01606daf..28c7e4f5ecf9 100644
--- a/Documentation/networking/ip-sysctl.rst
+++ b/Documentation/networking/ip-sysctl.rst
@@ -482,7 +482,9 @@ tcp_ecn_option - INTEGER
 	1 Send AccECN option sparingly according to the minimum option
 	  rules outlined in draft-ietf-tcpm-accurate-ecn.
 	2 Send AccECN option on every packet whenever it fits into TCP
-	  option space.
+	  option space except when AccECN fallback is triggered.
+	3 Send AccECN option on every packet whenever it fits into TCP
+	  option space even when AccECN fallback is triggered.
 	= ============================================================
 
 	Default: 2
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 683f38362977..32b031d09294 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -294,7 +294,8 @@ struct tcp_sock {
 	u8	nonagle     : 4,/* Disable Nagle algorithm?             */
 		rate_app_limited:1;  /* rate_{delivered,interval_us} limited? */
 	u8	received_ce_pending:4, /* Not yet transmit cnt of received_ce */
-		unused2:4;
+		accecn_opt_sent:1,/* Sent AccECN option in previous ACK */
+		unused2:3;
 	u8	accecn_minlen:2,/* Minimum length of AccECN option sent */
 		est_ecnfield:2,/* ECN field for AccECN delivered estimates */
 		accecn_opt_demand:2,/* Demand AccECN option for n next ACKs */
diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index bf7d3f9f22c7..41b593ece1dd 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -29,6 +29,7 @@ enum tcp_accecn_option {
 	TCP_ACCECN_OPTION_DISABLED = 0,
 	TCP_ACCECN_OPTION_MINIMUM = 1,
 	TCP_ACCECN_OPTION_FULL = 2,
+	TCP_ACCECN_OPTION_PERSIST = 3,
 };
 
 /* Apply either ECT(0) or ECT(1) based on TCP_CONG_ECT_1_NEGOTIATION flag */
@@ -413,6 +414,7 @@ static inline void tcp_accecn_init_counters(struct tcp_sock *tp)
 	tp->received_ce_pending = 0;
 	__tcp_accecn_init_bytes_counters(tp->received_ecn_bytes);
 	__tcp_accecn_init_bytes_counters(tp->delivered_ecn_bytes);
+	tp->accecn_opt_sent = 0;
 	tp->accecn_minlen = 0;
 	tp->accecn_opt_demand = 0;
 	tp->est_ecnfield = 0;
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index a1a50a5c80dc..385b5b986d23 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -749,7 +749,7 @@ static struct ctl_table ipv4_net_table[] = {
 		.mode		= 0644,
 		.proc_handler	= proc_dou8vec_minmax,
 		.extra1		= SYSCTL_ZERO,
-		.extra2		= SYSCTL_TWO,
+		.extra2		= SYSCTL_THREE,
 	},
 	{
 		.procname	= "tcp_ecn_option_beacon",
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index db361daebff8..2aae397b9d66 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4819,6 +4819,8 @@ static void tcp_dsack_extend(struct sock *sk, u32 seq, u32 end_seq)
 
 static void tcp_rcv_spurious_retrans(struct sock *sk, const struct sk_buff *skb)
 {
+	struct tcp_sock *tp = tcp_sk(sk);
+
 	/* When the ACK path fails or drops most ACKs, the sender would
 	 * timeout and spuriously retransmit the same segment repeatedly.
 	 * If it seems our ACKs are not reaching the other side,
@@ -4838,6 +4840,14 @@ static void tcp_rcv_spurious_retrans(struct sock *sk, const struct sk_buff *skb)
 	/* Save last flowlabel after a spurious retrans. */
 	tcp_save_lrcv_flowlabel(sk, skb);
 #endif
+	/* Check DSACK info to detect that the previous ACK carrying the
+	 * AccECN option was lost after the second retransmision, and then
+	 * stop sending AccECN option in all subsequent ACKs.
+	 */
+	if (tcp_ecn_mode_accecn(tp) &&
+	    TCP_SKB_CB(skb)->seq == tp->duplicate_sack[0].start_seq &&
+	    tp->accecn_opt_sent)
+		tcp_accecn_fail_mode_set(tp, TCP_ACCECN_OPT_FAIL_SEND);
 }
 
 static void tcp_send_dupack(struct sock *sk, const struct sk_buff *skb)
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 042e7e9b13cc..0cbba38ea87a 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -713,9 +713,12 @@ static void tcp_options_write(struct tcphdr *th, struct tcp_sock *tp,
 		if (tp) {
 			tp->accecn_minlen = 0;
 			tp->accecn_opt_tstamp = tp->tcp_mstamp;
+			tp->accecn_opt_sent = 1;
 			if (tp->accecn_opt_demand)
 				tp->accecn_opt_demand--;
 		}
+	} else if (tp) {
+		tp->accecn_opt_sent = 0;
 	}
 
 	if (unlikely(OPTION_SACK_ADVERTISE & options)) {
@@ -1187,7 +1190,9 @@ static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb
 	if (tcp_ecn_mode_accecn(tp)) {
 		int ecn_opt = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn_option);
 
-		if (ecn_opt && tp->saw_accecn_opt && !tcp_accecn_opt_fail_send(tp) &&
+		if (ecn_opt && tp->saw_accecn_opt &&
+		    (ecn_opt >= TCP_ACCECN_OPTION_PERSIST ||
+		     !tcp_accecn_opt_fail_send(tp)) &&
 		    (ecn_opt >= TCP_ACCECN_OPTION_FULL || tp->accecn_opt_demand ||
 		     tcp_accecn_option_beacon_check(sk))) {
 			opts->use_synack_ecn_bytes = 0;
-- 
2.34.1

[PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Add 2-bit tcpi_ecn_mode feild within tcp_info to indicate which ECN
mode is negotiated: ECN_MODE_DISABLED, ECN_MODE_RFC3168, ECN_MODE_ACCECN,
or ECN_MODE_PENDING. This is done by utilizing available bits from
tcpi_accecn_opt_seen (reduced from 16 bits to 2 bits) and
tcpi_accecn_fail_mode (reduced from 16 bits to 4 bits).

Also, an extra 24-bit tcpi_options2 field is identified to represent
newer options and connection features, as all 8 bits of tcpi_options
field have been used.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Co-developed-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
---
 include/net/tcp_ecn.h    | 11 -----------
 include/uapi/linux/tcp.h | 26 +++++++++++++++++++++++---
 net/ipv4/tcp.c           |  8 ++++++++
 3 files changed, 31 insertions(+), 14 deletions(-)

diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
index 41b593ece1dd..a31ba18b10d8 100644
--- a/include/net/tcp_ecn.h
+++ b/include/net/tcp_ecn.h
@@ -67,12 +67,6 @@ static inline void tcp_ecn_withdraw_cwr(struct tcp_sock *tp)
 	tp->ecn_flags &= ~TCP_ECN_QUEUE_CWR;
 }
 
-/* tp->accecn_fail_mode */
-#define TCP_ACCECN_ACE_FAIL_SEND	BIT(0)
-#define TCP_ACCECN_ACE_FAIL_RECV	BIT(1)
-#define TCP_ACCECN_OPT_FAIL_SEND	BIT(2)
-#define TCP_ACCECN_OPT_FAIL_RECV	BIT(3)
-
 static inline bool tcp_accecn_ace_fail_send(const struct tcp_sock *tp)
 {
 	return tp->accecn_fail_mode & TCP_ACCECN_ACE_FAIL_SEND;
@@ -105,11 +99,6 @@ static inline void tcp_accecn_ace_fail_send_set_retrans(struct request_sock *req
 		tcp_accecn_fail_mode_set(tp, TCP_ACCECN_ACE_FAIL_SEND);
 }
 
-#define TCP_ACCECN_OPT_NOT_SEEN		0x0
-#define TCP_ACCECN_OPT_EMPTY_SEEN	0x1
-#define TCP_ACCECN_OPT_COUNTER_SEEN	0x2
-#define TCP_ACCECN_OPT_FAIL_SEEN	0x3
-
 static inline u8 tcp_accecn_ace(const struct tcphdr *th)
 {
 	return (th->ae << 2) | (th->cwr << 1) | th->ece;
diff --git a/include/uapi/linux/tcp.h b/include/uapi/linux/tcp.h
index dce3113787a7..7be9044c5af3 100644
--- a/include/uapi/linux/tcp.h
+++ b/include/uapi/linux/tcp.h
@@ -226,6 +226,24 @@ enum tcp_ca_state {
 #define TCPF_CA_Loss	(1<<TCP_CA_Loss)
 };
 
+/* Values for tcpi_ecn_mode after negotiation */
+#define TCPI_ECN_MODE_DISABLED		0x0
+#define TCPI_ECN_MODE_RFC3168		0x1
+#define TCPI_ECN_MODE_ACCECN		0x2
+#define TCPI_ECN_MODE_PENDING		0x3
+
+/* Values for accecn_opt_seen */
+#define TCP_ACCECN_OPT_NOT_SEEN		0x0
+#define TCP_ACCECN_OPT_EMPTY_SEEN	0x1
+#define TCP_ACCECN_OPT_COUNTER_SEEN	0x2
+#define TCP_ACCECN_OPT_FAIL_SEEN	0x3
+
+/* Values for accecn_fail_mode */
+#define TCP_ACCECN_ACE_FAIL_SEND	BIT(0)
+#define TCP_ACCECN_ACE_FAIL_RECV	BIT(1)
+#define TCP_ACCECN_OPT_FAIL_SEND	BIT(2)
+#define TCP_ACCECN_OPT_FAIL_RECV	BIT(3)
+
 struct tcp_info {
 	__u8	tcpi_state;
 	__u8	tcpi_ca_state;
@@ -316,15 +334,17 @@ struct tcp_info {
 					 * in milliseconds, including any
 					 * unfinished recovery.
 					 */
-	__u32	tcpi_received_ce;    /* # of CE marks received */
+	__u32	tcpi_ecn_mode:2,
+		tcpi_accecn_opt_seen:2,
+		tcpi_accecn_fail_mode:4,
+		tcpi_options2:24;
+	__u32	tcpi_received_ce;    /* # of CE marked segments received */
 	__u32	tcpi_delivered_e1_bytes;  /* Accurate ECN byte counters */
 	__u32	tcpi_delivered_e0_bytes;
 	__u32	tcpi_delivered_ce_bytes;
 	__u32	tcpi_received_e1_bytes;
 	__u32	tcpi_received_e0_bytes;
 	__u32	tcpi_received_ce_bytes;
-	__u16	tcpi_accecn_fail_mode;
-	__u16	tcpi_accecn_opt_seen;
 };
 
 /* netlink attributes types for SCM_TIMESTAMPING_OPT_STATS */
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 37a6e0aa9176..f9e61e49f811 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -4321,6 +4321,14 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info)
 	if (tp->rto_stamp)
 		info->tcpi_total_rto_time += tcp_clock_ms() - tp->rto_stamp;
 
+	if (tcp_ecn_disabled(tp))
+		info->tcpi_ecn_mode = TCPI_ECN_MODE_DISABLED;
+	else if (tcp_ecn_mode_rfc3168(tp))
+		info->tcpi_ecn_mode = TCPI_ECN_MODE_RFC3168;
+	else if (tcp_ecn_mode_accecn(tp))
+		info->tcpi_ecn_mode = TCPI_ECN_MODE_ACCECN;
+	else if (tcp_ecn_mode_pending(tp))
+		info->tcpi_ecn_mode = TCPI_ECN_MODE_PENDING;
 	info->tcpi_accecn_fail_mode = tp->accecn_fail_mode;
 	info->tcpi_accecn_opt_seen = tp->saw_accecn_opt;
 	info->tcpi_received_ce = tp->received_ce;
-- 
2.34.1

[PATCH v9 net-next 14/15] tcp: accecn: enable AccECN

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Enable Accurate ECN negotiation and request for incoming and
outgoing connection by setting sysctl_tcp_ecn:

+==============+===========================================+
|              |  Highest ECN variant (Accurate ECN, ECN,  |
|   tcp_ecn    |  or no ECN) to be negotiated & requested  |
|              +---------------------+---------------------+
|              | Incoming connection | Outgoing connection |
+==============+=====================+=====================+
|      0       |        No ECN       |        No ECN       |
|      1       |         ECN         |         ECN         |
|      2       |         ECN         |        No ECN       |
+--------------+---------------------+---------------------+
|      3       |     Accurate ECN    |     Accurate ECN    |
|      4       |     Accurate ECN    |         ECN         |
|      5       |     Accurate ECN    |        No ECN       |
+==============+=====================+=====================+

Refer Documentation/networking/ip-sysctl.rst for more details.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
---
 net/ipv4/sysctl_net_ipv4.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 385b5b986d23..643763bc2142 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -47,7 +47,7 @@ static unsigned int udp_child_hash_entries_max = UDP_HTABLE_SIZE_MAX;
 static int tcp_plb_max_rounds = 31;
 static int tcp_plb_max_cong_thresh = 256;
 static unsigned int tcp_tw_reuse_delay_max = TCP_PAWS_MSL * MSEC_PER_SEC;
-static int tcp_ecn_mode_max = 2;
+static int tcp_ecn_mode_max = 5;
 static u32 icmp_errors_extension_mask_all =
 	GENMASK_U8(ICMP_ERR_EXT_COUNT - 1, 0);
 
-- 
2.34.1

[PATCH v9 net-next 15/15] selftests/net: packetdrill: add TCP Accurate ECN cases

[chia-yu.chang]

From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>

Linux Accurate ECN test sets using ACE counters and AccECN options to
cover several scenarios: Connection teardown, different ACK conditions,
counter wrapping, SACK space grabbing, fallback schemes, negotiation
retransmission/reorder/loss, AccECN option drop/loss, different
handshake reflectors, data with marking, and different sysctl values.

