* [PATCH net-next 1/5] tcp: move tp->delivered and tp->delivered_ce to tcp_sock_write_tx group
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
@ 2026-04-30 10:00 ` Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 2/5] tcp: move tp->segs_in and tp->segs_out to tcp_sock_write_txrx group Eric Dumazet
` (4 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Eric Dumazet @ 2026-04-30 10:00 UTC (permalink / raw)
To: David S . Miller, Jakub Kicinski, Paolo Abeni
Cc: Simon Horman, Neal Cardwell, Kuniyuki Iwashima, netdev,
eric.dumazet, Eric Dumazet
These counters are changed whenever sent data is acknowleged.
They do not belong to tcp_sock_write_txrx group, because TCP receivers
do not touch them.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
Documentation/networking/net_cachelines/tcp_sock.rst | 4 ++--
include/linux/tcp.h | 4 ++--
net/ipv4/tcp.c | 4 ++--
3 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/Documentation/networking/net_cachelines/tcp_sock.rst b/Documentation/networking/net_cachelines/tcp_sock.rst
index fecf61166a54ee2f64bcef5312c81dcc4aa9a124..f2eafc933b5fa6aff5222cc486c00c4dbd437f92 100644
--- a/Documentation/networking/net_cachelines/tcp_sock.rst
+++ b/Documentation/networking/net_cachelines/tcp_sock.rst
@@ -99,8 +99,8 @@ u32 snd_cwnd_stamp
u32 prior_cwnd
u32 prr_delivered
u32 prr_out read_mostly read_mostly tcp_rate_skb_sent,tcp_newly_delivered(tx);tcp_ack,tcp_rate_gen,tcp_clean_rtx_queue(rx)
-u32 delivered read_mostly read_write tcp_rate_skb_sent, tcp_newly_delivered(tx);tcp_ack, tcp_rate_gen, tcp_clean_rtx_queue (rx)
-u32 delivered_ce read_mostly read_write tcp_rate_skb_sent(tx);tcp_rate_gen(rx)
+u32 delivered read_write tcp_rate_skb_sent, tcp_newly_delivered(tx);tcp_ack, tcp_rate_gen, tcp_clean_rtx_queue (rx)
+u32 delivered_ce read_write tcp_rate_skb_sent(tx);tcp_rate_gen(rx)
u32 received_ce read_mostly read_write
u32[3] received_ecn_bytes read_mostly read_write
u8:4 received_ce_pending read_mostly read_write
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 6982f10e826b4004f210ea22c94f0488e52184d1..3e20bffd6ae908f18c1a88ed97db78f7ea8989ae 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -259,6 +259,8 @@ struct tcp_sock {
u32 data_segs_out; /* RFC4898 tcpEStatsPerfDataSegsOut
* total number of data segments sent.
*/
+ u32 delivered; /* Total data packets delivered incl. rexmits */
+ u32 delivered_ce; /* Like the above but only ECE marked packets */
u64 bytes_sent; /* RFC4898 tcpEStatsPerfHCDataOctetsOut
* total number of data bytes sent.
*/
@@ -307,8 +309,6 @@ struct tcp_sock {
u32 srtt_us; /* smoothed round trip time << 3 in usecs */
u32 packets_out; /* Packets which are "in flight" */
u32 snd_up; /* Urgent pointer */
- u32 delivered; /* Total data packets delivered incl. rexmits */
- u32 delivered_ce; /* Like the above but only ECE marked packets */
u32 received_ce; /* Like the above but for rcvd CE marked pkts */
u32 received_ecn_bytes[3]; /* received byte counters for three ECN
* types: INET_ECN_ECT_1, INET_ECN_ECT_0,
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 432fa28e47d4c8ef5d50339bfdf7da0ea8772b94..fe1e54321e181c307227c8f2f5a9c0fb2816c762 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -5232,6 +5232,8 @@ static void __init tcp_struct_check(void)
/* TX read-write hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, data_segs_out);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_ce);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, bytes_sent);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, snd_sml);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, chrono_start);
@@ -5258,8 +5260,6 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, srtt_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, packets_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, snd_up);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, delivered);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, delivered_ce);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, received_ce);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, received_ecn_bytes);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, app_limited);
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH net-next 2/5] tcp: move tp->segs_in and tp->segs_out to tcp_sock_write_txrx group
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 1/5] tcp: move tp->delivered and tp->delivered_ce to tcp_sock_write_tx group Eric Dumazet
@ 2026-04-30 10:00 ` Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 3/5] tcp: move tp->first_tx_mstamp and tp->delivered_mstamp to tcp_sock_write_tx Eric Dumazet
` (3 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Eric Dumazet @ 2026-04-30 10:00 UTC (permalink / raw)
To: David S . Miller, Jakub Kicinski, Paolo Abeni
Cc: Simon Horman, Neal Cardwell, Kuniyuki Iwashima, netdev,
eric.dumazet, Eric Dumazet
segs_in is changed for each incoming packet, including ACK packets.