The packetdrill used is commit cbe405666c9c8698ac1e72f5e8ffc551216dfa56
of repo: https://github.com/minuscat/packetdrill/tree/upstream_accecn.
And corresponding patches are sent to google/packetdrill email list.

Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Co-developed-by: Ilpo Järvinen <ij@kernel.org>
Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Co-developed-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>

--
v9:
- Update commit message

v8:
- Change patch title
- Rename all AccECN cases with tcp_accecn in the prefix
- Move all cases under tools/testing/selftests/net/packetdrill/
---
 .../tcp_accecn_2nd_data_as_first.pkt          | 24 +++++++
 .../tcp_accecn_2nd_data_as_first_connect.pkt  | 30 ++++++++
 .../tcp_accecn_3rd_ack_after_synack_rxmt.pkt  | 19 +++++
 ..._accecn_3rd_ack_ce_updates_received_ce.pkt | 18 +++++
 .../tcp_accecn_3rd_ack_lost_data_ce.pkt       | 22 ++++++
 .../net/packetdrill/tcp_accecn_3rd_dups.pkt   | 26 +++++++
 .../tcp_accecn_acc_ecn_disabled.pkt           | 14 ++++
 .../tcp_accecn_accecn_then_notecn_syn.pkt     | 28 ++++++++
 .../tcp_accecn_accecn_to_rfc3168.pkt          | 18 +++++
 .../tcp_accecn_client_accecn_options_drop.pkt | 34 +++++++++
 .../tcp_accecn_client_accecn_options_lost.pkt | 38 ++++++++++
 .../tcp_accecn_clientside_disabled.pkt        | 12 ++++
 ...cecn_close_local_close_then_remote_fin.pkt | 25 +++++++
 .../tcp_accecn_delivered_2ndlargeack.pkt      | 25 +++++++
 ..._accecn_delivered_falseoverflow_detect.pkt | 31 ++++++++
 .../tcp_accecn_delivered_largeack.pkt         | 24 +++++++
 .../tcp_accecn_delivered_largeack2.pkt        | 25 +++++++
 .../tcp_accecn_delivered_maxack.pkt           | 25 +++++++
 .../tcp_accecn_delivered_updates.pkt          | 70 +++++++++++++++++++
 .../net/packetdrill/tcp_accecn_ecn3.pkt       | 12 ++++
 .../tcp_accecn_ecn_field_updates_opt.pkt      | 35 ++++++++++
 .../packetdrill/tcp_accecn_ipflags_drop.pkt   | 14 ++++
 .../tcp_accecn_listen_opt_drop.pkt            | 16 +++++
 .../tcp_accecn_multiple_syn_ack_drop.pkt      | 28 ++++++++
 .../tcp_accecn_multiple_syn_drop.pkt          | 18 +++++
 .../tcp_accecn_negotiation_bleach.pkt         | 23 ++++++
 .../tcp_accecn_negotiation_connect.pkt        | 23 ++++++
 .../tcp_accecn_negotiation_listen.pkt         | 26 +++++++
 .../tcp_accecn_negotiation_noopt_connect.pkt  | 23 ++++++
 .../tcp_accecn_negotiation_optenable.pkt      | 23 ++++++
 .../tcp_accecn_no_ecn_after_accecn.pkt        | 20 ++++++
 .../net/packetdrill/tcp_accecn_noopt.pkt      | 27 +++++++
 .../net/packetdrill/tcp_accecn_noprogress.pkt | 27 +++++++
 .../tcp_accecn_notecn_then_accecn_syn.pkt     | 28 ++++++++
 .../tcp_accecn_rfc3168_to_fallback.pkt        | 18 +++++
 .../tcp_accecn_rfc3168_to_rfc3168.pkt         | 18 +++++
 .../tcp_accecn_sack_space_grab.pkt            | 28 ++++++++
 .../tcp_accecn_sack_space_grab_with_ts.pkt    | 39 +++++++++++
 ...tcp_accecn_serverside_accecn_disabled1.pkt | 20 ++++++
 ...tcp_accecn_serverside_accecn_disabled2.pkt | 20 ++++++
 .../tcp_accecn_serverside_broken.pkt          | 19 +++++
 .../tcp_accecn_serverside_ecn_disabled.pkt    | 19 +++++
 .../tcp_accecn_serverside_only.pkt            | 18 +++++
 ...n_syn_ace_flags_acked_after_retransmit.pkt | 18 +++++
 .../tcp_accecn_syn_ace_flags_drop.pkt         | 16 +++++
 ...n_ack_ace_flags_acked_after_retransmit.pkt | 27 +++++++
 .../tcp_accecn_syn_ack_ace_flags_drop.pkt     | 27 +++++++
 .../net/packetdrill/tcp_accecn_syn_ce.pkt     | 13 ++++
 .../net/packetdrill/tcp_accecn_syn_ect0.pkt   | 13 ++++
 .../net/packetdrill/tcp_accecn_syn_ect1.pkt   | 13 ++++
 .../net/packetdrill/tcp_accecn_synack_ce.pkt  | 28 ++++++++
 ..._accecn_synack_ce_updates_delivered_ce.pkt | 22 ++++++
 .../packetdrill/tcp_accecn_synack_ect0.pkt    | 24 +++++++
 .../packetdrill/tcp_accecn_synack_ect1.pkt    | 24 +++++++
 .../packetdrill/tcp_accecn_synack_rexmit.pkt  | 15 ++++
 .../packetdrill/tcp_accecn_synack_rxmt.pkt    | 25 +++++++
 .../packetdrill/tcp_accecn_tsnoprogress.pkt   | 26 +++++++
 .../net/packetdrill/tcp_accecn_tsprogress.pkt | 25 +++++++
 58 files changed, 1366 insertions(+)
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_after_synack_rxmt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_ce_updates_received_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_lost_data_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_dups.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_acc_ecn_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_then_notecn_syn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_to_rfc3168.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_lost.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_clientside_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_close_local_close_then_remote_fin.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_2ndlargeack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_falseoverflow_detect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack2.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_maxack.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_updates.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ecn3.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ecn_field_updates_opt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_ipflags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_listen_opt_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_ack_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_bleach.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_listen.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_noopt_connect.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_optenable.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_no_ecn_after_accecn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_noopt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_noprogress.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_notecn_then_accecn_syn.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_fallback.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_rfc3168.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab_with_ts.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled2.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_broken.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_ecn_disabled.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_only.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_acked_after_retransmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_acked_after_retransmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_drop.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect0.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce_updates_delivered_ce.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect0.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect1.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rexmit.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rxmt.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_tsnoprogress.pkt
 create mode 100644 tools/testing/selftests/net/packetdrill/tcp_accecn_tsprogress.pkt

diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first.pkt
new file mode 100644
index 000000000000..07e9936e70e6
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first.pkt
@@ -0,0 +1,24 @@
+// 3rd ACK + 1st data segment lost, data segments with ce
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
++0.05 < SEWA 0:0(0) win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+// 3rd ACK lost
+// 1st data segment lost
++0.05 < [ce] EAP. 1001:2001(1000) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] WA. 1:1(0) ack 1 <ECN e1b 1 ceb 1000 e0b 1,nop,nop,nop,sack 1001:2001>
++.002 accept(3, ..., ...) = 4
+
++0.2 < [ce] EAP. 1:1001(1000) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.001 > [ect0] EWA. 1:1(0) ack 2001 <ECN e1b 1 ceb 2000 e0b 1,nop>
+
++0.05 < [ce] EAP. 2001:3001(1000) ack 1 win 264
++.001 > [ect0] . 1:1(0) ack 3001 <ECN e1b 1 ceb 3000 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first_connect.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first_connect.pkt
new file mode 100644
index 000000000000..76b8422b34dc
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_2nd_data_as_first_connect.pkt
@@ -0,0 +1,30 @@
+// 3rd ACK + 1st data segment lost, 2nd data segments with ce
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] SW. 0:0(0) ack 1 win 32767 <mss 1016,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
+// 3rd ACK lost
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 write(4, ..., 2000) = 2000
+// 1st data segment lost + 2nd gets CE
++.002 > [ect0] .5 1:1005(1004) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.000 > [ect0] P.5 1005:2001(996) ack 1 <ECN e1b 1 ceb 0 e0b 1, nop>
++0.05 < [ect0] .6 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 996 e1b 1,nop,nop,nop,sack 1005:2001>
+
++0.01 %{ assert tcpi_delivered_ce == 1, tcpi_delivered_ce }%
+
++0.002~+0.1 > [ect0] .5 1:1005(1004) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.05 < [ect0] .6 1:1(0) ack 2001 win 264 <ECN e0b 1005 ceb 996 e1b 1,nop>
+
++0.01 write(4, ..., 1000) = 1000
++0~+0.002 > [ect0] P.5 2001:3001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.1 < [ect0] .5 1:1001(1000) ack 3001 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++0~+0.01 > [ect0] .5 3001:3001(0) ack 1001 <ECN e1b 1 ceb 0 e0b 1001,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_after_synack_rxmt.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_after_synack_rxmt.pkt
new file mode 100644
index 000000000000..84060e490589
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_after_synack_rxmt.pkt
@@ -0,0 +1,19 @@
+// Test 3rd ACK flags when SYN-ACK is rexmitted
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.1 < [ect0] S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// Our code currently sends a challenge ACK
+// when it receives a SYN in ESTABLISHED state
+// based on the latest SYN
++.002 > [ect0] A. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_ce_updates_received_ce.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_ce_updates_received_ce.pkt
new file mode 100644
index 000000000000..d3fe09d0606f
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_ce_updates_received_ce.pkt
@@ -0,0 +1,18 @@
+// Third ACK CE increases r.cep
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
++0.05 < SEWA 0:0(0) win 32767 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ce] W. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] WAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_lost_data_ce.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_lost_data_ce.pkt
new file mode 100644
index 000000000000..d28722db42b1
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_ack_lost_data_ce.pkt
@@ -0,0 +1,22 @@
+// 3rd ACK lost, CE for the first data segment
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
++0.05 < SEWA 0:0(0) win 32767 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+// 3rd ACK lost
++0.05 < [ce] EAP. 1:1001(1000) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] WA. 1:1(0) ack 1001 <ECN e1b 1 ceb 1000 e0b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.05 < [ce] EAP. 1001:2001(1000) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.001 > [ect0] EWA. 1:1(0) ack 2001 <ECN e1b 1 ceb 2000 e0b 1 ,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_dups.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_dups.pkt
new file mode 100644
index 000000000000..a4d808116e34
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_3rd_dups.pkt
@@ -0,0 +1,26 @@
+// Test SYN/ACK rexmit triggered 3rd ACK duplicate + CE on first data seg
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// SYN/ACK rexmitted => two 3rd ACKs in-flight
++1.0~+1.1 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+// Delivered 1st 3rd ACK
++0.05 < [ect0] W. 1:1(0) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
+// Duplicate 3rd ACK delivered
++1.05 < [ect0] W. 1:1(0) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
+
++0.05 < [ce] EAP. 1:1001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] WA. 1:1(0) ack 1001 <ECN e1b 1 ceb 1000 e0b 1,nop>
+   +0 read(4, ..., 1000) = 1000
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_acc_ecn_disabled.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_acc_ecn_disabled.pkt
new file mode 100644
index 000000000000..509838d5a4b2
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_acc_ecn_disabled.pkt
@@ -0,0 +1,14 @@
+// Test that when accurate ECN is disabled,
+// client uses RFC3168 ECN for SYN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=1
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEW 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] S. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,nop,nop,nop,wscale 8>
++.002 > [noecn] . 1:1(0) ack 1
+
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_then_notecn_syn.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_then_notecn_syn.pkt
new file mode 100644
index 000000000000..10728114b11b
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_then_notecn_syn.pkt
@@ -0,0 +1,28 @@
+// Test that SYN-ACK with ACE flags and without
+// ACE flags got dropped. Although we disable ECN,
+// we shouldn't consider this as blackholed as
+// these are dropped due to congestion
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [ect0] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SA. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN
++0.1 < [noecn] S 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
++0.1 < [noecn] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
+// Write with AccECN option but with ip-noecn since we received one SYN with ACE=0
++0.01 write(4, ..., 100) = 100
++.002 > [noecn] P5. 1:101(100) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_to_rfc3168.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_to_rfc3168.pkt
new file mode 100644
index 000000000000..04d928f0d44d
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_accecn_to_rfc3168.pkt
@@ -0,0 +1,18 @@
+// Test AccECN -> RFC3168 fallback when sysctl asks for RFC3168 ECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=1
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SE. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < . 1:1(0) ack 1 win 320
++.002 accept(3, ..., ...) = 4
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] P. 1:1001(1000) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_drop.pkt
new file mode 100644
index 000000000000..788af6bea69c
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_drop.pkt
@@ -0,0 +1,34 @@
+// Client negotiates AccECN and starts sending
+// AccECN option in last ACK and data segments
+// Middlebox drops AccECN option and client
+// reverts to ACE flags only
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+sysctl -q net.ipv4.tcp_ecn_option_beacon=1
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.05 < [ect0] EAP. 1:1001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EA. 1:1(0) ack 1001 <ECN e1b 1 ceb 0 e0b 1001,nop>
+   +0 read(4, ..., 1000) = 1000
+
++0.05 < [ect0] EAP. 1:1001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EA. 1:1(0) ack 1001 <ECN e1b 1 ceb 0 e0b 2001,nop,nop,nop,sack 1:1001>
+
++0.05 < [ect0] EAP. 1:1001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EA. 1:1(0) ack 1001 <nop,nop,sack 1:1001>
+
++0.05 < [ect0] EAP. 1001:2001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EA. 1:1(0) ack 2001
+   +0 read(4, ..., 1000) = 1000
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_lost.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_lost.pkt
new file mode 100644
index 000000000000..d04e11bba37c
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_client_accecn_options_lost.pkt
@@ -0,0 +1,38 @@
+// Client negotiates AccECN and starts sending
+// AccECN option in last ACK and data segments
+// Middlebox accepts AccECN option but some packets
+// are lost due to congestion. Client should
+// continue to send AccECN option
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1024,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] A. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
+// Send
++0.01 write(4, ..., 3000) = 3000
++.002 > [ect0] .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.002 > [ect0] P.5 1013:2025(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.002 > [ect0] P.5 2025:3001(976) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
+// First two segments were lost due to congestion as SACK was
+// received acknowledging 3rd segment
++0.1 < [ect0] .5 1:1(0) ack 1 win 264 <ECN e1b 1 ceb 0 e0b 977,nop,nop,nop,sack 2025:3001>
+
+// Since data with option was SACKed, we can
+// continue to use AccECN option for the rest of
+// the connection. This one is a rexmt
++.02~+0.5 > [ect0] .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.1 < [ect0] .5 1:1(0) ack 3001 win 264 <ECN e1b 1 ceb 0 e0b 3000,nop>
+
+// Send new data, it should contain AccECN option
++0.01 write(4, ..., 2000) = 2000
++.002 > [ect0] .5 3001:4013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.002 > [ect0] P.5 4013:5001(988) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_clientside_disabled.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_clientside_disabled.pkt
new file mode 100644
index 000000000000..c00b36d6a833
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_clientside_disabled.pkt
@@ -0,0 +1,12 @@
+// AccECN sysctl server-side only, no ECN/AccECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=5
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > S 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < S. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,nop,nop,nop,wscale 8>
++.002 > . 1:1(0) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_close_local_close_then_remote_fin.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_close_local_close_then_remote_fin.pkt
new file mode 100644
index 000000000000..f9c27f39f354
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_close_local_close_then_remote_fin.pkt
@@ -0,0 +1,25 @@
+// Test basic connection teardown where local process closes first:
+// the local process calls close() first, so we send a FIN, and receive an ACK.
+// Then we receive a FIN and ACK it.
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+ +.01...0.011 connect(3, ..., ...) = 0
+   +0 > [noecn] SEWA 0:0(0) <...>
+   +0 < [ect1] SW. 0:0(0) ack 1 win 32768 <mss 1000,nop,wscale 6,nop,nop,sackOK>
+   +0 > [ect0] EW. 1:1(0) ack 1
+
+   +0 write(3, ..., 1000) = 1000
+   +0 > [ect0] P5. 1:1001(1000) ack 1
+   +0 < [ect0] .5 1:1(0) ack 1001 win 257
+
+   +0 close(3) = 0
+   +0 > [ect0] F5. 1001:1001(0) ack 1
+   +0 < [ect0] .5 1:1(0) ack 1002 win 257
+
+   +0 < [ect0] F5. 1:1(0) ack 1002 win 257
+   +0 > [ect0] . 1002:1002(0) ack 2
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_2ndlargeack.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_2ndlargeack.pkt
new file mode 100644
index 000000000000..6d771234124a
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_2ndlargeack.pkt
@@ -0,0 +1,25 @@
+// Test a large ACK (> ACE field max)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 14600) = 14600
++.002 > [ect0] P.5 1:14601(14600) ack 1
++0.05 < [ect0] .5 1:1(0) ack 1461 win 264
++0.05 < [ect0] .5 1:1(0) ack 14601 win 264
+
++0.01 %{ assert tcpi_delivered_ce == 8, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_falseoverflow_detect.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_falseoverflow_detect.pkt
new file mode 100644
index 000000000000..76384f52b021
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_falseoverflow_detect.pkt
@@ -0,0 +1,31 @@
+// Test false overflow detection with option used to rule out overflow
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
+// Stop sending option to allow easier testing
++0 `sysctl -q net.ipv4.tcp_ecn_option=0`
+
++0.002 write(4, ..., 14600) = 14600
++.002 > [ect0] P.5 1:14601(14600) ack 1
+
++0.05 < [ect0] .5 1:1(0) ack 1460 win 264 <ECN e0b 1461 ceb 0 e1b 1,nop>
++0.05 < [ect0] .5 1:1(0) ack 14601 win 264 <ECN e0b 14601 ceb 0 e1b 1,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 0, tcpi_delivered_ce
+assert tcpi_delivered_e0_bytes == 14600, tcpi_delivered_e0_bytes
+}%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack.pkt
new file mode 100644
index 000000000000..8bce5dce35a2
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack.pkt
@@ -0,0 +1,24 @@
+// Test a large ACK (> ACE field max)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 14600) = 14600
++.002 > [ect0] P.5 1:14601(14600) ack 1
++0.05 < [ect0] .5 1:1(0) ack 14601 win 264
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack2.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack2.pkt
new file mode 100644
index 000000000000..5f2b147214f4
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_largeack2.pkt
@@ -0,0 +1,25 @@
+// Test a large ACK (> ACE field max)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 14600) = 14600
++.002 > [ect0] P.5 1:14601(14600) ack 1
+  // Fake CE
++0.05 < [ect0] .6 1:1(0) ack 14601 win 264
+
++0.01 %{ assert tcpi_delivered_ce == 1, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_maxack.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_maxack.pkt
new file mode 100644
index 000000000000..fd07bdc14f37
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_maxack.pkt
@@ -0,0 +1,25 @@
+// Test a large ACK (at ACE field max delta)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 14600) = 14600
++.002 > [ect0] P.5 1:14601(14600) ack 1
+  // Fake CE
++0.05 < [ect0] .4 1:1(0) ack 14601 win 264
+
++0.01 %{ assert tcpi_delivered_ce == 7, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_updates.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_updates.pkt
new file mode 100644
index 000000000000..cb1e70ff2d26
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_delivered_updates.pkt
@@ -0,0 +1,70 @@
+// Test basic AccECN CEP/CEB/E0B/E1B functionality & CEP wrapping
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{
+assert tcpi_delivered_ce == 0, tcpi_delivered_ce
+assert tcpi_delivered_ce_bytes == 0, tcpi_delivered_ce_bytes
+}%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake CE
++0.05 < [ect0] WA. 1:1(0) ack 1001 win 264 <ECN e0b 1 ceb 1000 e1b 1,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 1, tcpi_delivered_ce
+assert tcpi_delivered_ce_bytes == 1000, tcpi_delivered_ce_bytes
+}%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake ect0