segs_out is changed for each outgoing packet, including ACK packets.
They belong to tcp_sock_write_txrx group.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
.../networking/net_cachelines/tcp_sock.rst | 4 ++--
include/linux/tcp.h | 18 +++++++++---------
net/ipv4/tcp.c | 4 ++--
3 files changed, 13 insertions(+), 13 deletions(-)
diff --git a/Documentation/networking/net_cachelines/tcp_sock.rst b/Documentation/networking/net_cachelines/tcp_sock.rst
index f2eafc933b5fa6aff5222cc486c00c4dbd437f92..b914382cb8fd7ae739297180ff5101acb8dc8a14 100644
--- a/Documentation/networking/net_cachelines/tcp_sock.rst
+++ b/Documentation/networking/net_cachelines/tcp_sock.rst
@@ -13,13 +13,13 @@ u16 tcp_header_len read_mostly read_m
u16 gso_segs read_mostly tcp_xmit_size_goal
__be32 pred_flags read_write read_mostly tcp_select_window(tx);tcp_rcv_established(rx)
u64 bytes_received read_write tcp_rcv_nxt_update(rx)
-u32 segs_in read_write tcp_v6_rcv(rx)
+u32 segs_in read_write read_write tcp_segs_in(),tcp_v6_rcv(rx),tcp_v4_rcv()
u32 data_segs_in read_write tcp_v6_rcv(rx)
u32 rcv_nxt read_mostly read_write tcp_cleanup_rbuf,tcp_send_ack,tcp_inq_hint,tcp_transmit_skb,tcp_receive_window(tx);tcp_v6_do_rcv,tcp_rcv_established,tcp_data_queue,tcp_receive_window,tcp_rcv_nxt_update(write)(rx)
u32 copied_seq read_mostly tcp_cleanup_rbuf,tcp_rcv_space_adjust,tcp_inq_hint
u32 rcv_wup read_write __tcp_cleanup_rbuf,tcp_receive_window,tcp_receive_established
u32 snd_nxt read_write read_mostly tcp_rate_check_app_limited,__tcp_transmit_skb,tcp_event_new_data_sent(write)(tx);tcp_rcv_established,tcp_ack,tcp_clean_rtx_queue(rx)
-u32 segs_out read_write __tcp_transmit_skb
+u32 segs_out read_write read_write __tcp_transmit_skb
u32 data_segs_out read_write __tcp_transmit_skb,tcp_update_skb_after_send
u64 bytes_sent read_write __tcp_transmit_skb
u64 bytes_acked read_write tcp_snd_una_update/tcp_ack
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 3e20bffd6ae908f18c1a88ed97db78f7ea8989ae..89013be1519abd1713d42f93e1864fb1d61d303a 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -253,17 +253,14 @@ struct tcp_sock {
/* TX read-write hotpath cache lines */
__cacheline_group_begin(tcp_sock_write_tx) ____cacheline_aligned;
- u32 segs_out; /* RFC4898 tcpEStatsPerfSegsOut
- * The total number of segments sent.
- */
- u32 data_segs_out; /* RFC4898 tcpEStatsPerfDataSegsOut
- * total number of data segments sent.
- */
u32 delivered; /* Total data packets delivered incl. rexmits */
u32 delivered_ce; /* Like the above but only ECE marked packets */
u64 bytes_sent; /* RFC4898 tcpEStatsPerfHCDataOctetsOut
* total number of data bytes sent.
*/
+ u32 data_segs_out; /* RFC4898 tcpEStatsPerfDataSegsOut
+ * total number of data segments sent.