++0.05 < [ect0] WA. 1:1(0) ack 2001 win 264 <ECN e0b 1001 ceb 1000 e1b 1,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 1, tcpi_delivered_ce
+assert tcpi_delivered_e0_bytes == 1000, tcpi_delivered_e0_bytes
+}%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 2001:3001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake ce
++0.05 < [ect0] EWA. 1:1(0) ack 3001 win 264 <ECN e0b 1001 ceb 2000 e1b 1,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 2, tcpi_delivered_ce
+assert tcpi_delivered_ce_bytes == 2000, tcpi_delivered_ce_bytes
+}%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 3001:4001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake ect1
++0.05 < [ect0] EWA. 1:1(0) ack 4001 win 264 <ECN e0b 1001 ceb 2000 e1b 1001,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 2, tcpi_delivered_ce
+assert tcpi_delivered_e1_bytes == 1000, tcpi_delivered_e1_bytes
+}%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 4001:5001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake ce
++0.05 < [ect0] . 1:1(0) ack 5001 win 264 <ECN e0b 1001 ceb 3000 e1b 1001,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 3, tcpi_delivered_ce
+assert tcpi_delivered_ce_bytes == 3000, tcpi_delivered_ce_bytes
+}%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn3.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn3.pkt
new file mode 100644
index 000000000000..6627c7bb2d26
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn3.pkt
@@ -0,0 +1,12 @@
+// Test that tcp_ecn=4 uses RFC3168 ECN for SYN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=4
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.05 connect(4, ..., ...) = 0
+
++.002 > SEW 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > . 1:1(0) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn_field_updates_opt.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn_field_updates_opt.pkt
new file mode 100644
index 000000000000..51879477bb50
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_ecn_field_updates_opt.pkt
@@ -0,0 +1,35 @@
+// Test basic AccECN CEP/CEB/E0B/E1B functionality & CEP wrapping
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.05 < [ce] EAP. 1:1001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] WA. 1:1(0) ack 1001 <ECN e1b 1 ceb 1000 e0b 1,nop>
+   +0 read(4, ..., 1000) = 1000
+
++0.05 < [ect0] EAP. 1001:2001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] WA. 1:1(0) ack 2001 <ECN e1b 1 ceb 1000 e0b 1001,nop>
+   +0 read(4, ..., 1000) = 1000
+
++0.05 < [ce] EAP. 2001:3001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EWA. 1:1(0) ack 3001 <ECN e1b 1 ceb 2000 e0b 1001,nop>
+   +0 read(4, ..., 1000) = 1000
+
++0.05 < [ect1] EAP. 3001:4001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] EWA. 1:1(0) ack 4001 <ECN e1b 1001 ceb 2000 e0b 1001,nop>
+   +0 read(4, ..., 1000) = 1000
+
++0.05 < [ce] EAP. 4001:5001(1000) ack 1 win 257 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] . 1:1(0) ack 5001 <ECN e1b 1001 ceb 3000 e0b 1001,nop>
+   +0 read(4, ..., 1000) = 1000
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_ipflags_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_ipflags_drop.pkt
new file mode 100644
index 000000000000..0c72fa4a1251
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_ipflags_drop.pkt
@@ -0,0 +1,14 @@
+// Test IP flags drop
+--tolerance_usecs=50000
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 1.1 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++.02 ~ +1.1 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] . 1:1(0) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_listen_opt_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_listen_opt_drop.pkt
new file mode 100644
index 000000000000..171f9433e55f
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_listen_opt_drop.pkt
@@ -0,0 +1,16 @@
+// SYN/ACK option drop test
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.02 ~+2 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.02 ~+5 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.02 ~+8 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_ack_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_ack_drop.pkt
new file mode 100644
index 000000000000..0f65cf56cd2b
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_ack_drop.pkt
@@ -0,0 +1,28 @@
+// Test that SYN-ACK with ACE flags and without
+// ACE flags got dropped. Although we disable ECN,
+// we shouldn't consider this as blackholed as
+// these are dropped due to congestion
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [noecn] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN-ACK without option
++1~+1.1 > [noecn] SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// SYN-ACK maybe getting blackholed, disable ECN
++2~+2.2 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++4~+4.4 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// Received an ACK after sending 3rd retransmission, not a blackhole
++0.1 < [noecn] . 1:1(0) ack 1 win 320
++.002 accept(3, ..., ...) = 4
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_drop.pkt
new file mode 100644
index 000000000000..343181633980
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_multiple_syn_drop.pkt
@@ -0,0 +1,18 @@
+// Test that SYN with ACE flags and without
+// ACE flags got dropped. Although we disable
+// ECN, we shouldn't consider this as blackholed
+// as these are dropped due to congestion
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 3.1 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++.02~+1.1 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++.02~+1.1 > [noecn] S 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++.02~+1.1 > [noecn] S 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.1 < [noecn] S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0~+0.01 > [noecn] . 1:1(0) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_bleach.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_bleach.pkt
new file mode 100644
index 000000000000..37dabc4603c8
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_bleach.pkt
@@ -0,0 +1,23 @@
+// Test AccECN flags bleach
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] . 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [noecn] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.05 < [ect0] EAP. 1:1(0) ack 1001 win 320
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_connect.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_connect.pkt
new file mode 100644
index 000000000000..5b14892fda51
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_connect.pkt
@@ -0,0 +1,23 @@
+// Test basic AccECN negotiation
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++.05 < [ect0] EAP. 1:1(0) ack 1001 win 256 <ECN e0b 1001 ceb 0 e1b 0,nop>
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_listen.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_listen.pkt
new file mode 100644
index 000000000000..25f7cb2feb25
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_listen.pkt
@@ -0,0 +1,26 @@
+// Test basic AccECN negotiation
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.05 < [ect0] EAP. 1:1(0) ack 1001 win 320
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_noopt_connect.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_noopt_connect.pkt
new file mode 100644
index 000000000000..50e08c492a69
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_noopt_connect.pkt
@@ -0,0 +1,23 @@
+// Test basic AccECN negotiation without option
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < SW. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1
++.05 < [ect0] EAP. 1:1(0) ack 1001 win 256
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_optenable.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_optenable.pkt
new file mode 100644
index 000000000000..2904f1ba9975
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_negotiation_optenable.pkt
@@ -0,0 +1,23 @@
+// Test basic AccECN negotiation, late option enable
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < SW. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1
++.05 < [ect0] EAP. 1:1(0) ack 1001 win 256 <ECN e0b 1001 ceb 0 e1b 1,nop>
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_no_ecn_after_accecn.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_no_ecn_after_accecn.pkt
new file mode 100644
index 000000000000..64e0fc1c1f14
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_no_ecn_after_accecn.pkt
@@ -0,0 +1,20 @@
+// Test client behavior on receiving a non ECN SYN-ACK
+// after receiving an AccECN SYN-ACK and moving to
+// ESTABLISHED state
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
+// Receive an AccECN SYN-ACK and move to ESTABLISHED
++0.05 < [noecn] SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
+// Receive a non ECN SYN-ACK and send a challenge ACK with ACE feedback
++0.1 < [noecn] S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_noopt.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_noopt.pkt
new file mode 100644
index 000000000000..f407c629a3f7
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_noopt.pkt
@@ -0,0 +1,27 @@
+// Test basic AccECN negotiation with option off using sysctl
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1
++0.05 < [ect0] EAP. 1:1(0) ack 1001 win 320
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1001:2001(1000) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_noprogress.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_noprogress.pkt
new file mode 100644
index 000000000000..32454e7187f9
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_noprogress.pkt
@@ -0,0 +1,27 @@
+// Test no progress filtering
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+  // Fake CE and claim no progress
++0.05 < [ect0] WA. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 1000 e1b 1,nop>
+
++0.01 %{
+assert tcpi_delivered_ce == 0, tcpi_delivered_ce
+assert tcpi_delivered_ce_bytes == 0, tcpi_delivered_ce_bytes
+}%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_notecn_then_accecn_syn.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_notecn_then_accecn_syn.pkt
new file mode 100644
index 000000000000..6597d5f2d778
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_notecn_then_accecn_syn.pkt
@@ -0,0 +1,28 @@
+// Test that SYN-ACK with ACE flags and without
+// ACE flags got dropped. Although we disable ECN,
+// we shouldn't consider this as blackholed as
+// these are dropped due to congestion
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [noecn] S 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN
++0.1 < [ect0] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
++0.1 < [noecn] . 1:1(0) ack 1 win 320
++.002 accept(3, ..., ...) = 4
+
+// Write with AccECN option but with ip-noecn since we received one SYN with ACE=0
++0.01 write(4, ..., 100) = 100
++.002 > [noecn] P. 1:101(100) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_fallback.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_fallback.pkt
new file mode 100644
index 000000000000..0f97dfcfa82d
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_fallback.pkt
@@ -0,0 +1,18 @@
+// Test RFC3168 fallback when sysctl asks for AccECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEW 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SE. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < . 1:1(0) ack 1 win 320
++.002 accept(3, ..., ...) = 4
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] P. 1:1001(1000) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_rfc3168.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_rfc3168.pkt
new file mode 100644
index 000000000000..9baffdd66fe5
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_rfc3168_to_rfc3168.pkt
@@ -0,0 +1,18 @@
+// Test RFC3168 ECN when sysctl asks for RFC3168 ECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=1
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEW 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SE. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.05 < . 1:1(0) ack 1 win 320
++.002 accept(3, ..., ...) = 4
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] P. 1:1001(1000) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab.pkt
new file mode 100644
index 000000000000..3fc56f9c6a6f
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab.pkt
@@ -0,0 +1,28 @@
+// Test SACK space grab to fit AccECN option
+--tcp_ts_tick_usecs=1000
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++.01 < [ect1] EAP. 1001:2001(1000) ack 1 win 264
++0.002 > [ect0] EA. 1:1(0) ack 1 <ECN e1b 1001 ceb 0 e0b 1,nop,nop,nop,sack 1001:2001>
++.01 < [ect0] EAP. 3001:4001(1000) ack 1 win 264
++0.002 > [ect0] EA. 1:1(0) ack 1 <ECN e1b 1001 ceb 0 e0b 1001,nop,nop,nop,sack 3001:4001 1001:2001>
++.01 < [ce] EAP. 5001:6001(1000) ack 1 win 264
++0.002 > [ect0] WA. 1:1(0) ack 1 <ECN e1b 1001 ceb 1000 e0b 1001,nop,nop,nop,sack 5001:6001 3001:4001 1001:2001>
+// DSACK works?
++.01 < [ect0] EAP. 5001:6001(1000) ack 1 win 264
++0.002 > [ect0] WA. 1:1(0) ack 1 <ECN e1b 1001 ceb 1000 e0b 2001,nop,nop,nop,sack 5001:6001 5001:6001 3001:4001>
++.01 < [ect1] EAP. 6001:7001(1000) ack 1 win 264
++0.002 > [ect0] WA. 1:1(0) ack 1 <ECN e1b 2001 ceb 1000 e0b 2001,nop,nop,nop,sack 5001:7001 3001:4001 1001:2001>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab_with_ts.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab_with_ts.pkt
new file mode 100644
index 000000000000..1c075b5d81ae
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_sack_space_grab_with_ts.pkt
@@ -0,0 +1,39 @@
+// Test SACK space grab to fit AccECN option
+--tcp_ts_tick_usecs=1000
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,sackOK,TS val 1 ecr 0,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,sackOK,TS val 100 ecr 1,ECN e1b 1 ceb 0 e0b 1,nop,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <nop,nop,TS val 2 ecr 100,ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
+// One SACK block should allow all 3 AccECN fields:
++.01 < [ect1] EAP. 1001:2001(1000) ack 1 win 264 <nop,nop,TS val 3 ecr 100>
++0.002 > [ect0] EA. 1:1(0) ack 1 <nop,nop,TS val 160 ecr 2,ECN e1b 1001 ceb 0 e0b 1,nop,nop,nop,sack 1001:2001>
+
+// Two SACK blocks should fit w/ AccECN if we only need to use 2 AccECN fields: check ect1 arriving.
++.01 < [ect1] EAP. 3001:4001(1000) ack 1 win 264 <nop,nop,TS val 4 ecr 100>
++0.002 > [ect0] EA. 1:1(0) ack 1 <nop,nop,TS val 172 ecr 2,ECN e1b 2001 ceb 0,nop,nop,sack 3001:4001 1001:2001>
+
+// Two SACK blocks should fit w/ AccECN if we only need to use 2 AccECN fields: check CE arriving.
++.01 < [ce] EAP. 5001:6001(1000) ack 1 win 264 <nop,nop,TS val 5 ecr 100>
++0.002 > [ect0] WA. 1:1(0) ack 1 <nop,nop,TS val 184 ecr 2,ECN e1b 2001 ceb 1000,nop,nop,sack 5001:6001 3001:4001>
+
+// Check that DSACK works, using 2 SACK blocks in total, if we only need to use 2 AccECN fields: check ect1 arriving.
++.01 < [ect1] EAP. 5001:6001(1000) ack 1 win 264 <nop,nop,TS val 5 ecr 100>
++0.002 > [ect0] WA. 1:1(0) ack 1 <nop,nop,TS val 196 ecr 2,ECN e1b 3001 ceb 1000,nop,nop,sack 5001:6001 5001:6001>
+
+// Check the case where the AccECN option doesn't fit, because sending ect0
+// with order 1 would rquire 3 AccECN fields,
+// and TS (12 bytes) + 2 SACK blocks (20 bytes) + 3 AccECN fields (2 + 3*3 bytes) > 40 bytes.
+// That's OK; Linux TCP AccECN is optimized for the ECT1 case, not ECT0.
++.01 < [ect0] EAP. 6001:7001(1000) ack 1 win 264 <nop,nop,TS val 5 ecr 100>
++0.002 > [ect0] WA. 1:1(0) ack 1 <nop,nop,TS val 204 ecr 2,nop,nop,sack 5001:7001 3001:4001 1001:2001>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled1.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled1.pkt
new file mode 100644
index 000000000000..6b88ab78bfce
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled1.pkt
@@ -0,0 +1,20 @@
+// Test against classic ECN server
+// Not-ECT on SYN and server sets 1|0|1 (AE is unused for classic ECN)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] SEA. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,TS val 700 ecr 100,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <nop, nop, TS val 200 ecr 700>
+
++0 write(4, ..., 100) = 100
++.002 > [ect0] P.5 1:101(100) ack 1 <nop,nop,TS val 300 ecr 700>
++0 close(4) = 0
+
++.002 > [ect0] F.5 101:101(0) ack 1 <nop,nop,TS val 400 ecr 700>
++0.1 < [noecn] R. 1:1(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled2.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled2.pkt
new file mode 100644
index 000000000000..d24ada008ece
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_accecn_disabled2.pkt
@@ -0,0 +1,20 @@
+// Test against classic ECN server
+// Not-ECT on SYN and server sets 0|0|1
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] SE. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,TS val 700 ecr 100,nop,wscale 8>
++.002 > [noecn] . 1:1(0) ack 1 <nop, nop, TS val 200 ecr 700>
+
++0 write(4, ..., 100) = 100
++.002 > [ect0] P. 1:101(100) ack 1 <nop,nop,TS val 300 ecr 700>
++0 close(4) = 0
+
++0 > [noecn] F. 101:101(0) ack 1 <...>
++0.1 < R. 1:1(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_broken.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_broken.pkt
new file mode 100644
index 000000000000..a20d7e890ee1
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_broken.pkt
@@ -0,0 +1,19 @@
+// Test against broken server (1|1|1)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] SEWA. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,TS val 700 ecr 100,nop,wscale 8>
++.002 > [noecn] . 1:1(0) ack 1 <nop, nop, TS val 200 ecr 700>
+
++0 write(4, ..., 100) = 100
++.002 > [noecn] P. 1:101(100) ack 1 <nop,nop,TS val 300 ecr 700>
++0 close(4) = 0
+
++.002 > [noecn] F. 101:101(0) ack 1 <...>
++0.1 < [noecn] R. 1:1(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_ecn_disabled.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_ecn_disabled.pkt
new file mode 100644
index 000000000000..428255bedab7
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_ecn_disabled.pkt
@@ -0,0 +1,19 @@
+// Test against Non ECN server (0|0|0)
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [noecn] S. 0:0(0) ack 1 win 32767 <mss 1460,sackOK,TS val 700 ecr 100,nop,wscale 8>
++.002 > [noecn] . 1:1(0) ack 1 <nop, nop, TS val 200 ecr 700>
+
++0 write(4, ..., 100) = 100
++.002 > [noecn] P. 1:101(100) ack 1 <nop,nop,TS val 300 ecr 700>
++0 close(4) = 0
+
++.002 > [noecn] F. 101:101(0) ack 1 <nop,nop,TS val 400 ecr 700>
++0.1 < [noecn] R. 1:1(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_only.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_only.pkt
new file mode 100644
index 000000000000..e9a5a0d3677c
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_serverside_only.pkt
@@ -0,0 +1,18 @@
+// Test AccECN with sysctl set to server-side only
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=5
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_acked_after_retransmit.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_acked_after_retransmit.pkt
new file mode 100644
index 000000000000..412fa903105c
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_acked_after_retransmit.pkt
@@ -0,0 +1,18 @@
+// Test that SYN with ACE flags was Acked
+// after 2nd retransmission. In this case,
+// since we got SYN-ACK that supports Accurate
+// ECN, we consider this as successful negotiation
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 2.1 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++1~+1.1 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++1~+1.1 > [noecn] S 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
+
++0.1 < [noecn] SW. 0:0(0) ack 1 win 32767 <mss 1016,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++0~+0.01 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_drop.pkt
new file mode 100644
index 000000000000..4622754a2270
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ace_flags_drop.pkt
@@ -0,0 +1,16 @@
+// Test that SYN with ACE flags got dropped
+// We retry one more time with ACE and then
+// fallback to disabled ECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 2.1 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++1~+1.1 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++1~+1.1 > [noecn] S 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.1 < [noecn] S. 0:0(0) ack 1 win 32767 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0~+0.01 > [noecn] . 1:1(0) ack 1
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_acked_after_retransmit.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_acked_after_retransmit.pkt
new file mode 100644
index 000000000000..ee15f108cafe
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_acked_after_retransmit.pkt
@@ -0,0 +1,27 @@
+// Test that SYN-ACK with ACE flags was Acked
+// after 2nd retransmission. In this case,
+// since we got the last ACK that supports Accurate
+// ECN, we consider this as successful negotiation
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [noecn] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN-ACK without option
++1~+1.1 > [noecn] SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// SYN-ACK maybe getting blackholed, disable ECN
++2~+2.2 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// Received an ACK with ACE flags, state should be set to negotiation succeeded
++0.1 < [noecn] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_drop.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_drop.pkt
new file mode 100644
index 000000000000..3807e7fafafb
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ack_ace_flags_drop.pkt
@@ -0,0 +1,27 @@
+// Test that SYN-ACK with ACE flags got dropped
+// We retry one more time with ACE and then
+// fallback to disabled ECN
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [noecn] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN-ACK without option
++1~+1.1 > [noecn] SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// SYN-ACK maybe getting blackholed, disable ECN
++2~+2.2 > [noecn] S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+
+// Received an ACK with no ACE flags, state should be set to blackholed
++0.1 < [noecn] . 1:1(0) ack 1 win 320
++0 accept(3, ..., ...) = 4
+
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ce.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ce.pkt
new file mode 100644
index 000000000000..dc83f7a18180
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ce.pkt
@@ -0,0 +1,13 @@
+// Test AccECN ECN field reflector in SYNACK
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < [ce] SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SWA. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect0.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect0.pkt
new file mode 100644
index 000000000000..e63a8d018c37
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect0.pkt
@@ -0,0 +1,13 @@
+// Test AccECN ECN field reflector in SYNACK
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < [ect0] SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SA. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect1.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect1.pkt
new file mode 100644
index 000000000000..23c0e43b3dbe
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_syn_ect1.pkt
@@ -0,0 +1,13 @@
+// Test AccECN ECN field reflector in SYNACK
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < [ect1] SEWA 0:0(0) win 32792 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SEW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce.pkt
new file mode 100644
index 000000000000..44add14c57f4
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce.pkt
@@ -0,0 +1,28 @@
+// Test SYNACK CE & received_ce update
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > [noecn] SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [ce] SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] WA. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 write(4, ..., 100) = 100
++.002 > [ect0] P.6 1:101(100) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.05 < [ect0] P.5 1:101(100) ack 101 win 256 <ECN e0b 101 ceb 0 e1b 1,nop>
++.002 > [ect0] .6  101:101(0) ack 101 <ECN e1b 1 ceb 0 e0b 101,nop>
+
++0.01 write(4, ..., 100) = 100
++.002 > [ect0] P.6 101:201(100) ack 101 <ECN e1b 1 ceb 0 e0b 101,nop>
+
++0.1  < [ect1] P.5 201:301(100) ack 201 win 256 <ECN e0b 101 ceb 0 e1b 1,nop>
++.002 > [ect0] .6 201:201(0) ack 101 <ECN e1b 101 ceb 0 e0b 101,nop,nop,nop,sack 201:301>
+
++0.01 < [ce] .6 401:501(100) ack 201 win 256 <ECN e0b 101 ceb 0 e1b 1,nop>
++.002 > [ect0] .7 201:201(0) ack 101 <ECN e1b 101 ceb 100 e0b 101,nop,nop,nop,sack 401:501 201:301>
+
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce_updates_delivered_ce.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce_updates_delivered_ce.pkt
new file mode 100644
index 000000000000..5fd77f466572
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ce_updates_delivered_ce.pkt
@@ -0,0 +1,22 @@
+// Reflected SYNACK CE mark increases delivered_ce
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_fallback=0
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
++0.05 < SEWA 0:0(0) win 32767 <mss 1050,nop,nop,sackOK,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+// Fake ce for prev, ECT validator must be disabled for this to work
++0.05 < [ect0] WA. 1:1(0) ack 1 win 264 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 1, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect0.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect0.pkt
new file mode 100644
index 000000000000..f6ad1ea5c0c4
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect0.pkt
@@ -0,0 +1,24 @@
+// Test SYN=0 reflector
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] A. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 write(4, ..., 100) = 100
++.002 > [ect0] P.5 1:101(100) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.05 < [ect0] P.5 1:1(0) ack 101 win 256 <ECN e0b 101 ceb 0 e1b 1,nop>
+
++0.01 < [ect0] P.5 1:101(100) ack 101 win 256 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] .5 101:101(0) ack 101 <ECN e1b 1 ceb 0 e0b 101,nop>
++0 read(4, ..., 100) = 100
+
++0 close(4) = 0
++0 > F.5 101:101(0) ack 101 <...>
++0.1 < R. 101:101(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect1.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect1.pkt
new file mode 100644
index 000000000000..7ecfc5fb9dbb
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_ect1.pkt
@@ -0,0 +1,24 @@
+// Test SYN=0 reflector
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < [ect1] SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] EW. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.01 write(4, ..., 100) = 100
++.002 > [ect0] P.5 1:101(100) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
++0.05 < [ect1] P.5 1:1(0) ack 101 win 256 <ECN e0b 101 ceb 0 e1b 1,nop>
+
++0.01 < [ect1] P.5 1:101(100) ack 101 win 256 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 > [ect0] .5 101:101(0) ack 101 <ECN e1b 101 ceb 0 e0b 1,nop>
++0 read(4, ..., 100) = 100
+
++0 close(4) = 0
++0 > F5. 101:101(0) ack 101 <...>
++0.1 < R. 101:101(0) ack 102 win 4242
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rexmit.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rexmit.pkt
new file mode 100644
index 000000000000..9e0959782ef5
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rexmit.pkt
@@ -0,0 +1,15 @@
+// Test 3rd ACK flags when SYN-ACK is rexmitted
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
++.002 ... 0.052 connect(4, ..., ...) = 0
+
++.002 > SEWA 0:0(0) <mss 1460,sackOK,TS val 100 ecr 0,nop,wscale 8>
++0.05 < SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
+
++0.05 < SW. 0:0(0) ack 1 win 32767 <mss 1460,ECN e0b 1 ceb 0 e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
++.002 > [ect0] W. 1:1(0) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rxmt.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rxmt.pkt
new file mode 100644
index 000000000000..a5a41633af07
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_synack_rxmt.pkt
@@ -0,0 +1,25 @@
+// Test that we retransmit SYN-ACK with ACE and without
+// AccECN options after
+// SYN-ACK was lost and TCP moved to TCPS_SYN_RECEIVED
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+sysctl -q net.ipv4.tcp_ecn_option=2
+`
+
++0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
++0 bind(3, ..., ...) = 0
++0 listen(3, 1) = 0
+
++0 < [noecn] SEWA 0:0(0) win 32792 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++.002 > [noecn] SW. 0:0(0) ack 1 <mss 1460,ECN e1b 1 ceb 0 e0b 1,nop,nop,nop,sackOK,nop,wscale 8>
+
+// Retransmit SYN-ACK without option
++1~+1.1 > [noecn] SW. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
++0.1 < [noecn] W. 1:1(0) ack 1 win 320 <ECN e0b 1 ceb 0 e1b 1,nop>
++.002 accept(3, ..., ...) = 4
+
+// We try to write with AccECN option
++0.01 write(4, ..., 100) = 100
++.002 > [ect0] P5. 1:101(100) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_tsnoprogress.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_tsnoprogress.pkt
new file mode 100644
index 000000000000..f3fe2f098966
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_tsnoprogress.pkt
@@ -0,0 +1,26 @@
+// Test TS progress filtering
+--tcp_ts_tick_usecs=1000
+--tolerance_usecs=7000
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,sackOK,TS val 1 ecr 0,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,sackOK,TS val 10 ecr 1,ECN e1b 1 ceb 0 e0b 1,nop,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <nop,nop,TS val 2 ecr 10>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <nop,nop,TS val 83 ecr 2>
+  // Fake CE and claim no progress
++0.05 < [ect0] WA. 1:1(0) ack 1 win 264 <nop,nop,TS val 2 ecr 83>
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
diff --git a/tools/testing/selftests/net/packetdrill/tcp_accecn_tsprogress.pkt b/tools/testing/selftests/net/packetdrill/tcp_accecn_tsprogress.pkt
new file mode 100644
index 000000000000..1446799d2481
--- /dev/null
+++ b/tools/testing/selftests/net/packetdrill/tcp_accecn_tsprogress.pkt
@@ -0,0 +1,25 @@
+// Test TS progress filtering
+--tcp_ts_tick_usecs=1000
+
+`./defaults.sh
+sysctl -q net.ipv4.tcp_ecn=3
+`
+
+    0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+   +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+   +0 bind(3, ..., ...) = 0
+   +0 listen(3, 1) = 0
+
+   +0 < SEWA 0:0(0) win 32792 <mss 1050,sackOK,TS val 1 ecr 0,nop,wscale 8>
++.002 > SW. 0:0(0) ack 1 <mss 1460,sackOK,TS val 10 ecr 1,ECN e1b 1 ceb 0 e0b 1,nop,nop,wscale 8>
++0.05 < [ect0] W. 1:1(0) ack 1 win 264 <nop,nop,TS val 2 ecr 10>
++.002 accept(3, ..., ...) = 4
+
++0.01 %{ assert tcpi_delivered_ce == 0, tcpi_delivered_ce }%
+
++0.01 write(4, ..., 1000) = 1000
++.002 > [ect0] EAP. 1:1001(1000) ack 1 <nop,nop,TS val 83 ecr 2>
+  // Fake CE and claim no progress
++0.05 < [ect0] WA. 1:1(0) ack 1 win 264 <nop,nop,TS val 3 ecr 83>
+
++0.01 %{ assert tcpi_delivered_ce == 1, tcpi_delivered_ce }%
-- 
2.34.1

Re: [PATCH v9 net-next 01/15] tcp: try to avoid safer when ACKs are thinned

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Ilpo Järvinen <ij@kernel.org>
>
> Add newly acked pkts EWMA. When ACK thinning occurs, select
> between safer and unsafe cep delta in AccECN processing based
> on it. If the packets ACKed per ACK tends to be large, don't
> conservatively assume ACE field overflow.
>
> This patch uses the existing 2-byte holes in the rx group for new
> u16 variables withtout creating more holes. Below are the pahole
> outcomes before and after this patch:

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 02/15] gro: flushing when CWR is set negatively affects AccECN

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Ilpo Järvinen <ij@kernel.org>
>
> As AccECN may keep CWR bit asserted due to different
> interpretation of the bit, flushing with GRO because of
> CWR may effectively disable GRO until AccECN counter
> field changes such that CWR-bit becomes 0.
>
> There is no harm done from not immediately forwarding the
> CWR'ed segment with RFC3168 ECN.

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 03/15] selftests/net: gro: add self-test for TCP CWR flag

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Currently, GRO does not flush packets when the CWR bit is set.
> A corresponding self-test is being added, in which the CWR flag
> is set for two consecutive packets, but the first packet with the
> CWR flag set will not be flushed immediately.

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 04/15] tcp: ECT_1_NEGOTIATION and NEEDS_ACCECN identifiers

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Two CA module flags are added in this patch related to AccECN negotiation.
> First, a new CA module flag (TCP_CONG_NEEDS_ACCECN) defines that the CA
> expects to negotiate AccECN functionality using the ECE, CWR and AE flags
> in the TCP header.
>
> Second, during ECN negotiation, ECT(0) in the IP header is used. This patch
> enables CA to control whether ECT(0) or ECT(1) should be used on a per-segment
> basis. A new flag (TCP_CONG_ECT_1_NEGOTIATION) defines the expected ECT value
> in the IP header by the CA when not-yet initialized for the connection.
>
> The detailed AccECN negotiaotn during the 3WHS can be found in the AccECN spec:
>   https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt

While for some reason linux uses icsk_ca_ops, I think the terminology
is about "CC : Congestion Control"

Not sure what CA means...

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 05/15] tcp: disable RFC3168 fallback identifier for CC modules

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> When AccECN is not successfully negociated for a TCP flow, it defaults
> fallback to classic ECN (RFC3168). However, L4S service will fallback
> to non-ECN.
>
> This patch enables congestion control module to control whether it
> should not fallback to classic ECN after unsuccessful AccECN negotiation.
> A new CA module flag (TCP_CONG_NO_FALLBACK_RFC3168) identifies this
> behavior expected by the CA.
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> Acked-by: Paolo Abeni <pabeni@redhat.com>

Reviewed-by: Eric Dumazet <edumazet@google.com>

RE: [PATCH v9 net-next 04/15] tcp: ECT_1_NEGOTIATION and NEEDS_ACCECN identifiers

[Chia-Yu Chang (Nokia)]

> -----Original Message-----
> From: Eric Dumazet <edumazet@google.com> 
> Sent: Tuesday, January 20, 2026 10:54 AM
> To: Chia-Yu Chang (Nokia) <chia-yu.chang@nokia-bell-labs.com>
> Cc: pabeni@redhat.com; parav@nvidia.com; linux-doc@vger.kernel.org; corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; dave.taht@gmail.com; jhs@mojatatu.com; kuba@kernel.org; stephen@networkplumber.org; xiyou.wangcong@gmail.com; jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; ncardwell@google.com; Koen De Schepper (Nokia) <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; cheshire <cheshire@apple.com>; rs.ietf@gmx.at; Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>; Olivier Tilmans (Nokia) <olivier.tilmans@nokia.com>
> Subject: Re: [PATCH v9 net-next 04/15] tcp: ECT_1_NEGOTIATION and NEEDS_ACCECN identifiers
> 
> 
> CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.
> 
> 
> 
> On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
> >
> > From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> >
> > Two CA module flags are added in this patch related to AccECN negotiation.
> > First, a new CA module flag (TCP_CONG_NEEDS_ACCECN) defines that the 
> > CA expects to negotiate AccECN functionality using the ECE, CWR and AE 
> > flags in the TCP header.
> >
> > Second, during ECN negotiation, ECT(0) in the IP header is used. This 
> > patch enables CA to control whether ECT(0) or ECT(1) should be used on 
> > a per-segment basis. A new flag (TCP_CONG_ECT_1_NEGOTIATION) defines 
> > the expected ECT value in the IP header by the CA when not-yet initialized for the connection.
> >
> > The detailed AccECN negotiaotn during the 3WHS can be found in the AccECN spec:
> >   https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt
> 
> While for some reason linux uses icsk_ca_ops, I think the terminology is about "CC : Congestion Control"
> 
> Not sure what CA means...
> 
> Reviewed-by: Eric Dumazet <edumazet@google.com>

Thanks for feedback, I will change into CC in the next version.

Chia-Yu

Re: [PATCH v9 net-next 06/15] tcp: accecn: handle unexpected AccECN negotiation feedback

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> According to Section 3.1.2 of AccECN spec (RFC9768), if a TCP Client
> has sent a SYN requesting AccECN feedback with (AE,CWR,ECE) = (1,1,1)
> then receives a SYN/ACK with the currently reserved combination
> (AE,CWR,ECE) = (1,0,1) but it does not have logic specific to such a
> combination, the Client MUST enable AccECN mode as if the SYN/ACK
> confirmed that the Server supported AccECN and as if it fed back that
> the IP-ECN field on the SYN had arrived unchanged.

I find this a bit confusing.

3.1.2 has :

An AccECN implementation has no need to recognize or support the Server
response labelled 'Nonce' or ECN-nonce feedback more generally , as RFC 3540
has been reclassified as Historic . AccECN is compatible with alternative
 ECN feedback integrity approaches to the nonce (see Section 5.3).
 The SYN/ACK labelled 'Nonce' with (AE,CWR,ECE) = (1,0,1) is reserved
for future use.
A TCP Client (A) that receives such a SYN/ ACK follows the procedure
for forward compatibility given in Section 3.1.3.

The relevant section in the RFC is 3.1.2 _and_ 3.1.3 ?

Honestly, AccECN is way too complex for my taste :/

Please copy/paste the precise RFC parts, it will help future maintenance.

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 07/15] tcp: accecn: retransmit downgraded SYN in AccECN negotiation

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Based on AccECN spec (RFC9768), if the sender of an AccECN SYN
> (the TCP Client) times out before receiving the SYN/ACK, it SHOULD
> attempt to negotiate the use of AccECN at least one more time by
> continuing to set all three TCP ECN flags (AE,CWR,ECE) = (1,1,1) on
> the first retransmitted SYN (using the usual retransmission time-outs).
>
> If this first retransmission also fails to be acknowledged, in
> deployment scenarios where AccECN path traversal might be problematic,
> the TCP Client SHOULD send subsequent retransmissions of the SYN with
> the three TCP-ECN flags cleared (AE,CWR,ECE) = (0,0,0).
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> Acked-by: Paolo Abeni <pabeni@redhat.com>
>

Please amend the changelog to give the RFC precise relevant chapter
(3.1.4.1 if I am not mistaken)

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 08/15] tcp: add TCP_SYNACK_RETRANS synack_type

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Before this patch, retransmitted SYN/ACK did not have a specific synack_type;
> however, the upcoming patch needs to distinguish between retransmitted and
> non-retransmitted SYN/ACK for AccECN negotiation to transmit the fallback
> SYN/ACK during AccECN negotiation. Therefore, this patch introduces a new
> synack_type (TCP_SYNACK_RETRANS).
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> Acked-by: Paolo Abeni <pabeni@redhat.com>

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 09/15] tcp: accecn: retransmit SYN/ACK without AccECN option or non-AccECN SYN/ACK

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> For Accurate ECN, the first SYN/ACK sent by the TCP server shall set the
> ACE flag (see Table 1 of RFC9768) and the AccECN option to complete the
> capability negotiation. However, if the TCP server needs to retransmit such
> a SYN/ACK (for example, because it did not receive an ACK acknowledging its
> SYN/ACK, or received a second SYN requesting AccECN support), the TCP server
> retransmits the SYN/ACK without the AccECN option. This is because the
> SYN/ACK may be lost due to congestion, or a middlebox may block the AccECN
> option. Furthermore, if this retransmission also times out, to expedite
> connection establishment, the TCP server should retransmit the SYN/ACK with
> (AE,CWR,ECE) = (0,0,0) and without the AccECN option, while maintaining
> AccECN feedback mode.
>

Reviewed-by: Eric Dumazet <edumazet@google.com>

Re: [PATCH v9 net-next 10/15] tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Based on specification:
>   https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt
>
> Based on Section 3.1.5 of AccECN spec (RFC9768), a TCP Server in
> AccECN mode MUST NOT set ECT on any packet for the rest of the connection,
> if it has received or sent at least one valid SYN or Acceptable SYN/ACK
> with (AE,CWR,ECE) = (0,0,0) during the handshake.
>
> In addition, a host in AccECN mode that is feeding back the IP-ECN
> field on a SYN or SYN/ACK MUST feed back the IP-ECN field on the
> latest valid SYN or acceptable SYN/ACK to arrive.
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> ---
> v8:
> - Add new helper function tcp_accecn_ace_fail_send_set_retrans()
>
> v6:
> - Do not cast const struct request_sock into struct request_sock
> - Set tcp_accecn_fail_mode after calling tcp_rtx_synack().
> ---
>  include/net/tcp_ecn.h           |  7 +++++++
>  net/ipv4/inet_connection_sock.c |  3 +++
>  net/ipv4/tcp_input.c            |  2 ++
>  net/ipv4/tcp_minisocks.c        | 36 ++++++++++++++++++++++++---------
>  net/ipv4/tcp_output.c           |  3 ++-
>  net/ipv4/tcp_timer.c            |  2 ++
>  6 files changed, 42 insertions(+), 11 deletions(-)
>
> diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h
> index 796c613b5ef3..f5e1f6b1bec3 100644
> --- a/include/net/tcp_ecn.h
> +++ b/include/net/tcp_ecn.h
> @@ -97,6 +97,13 @@ static inline void tcp_accecn_fail_mode_set(struct tcp_sock *tp, u8 mode)
>         tp->accecn_fail_mode |= mode;
>  }
>
> +static inline void tcp_accecn_ace_fail_send_set_retrans(struct request_sock *req,
> +                                                       struct tcp_sock *tp)
> +{
> +       if (req->num_retrans > 1 && tcp_rsk(req)->accecn_ok)
> +               tcp_accecn_fail_mode_set(tp, TCP_ACCECN_ACE_FAIL_SEND);
> +}
> +
>  #define TCP_ACCECN_OPT_NOT_SEEN                0x0
>  #define TCP_ACCECN_OPT_EMPTY_SEEN      0x1
>  #define TCP_ACCECN_OPT_COUNTER_SEEN    0x2
> diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
> index 97d57c52b9ad..9d16cb9c3db4 100644
> --- a/net/ipv4/inet_connection_sock.c
> +++ b/net/ipv4/inet_connection_sock.c
> @@ -20,6 +20,7 @@
>  #include <net/tcp_states.h>
>  #include <net/xfrm.h>
>  #include <net/tcp.h>
> +#include <net/tcp_ecn.h>
>  #include <net/sock_reuseport.h>
>  #include <net/addrconf.h>
>
> @@ -1103,6 +1104,8 @@ static void reqsk_timer_handler(struct timer_list *t)
>             (!resend ||
>              !tcp_rtx_synack(sk_listener, req) ||
>              inet_rsk(req)->acked)) {
> +               tcp_accecn_ace_fail_send_set_retrans(req,
> +                                                    tcp_sk(sk_listener));


Ouch.

I think you missed the fact that a listener is shared by many SYN_RECV requests.

Consider it as read-only here.

Re: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Add 2-bit tcpi_ecn_mode feild within tcp_info to indicate which ECN
> mode is negotiated: ECN_MODE_DISABLED, ECN_MODE_RFC3168, ECN_MODE_ACCECN,
> or ECN_MODE_PENDING. This is done by utilizing available bits from
> tcpi_accecn_opt_seen (reduced from 16 bits to 2 bits) and
> tcpi_accecn_fail_mode (reduced from 16 bits to 4 bits).
>
> Also, an extra 24-bit tcpi_options2 field is identified to represent
> newer options and connection features, as all 8 bits of tcpi_options
> field have been used.
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> Co-developed-by: Neal Cardwell <ncardwell@google.com>
> Signed-off-by: Neal Cardwell <ncardwell@google.com>

Are you sure Neal Cardwell really is ok with this patch, saw it and
gave his SOB ?

> ---

>  struct tcp_info {
>         __u8    tcpi_state;
>         __u8    tcpi_ca_state;
> @@ -316,15 +334,17 @@ struct tcp_info {
>                                          * in milliseconds, including any
>                                          * unfinished recovery.
>                                          */
> -       __u32   tcpi_received_ce;    /* # of CE marks received */
> +       __u32   tcpi_ecn_mode:2,
> +               tcpi_accecn_opt_seen:2,
> +               tcpi_accecn_fail_mode:4,
> +               tcpi_options2:24;
> +       __u32   tcpi_received_ce;    /* # of CE marked segments received */
>         __u32   tcpi_delivered_e1_bytes;  /* Accurate ECN byte counters */
>         __u32   tcpi_delivered_e0_bytes;
>         __u32   tcpi_delivered_ce_bytes;
>         __u32   tcpi_received_e1_bytes;
>         __u32   tcpi_received_e0_bytes;
>         __u32   tcpi_received_ce_bytes;
> -       __u16   tcpi_accecn_fail_mode;
> -       __u16   tcpi_accecn_opt_seen;
>  };

tcp_info is ABI.

We can not add/remove fields in the middle.

You must add fields at the end of it only.

Re: [PATCH v9 net-next 14/15] tcp: accecn: enable AccECN

[Eric Dumazet]

On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
>
> From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
>
> Enable Accurate ECN negotiation and request for incoming and
> outgoing connection by setting sysctl_tcp_ecn:
>
> +==============+===========================================+
> |              |  Highest ECN variant (Accurate ECN, ECN,  |
> |   tcp_ecn    |  or no ECN) to be negotiated & requested  |
> |              +---------------------+---------------------+
> |              | Incoming connection | Outgoing connection |
> +==============+=====================+=====================+
> |      0       |        No ECN       |        No ECN       |
> |      1       |         ECN         |         ECN         |
> |      2       |         ECN         |        No ECN       |
> +--------------+---------------------+---------------------+
> |      3       |     Accurate ECN    |     Accurate ECN    |
> |      4       |     Accurate ECN    |         ECN         |
> |      5       |     Accurate ECN    |        No ECN       |
> +==============+=====================+=====================+
>
> Refer Documentation/networking/ip-sysctl.rst for more details.
>
> Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> Acked-by: Paolo Abeni <pabeni@redhat.com>
> ---

Reviewed-by: Eric Dumazet <edumazet@google.com>

RE: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info

[Chia-Yu Chang (Nokia)]

> -----Original Message-----
> From: Eric Dumazet <edumazet@google.com> 
> Sent: Tuesday, January 20, 2026 12:18 PM
> To: Chia-Yu Chang (Nokia) <chia-yu.chang@nokia-bell-labs.com>
> Cc: pabeni@redhat.com; parav@nvidia.com; linux-doc@vger.kernel.org; corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; dave.taht@gmail.com; jhs@mojatatu.com; kuba@kernel.org; stephen@networkplumber.org; xiyou.wangcong@gmail.com; jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; ncardwell@google.com; Koen De Schepper (Nokia) <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; cheshire <cheshire@apple.com>; rs.ietf@gmx.at; Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>
> Subject: Re: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info
> 
> 
> CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.
> 
> 
> 
> On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
> >
> > From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> >
> > Add 2-bit tcpi_ecn_mode feild within tcp_info to indicate which ECN 
> > mode is negotiated: ECN_MODE_DISABLED, ECN_MODE_RFC3168, 
> > ECN_MODE_ACCECN, or ECN_MODE_PENDING. This is done by utilizing 
> > available bits from tcpi_accecn_opt_seen (reduced from 16 bits to 2 
> > bits) and tcpi_accecn_fail_mode (reduced from 16 bits to 4 bits).
> >
> > Also, an extra 24-bit tcpi_options2 field is identified to represent 
> > newer options and connection features, as all 8 bits of tcpi_options 
> > field have been used.
> >
> > Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> > Co-developed-by: Neal Cardwell <ncardwell@google.com>
> > Signed-off-by: Neal Cardwell <ncardwell@google.com>
> 
> Are you sure Neal Cardwell really is ok with this patch, saw it and gave his SOB ?
> 
> > ---
> 
> >  struct tcp_info {
> >         __u8    tcpi_state;
> >         __u8    tcpi_ca_state;
> > @@ -316,15 +334,17 @@ struct tcp_info {
> >                                          * in milliseconds, including any
> >                                          * unfinished recovery.
> >                                          */
> > -       __u32   tcpi_received_ce;    /* # of CE marks received */
> > +       __u32   tcpi_ecn_mode:2,
> > +               tcpi_accecn_opt_seen:2,
> > +               tcpi_accecn_fail_mode:4,
> > +               tcpi_options2:24;
> > +       __u32   tcpi_received_ce;    /* # of CE marked segments received */
> >         __u32   tcpi_delivered_e1_bytes;  /* Accurate ECN byte counters */
> >         __u32   tcpi_delivered_e0_bytes;
> >         __u32   tcpi_delivered_ce_bytes;
> >         __u32   tcpi_received_e1_bytes;
> >         __u32   tcpi_received_e0_bytes;
> >         __u32   tcpi_received_ce_bytes;
> > -       __u16   tcpi_accecn_fail_mode;
> > -       __u16   tcpi_accecn_opt_seen;
> >  };
> 
> tcp_info is ABI.
> 
> We can not add/remove fields in the middle.
> 
> You must add fields at the end of it only.

Sure, I will update this is the next version and submit new corresponding patch to packetdrill (after this patch is accepted.)

Chia-Yu

RE: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info

[Chia-Yu Chang (Nokia)]

> -----Original Message-----
> From: Chia-Yu Chang (Nokia) 
> Sent: Tuesday, January 20, 2026 12:38 PM
> To: 'Eric Dumazet' <edumazet@google.com>
> Cc: pabeni@redhat.com; parav@nvidia.com; linux-doc@vger.kernel.org; corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; dave.taht@gmail.com; jhs@mojatatu.com; kuba@kernel.org; stephen@networkplumber.org; xiyou.wangcong@gmail.com; jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; ncardwell@google.com; Koen De Schepper (Nokia) <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; cheshire <cheshire@apple.com>; rs.ietf@gmx.at; Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>
> Subject: RE: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode and tcpi_option2 in tcp_info
> 
> > -----Original Message-----
> > From: Eric Dumazet <edumazet@google.com>
> > Sent: Tuesday, January 20, 2026 12:18 PM
> > To: Chia-Yu Chang (Nokia) <chia-yu.chang@nokia-bell-labs.com>
> > Cc: pabeni@redhat.com; parav@nvidia.com; linux-doc@vger.kernel.org; 
> > corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; 
> > kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; 
> > dave.taht@gmail.com; jhs@mojatatu.com; kuba@kernel.org; 
> > stephen@networkplumber.org; xiyou.wangcong@gmail.com; 
> > jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; 
> > donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; 
> > shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; 
> > ncardwell@google.com; Koen De Schepper (Nokia) 
> > <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; 
> > ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; 
> > cheshire <cheshire@apple.com>; rs.ietf@gmx.at; 
> > Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>
> > Subject: Re: [PATCH v9 net-next 13/15] tcp: accecn: add tcpi_ecn_mode 
> > and tcpi_option2 in tcp_info
> > 
> > 
> > CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.
> > 
> > 
> > 
> > On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
> > >
> > > From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> > >
> > > Add 2-bit tcpi_ecn_mode feild within tcp_info to indicate which ECN 
> > > mode is negotiated: ECN_MODE_DISABLED, ECN_MODE_RFC3168, 
> > > ECN_MODE_ACCECN, or ECN_MODE_PENDING. This is done by utilizing 
> > > available bits from tcpi_accecn_opt_seen (reduced from 16 bits to 2
> > > bits) and tcpi_accecn_fail_mode (reduced from 16 bits to 4 bits).
> > >
> > > Also, an extra 24-bit tcpi_options2 field is identified to represent 
> > > newer options and connection features, as all 8 bits of tcpi_options 
> > > field have been used.
> > >
> > > Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> > > Co-developed-by: Neal Cardwell <ncardwell@google.com>
> > > Signed-off-by: Neal Cardwell <ncardwell@google.com>
> > 
> > Are you sure Neal Cardwell really is ok with this patch, saw it and gave his SOB ?

This was discussed in another thread with Neal: "[PATCH net-next 1/1] selftests/net: Add packetdrill packetdrill cases". And Neal/me discussed this change.
But it was my miss that not syncing live update of tcp_info in packetdrill.
So, I will fix in the next net-next, thanks.

Chia-Yu

RE: [PATCH v9 net-next 10/15] tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion

[Chia-Yu Chang (Nokia)]

> -----Original Message-----
> From: Eric Dumazet <edumazet@google.com> 
> Sent: Tuesday, January 20, 2026 12:05 PM
> To: Chia-Yu Chang (Nokia) <chia-yu.chang@nokia-bell-labs.com>
> Cc: pabeni@redhat.com; parav@nvidia.com; linux-doc@vger.kernel.org; corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; dave.taht@gmail.com; jhs@mojatatu.com; kuba@kernel.org; stephen@networkplumber.org; xiyou.wangcong@gmail.com; jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; ncardwell@google.com; Koen De Schepper (Nokia) <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; cheshire <cheshire@apple.com>; rs.ietf@gmx.at; Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>
> Subject: Re: [PATCH v9 net-next 10/15] tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion
> 
> 
> CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.
> 
> 
> 
> On Mon, Jan 19, 2026 at 7:59 PM <chia-yu.chang@nokia-bell-labs.com> wrote:
> >
> > From: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> >
> > Based on specification:
> >   https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt
> >
> > Based on Section 3.1.5 of AccECN spec (RFC9768), a TCP Server in 
> > AccECN mode MUST NOT set ECT on any packet for the rest of the 
> > connection, if it has received or sent at least one valid SYN or 
> > Acceptable SYN/ACK with (AE,CWR,ECE) = (0,0,0) during the handshake.
> >
> > In addition, a host in AccECN mode that is feeding back the IP-ECN 
> > field on a SYN or SYN/ACK MUST feed back the IP-ECN field on the 
> > latest valid SYN or acceptable SYN/ACK to arrive.
> >
> > Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
> >
> > ---
> > v8:
> > - Add new helper function tcp_accecn_ace_fail_send_set_retrans()
> >
> > v6:
> > - Do not cast const struct request_sock into struct request_sock
> > - Set tcp_accecn_fail_mode after calling tcp_rtx_synack().
> > ---
> >  include/net/tcp_ecn.h           |  7 +++++++
> >  net/ipv4/inet_connection_sock.c |  3 +++
> >  net/ipv4/tcp_input.c            |  2 ++
> >  net/ipv4/tcp_minisocks.c        | 36 ++++++++++++++++++++++++---------
> >  net/ipv4/tcp_output.c           |  3 ++-
> >  net/ipv4/tcp_timer.c            |  2 ++
> >  6 files changed, 42 insertions(+), 11 deletions(-)
> >
> > diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h index 
> > 796c613b5ef3..f5e1f6b1bec3 100644
> > --- a/include/net/tcp_ecn.h
> > +++ b/include/net/tcp_ecn.h
> > @@ -97,6 +97,13 @@ static inline void tcp_accecn_fail_mode_set(struct tcp_sock *tp, u8 mode)
> >         tp->accecn_fail_mode |= mode;
> >  }
> >
> > +static inline void tcp_accecn_ace_fail_send_set_retrans(struct request_sock *req,
> > +                                                       struct 
> > +tcp_sock *tp) {
> > +       if (req->num_retrans > 1 && tcp_rsk(req)->accecn_ok)
> > +               tcp_accecn_fail_mode_set(tp, 
> > +TCP_ACCECN_ACE_FAIL_SEND); }
> > +
> >  #define TCP_ACCECN_OPT_NOT_SEEN                0x0
> >  #define TCP_ACCECN_OPT_EMPTY_SEEN      0x1
> >  #define TCP_ACCECN_OPT_COUNTER_SEEN    0x2
> > diff --git a/net/ipv4/inet_connection_sock.c 
> > b/net/ipv4/inet_connection_sock.c index 97d57c52b9ad..9d16cb9c3db4 
> > 100644
> > --- a/net/ipv4/inet_connection_sock.c
> > +++ b/net/ipv4/inet_connection_sock.c
> > @@ -20,6 +20,7 @@
> >  #include <net/tcp_states.h>
> >  #include <net/xfrm.h>
> >  #include <net/tcp.h>
> > +#include <net/tcp_ecn.h>
> >  #include <net/sock_reuseport.h>
> >  #include <net/addrconf.h>
> >
> > @@ -1103,6 +1104,8 @@ static void reqsk_timer_handler(struct timer_list *t)
> >             (!resend ||
> >              !tcp_rtx_synack(sk_listener, req) ||
> >              inet_rsk(req)->acked)) {
> > +               tcp_accecn_ace_fail_send_set_retrans(req,
> > +                                                    
> > + tcp_sk(sk_listener));
> 
> 
> Ouch.
> 
> I think you missed the fact that a listener is shared by many SYN_RECV requests.
> 
> Consider it as read-only here.

Hi Eric,

Thanks for the feedback.
Do you mean sk_listener here is read-only despite there is no const here?

Then, could you help to suggest the way please?
Beacuse for AccECN, here we need to set fail flag after retransmitting SYN/ACK > 1 time.
And this was done within tcp_make_synack(), but now move to every place where could retransmit SYN/ACK.

Thanks.
Chia-Yu

Re: [PATCH v9 net-next 10/15] tcp: accecn: unset ECT if receive or send ACE=0 in AccECN negotiaion

[Eric Dumazet]

On Tue, Jan 20, 2026 at 7:11 PM Chia-Yu Chang (Nokia)
<chia-yu.chang@nokia-bell-labs.com> wrote:
>
>
> Hi Eric,
>
> Thanks for the feedback.
> Do you mean sk_listener here is read-only despite there is no const here?

It is not const because we probably need to increment reference counts on it.

But if you have 1000 SYN_RECV, they might share the same listener
socket, and we do not lock the listener socket,
this would not scale very well on servers with 10,000,000 tcp sockets :)

So using any listener-fields to store 'per-syn-recv' information is racy.

>
> Then, could you help to suggest the way please?
> Beacuse for AccECN, here we need to set fail flag after retransmitting SYN/ACK > 1 time.

Why not use state in req itself ? (Or tcp_rsk())

> And this was done within tcp_make_synack(), but now move to every place where could retransmit SYN/ACK.
>
> Thanks.
> Chia-Yu

Re: [PATCH v9 net-next 15/15] selftests/net: packetdrill: add TCP Accurate ECN cases

[Jakub Kicinski]

On Mon, 19 Jan 2026 19:58:52 +0100 chia-yu.chang@nokia-bell-labs.com
wrote:
> Linux Accurate ECN test sets using ACE counters and AccECN options to
> cover several scenarios: Connection teardown, different ACK conditions,
> counter wrapping, SACK space grabbing, fallback schemes, negotiation
> retransmission/reorder/loss, AccECN option drop/loss, different
> handshake reflectors, data with marking, and different sysctl values.

Thank you for closing the packetdrill side, and big thanks to Neal 
for prioritizing getting it reviewed and merged!

I updated the packetdrill build in netdev CI and looks like one of
the cases is flaking a little. Since it looks like you'll have to
respin, please try to fix:

# 1..2
# tcp_accecn_client_accecn_options_lost.pkt:32: error handling packet: timing error: expected outbound packet in relative time range +0.020000~+0.500000 sec but happened at +0.015816 sec
# script packet:  0.181936 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
# actual packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 <ECN e1b 1 ceb 0 e0b 1,nop>
# not ok 1 ipv4
# tcp_accecn_client_accecn_options_lost.pkt:32: error handling packet: timing error: expected outbound packet in relative time range +0.020000~+0.500000 sec but happened at +0.015800 sec
# script packet:  0.181952 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
# actual packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 <ECN e1b 1 ceb 0 e0b 1,nop>
# not ok 2 ipv6
# # Totals: pass:0 fail:2 xfail:0 xpass:0 skip:0 error:0

https://netdev-ctrl.bots.linux.dev/logs/vmksft/packetdrill/results/482201/115-tcp-accecn-client-accecn-options-lost-pkt/stdout

Re: [PATCH v9 net-next 15/15] selftests/net: packetdrill: add TCP Accurate ECN cases

[Neal Cardwell]

On Tue, Jan 20, 2026 at 1:53 PM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Mon, 19 Jan 2026 19:58:52 +0100 chia-yu.chang@nokia-bell-labs.com
> wrote:
> > Linux Accurate ECN test sets using ACE counters and AccECN options to
> > cover several scenarios: Connection teardown, different ACK conditions,
> > counter wrapping, SACK space grabbing, fallback schemes, negotiation
> > retransmission/reorder/loss, AccECN option drop/loss, different
> > handshake reflectors, data with marking, and different sysctl values.
>
> Thank you for closing the packetdrill side, and big thanks to Neal
> for prioritizing getting it reviewed and merged!
>
> I updated the packetdrill build in netdev CI and looks like one of
> the cases is flaking a little. Since it looks like you'll have to
> respin, please try to fix:
>
> # 1..2
> # tcp_accecn_client_accecn_options_lost.pkt:32: error handling packet: timing error: expected outbound packet in relative time range +0.020000~+0.500000 sec but happened at +0.015816 sec
> # script packet:  0.181936 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
> # actual packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 <ECN e1b 1 ceb 0 e0b 1,nop>
> # not ok 1 ipv4
> # tcp_accecn_client_accecn_options_lost.pkt:32: error handling packet: timing error: expected outbound packet in relative time range +0.020000~+0.500000 sec but happened at +0.015800 sec
> # script packet:  0.181952 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop>
> # actual packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 <ECN e1b 1 ceb 0 e0b 1,nop>
> # not ok 2 ipv6
> # # Totals: pass:0 fail:2 xfail:0 xpass:0 skip:0 error:0
>
> https://netdev-ctrl.bots.linux.dev/logs/vmksft/packetdrill/results/482201/115-tcp-accecn-client-accecn-options-lost-pkt/stdout

Probably this is happening because the SRTT is around 56ms:

.050 * 7/8 + 1/8 * .1 = .05625 sec

So the RACK fast recovery starts afte rabout 15ms due to .25 * srtt
being about 14ms:
(.050 * 7/8 + 1/8 * .1) * .25 = .0140625 sec

If we make the SRTT 100ms then the fast retransmit should be around:

(.1 * 7/8 + 1/8 * .1) * .25 = .025 sec

So I'd suggest changing the timing of the SYNACK from 50ms to 100ms:

old:
+0.05 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1024,ECN e0b 1 ceb 0
e1b 1,nop,nop,nop,sackOK,nop,wscale 8>

new:
+.1 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1024,ECN e0b 1 ceb 0 e1b
1,nop,nop,nop,sackOK,nop,wscale 8>

neal

RE: [PATCH v9 net-next 15/15] selftests/net: packetdrill: add TCP Accurate ECN cases

[Chia-Yu Chang (Nokia)]

> -----Original Message-----
> From: Neal Cardwell <ncardwell@google.com> 
> Sent: Tuesday, January 20, 2026 8:35 PM
> To: Jakub Kicinski <kuba@kernel.org>
> Cc: Chia-Yu Chang (Nokia) <chia-yu.chang@nokia-bell-labs.com>; pabeni@redhat.com; edumazet@google.com; parav@nvidia.com; linux-doc@vger.kernel.org; corbet@lwn.net; horms@kernel.org; dsahern@kernel.org; kuniyu@google.com; bpf@vger.kernel.org; netdev@vger.kernel.org; dave.taht@gmail.com; jhs@mojatatu.com; stephen@networkplumber.org; xiyou.wangcong@gmail.com; jiri@resnulli.us; davem@davemloft.net; andrew+netdev@lunn.ch; donald.hunter@gmail.com; ast@fiberby.net; liuhangbin@gmail.com; shuah@kernel.org; linux-kselftest@vger.kernel.org; ij@kernel.org; Koen De Schepper (Nokia) <koen.de_schepper@nokia-bell-labs.com>; g.white@cablelabs.com; ingemar.s.johansson@ericsson.com; mirja.kuehlewind@ericsson.com; cheshire <cheshire@apple.com>; rs.ietf@gmx.at; Jason_Livingood@comcast.com; Vidhi Goel <vidhi_goel@apple.com>
> Subject: Re: [PATCH v9 net-next 15/15] selftests/net: packetdrill: add TCP Accurate ECN cases
> 
> 
> CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.
> 
> 
> 
> On Tue, Jan 20, 2026 at 1:53 PM Jakub Kicinski <kuba@kernel.org> wrote:
> >
> > On Mon, 19 Jan 2026 19:58:52 +0100 chia-yu.chang@nokia-bell-labs.com
> > wrote:
> > > Linux Accurate ECN test sets using ACE counters and AccECN options 
> > > to cover several scenarios: Connection teardown, different ACK 
> > > conditions, counter wrapping, SACK space grabbing, fallback schemes, 
> > > negotiation retransmission/reorder/loss, AccECN option drop/loss, 
> > > different handshake reflectors, data with marking, and different sysctl values.
> >
> > Thank you for closing the packetdrill side, and big thanks to Neal for 
> > prioritizing getting it reviewed and merged!
> >
> > I updated the packetdrill build in netdev CI and looks like one of the 
> > cases is flaking a little. Since it looks like you'll have to respin, 
> > please try to fix:
> >
> > # 1..2
> > # tcp_accecn_client_accecn_options_lost.pkt:32: error handling packet: 
> > timing error: expected outbound packet in relative time range 
> > +0.020000~+0.500000 sec but happened at +0.015816 sec # script packet:  
> > 0.181936 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 e0b 1,nop> # actual 
> > packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 <ECN e1b 1 ceb 0 e0b 
> > 1,nop> # not ok 1 ipv4 # tcp_accecn_client_accecn_options_lost.pkt:32: 
> > error handling packet: timing error: expected outbound packet in 
> > relative time range +0.020000~+0.500000 sec but happened at +0.015800 
> > sec # script packet:  0.181952 .5 1:1013(1012) ack 1 <ECN e1b 1 ceb 0 
> > e0b 1,nop> # actual packet:  0.177752 .EA 1:1013(1012) ack 1 win 1050 
> > <ECN e1b 1 ceb 0 e0b 1,nop> # not ok 2 ipv6 # # Totals: pass:0 fail:2 
> > xfail:0 xpass:0 skip:0 error:0
> >
> > https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fnetd
> > ev-ctrl.bots.linux.dev%2Flogs%2Fvmksft%2Fpacketdrill%2Fresults%2F48220
> > 1%2F115-tcp-accecn-client-accecn-options-lost-pkt%2Fstdout&data=05%7C0
> > 2%7Cchia-yu.chang%40nokia-bell-labs.com%7Cf125ccafc7134b620bad08de585b
> > 1a35%7C5d4717519675428d917b70f44f9630b0%7C0%7C0%7C639045345459758258%7
> > CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlA
> > iOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=kn3Pb
> > VXw%2Bkznnf7VaYzwP%2FL3IO2LYGYQzZOWS3HRZ6w%3D&reserved=0
> 
> Probably this is happening because the SRTT is around 56ms:
> 
> .050 * 7/8 + 1/8 * .1 = .05625 sec
> 
> So the RACK fast recovery starts afte rabout 15ms due to .25 * srtt being about 14ms:
> (.050 * 7/8 + 1/8 * .1) * .25 = .0140625 sec
> 
> If we make the SRTT 100ms then the fast retransmit should be around:
> 
> (.1 * 7/8 + 1/8 * .1) * .25 = .025 sec
> 
> So I'd suggest changing the timing of the SYNACK from 50ms to 100ms:
> 
> old:
> +0.05 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1024,ECN e0b 1 ceb 0
> e1b 1,nop,nop,nop,sackOK,nop,wscale 8>
> 
> new:
> +.1 < [ect0] SW. 0:0(0) ack 1 win 32767 <mss 1024,ECN e0b 1 ceb 0 e1b
> 1,nop,nop,nop,sackOK,nop,wscale 8>
> 
> neal

Thanks Neal and Eric, I've fixed this issue as well as the concerned raised in patch 10 of listen socket.
All AccECN packetdrill still pass from my end, so I will submit v10.

Thanks.
Chai-Yu