+ */
u32 snd_sml; /* Last byte of the most recently transmitted small packet */
u8 chrono_type; /* current chronograph type */
u32 chrono_start; /* Start time in jiffies of a TCP chrono */
@@ -324,6 +321,12 @@ struct tcp_sock {
* Options received (usually on last packet, some only on SYN packets).
*/
struct tcp_options_received rx_opt;
+ u32 segs_in; /* RFC4898 tcpEStatsPerfSegsIn
+ * total number of segments in.
+ */
+ u32 segs_out; /* RFC4898 tcpEStatsPerfSegsOut
+ * The total number of segments sent.
+ */
__cacheline_group_end(tcp_sock_write_txrx);
/* RX read-write hotpath cache lines */
@@ -333,9 +336,6 @@ struct tcp_sock {
* sum(delta(rcv_nxt)), or how many bytes
* were acked.
*/
- u32 segs_in; /* RFC4898 tcpEStatsPerfSegsIn
- * total number of segments in.
- */
u32 data_segs_in; /* RFC4898 tcpEStatsPerfDataSegsIn
* total number of data segments in.
*/
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index fe1e54321e181c307227c8f2f5a9c0fb2816c762..135a5db6c3fcc1e33227403fd9376642c9826412 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -5230,7 +5230,6 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, snd_ssthresh);
/* TX read-write hotpath cache lines */
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, data_segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_ce);
@@ -5267,10 +5266,11 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rcv_mwnd_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rcv_tstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rx_opt);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, segs_in);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, segs_out);
/* RX read-write hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_received);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, segs_in);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, data_segs_in);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_wup);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, max_packets_out);
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH net-next 3/5] tcp: move tp->first_tx_mstamp and tp->delivered_mstamp to tcp_sock_write_tx
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 1/5] tcp: move tp->delivered and tp->delivered_ce to tcp_sock_write_tx group Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 2/5] tcp: move tp->segs_in and tp->segs_out to tcp_sock_write_txrx group Eric Dumazet
@ 2026-04-30 10:00 ` Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 4/5] tcp: move tp->bytes_acked to tcp_sock_write_tx group Eric Dumazet
` (2 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Eric Dumazet @ 2026-04-30 10:00 UTC (permalink / raw)
To: David S . Miller, Jakub Kicinski, Paolo Abeni
Cc: Simon Horman, Neal Cardwell, Kuniyuki Iwashima, netdev,
eric.dumazet, Eric Dumazet
These fields are touched in when payload is sent.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
include/linux/tcp.h | 4 ++--
net/ipv4/tcp.c | 4 ++--
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 89013be1519abd1713d42f93e1864fb1d61d303a..e9adc88b73b43c39c734361da149ad29ef48e922 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -258,6 +258,8 @@ struct tcp_sock {
u64 bytes_sent; /* RFC4898 tcpEStatsPerfHCDataOctetsOut
* total number of data bytes sent.
*/
+ u64 first_tx_mstamp; /* start of window send phase */
+ u64 delivered_mstamp; /* time we reached "delivered" */
u32 data_segs_out; /* RFC4898 tcpEStatsPerfDataSegsOut
* total number of data segments sent.
*/
@@ -347,8 +349,6 @@ struct tcp_sock {
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
* sum(delta(snd_una)), or how many bytes
* were acked.
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 135a5db6c3fcc1e33227403fd9376642c9826412..d25698a0e6819cb4d3563461bb1ae89d5104669f 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -5234,6 +5234,8 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_ce);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, bytes_sent);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, first_tx_mstamp);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_mstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, snd_sml);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, chrono_start);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, chrono_stat);
@@ -5280,8 +5282,6 @@ static void __init tcp_struct_check(void)
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);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_rtt_est);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcvq_space);
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH net-next 4/5] tcp: move tp->bytes_acked to tcp_sock_write_tx group
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
` (2 preceding siblings ...)
2026-04-30 10:00 ` [PATCH net-next 3/5] tcp: move tp->first_tx_mstamp and tp->delivered_mstamp to tcp_sock_write_tx Eric Dumazet
@ 2026-04-30 10:00 ` Eric Dumazet
2026-04-30 10:00 ` [PATCH net-next 5/5] tcp: move max_packets_out, cwnd_usage_seq, rate_delivered and rate_interval_us " Eric Dumazet
2026-05-02 1:00 ` [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Eric Dumazet @ 2026-04-30 10:00 UTC (permalink / raw)
To: David S . Miller, Jakub Kicinski, Paolo Abeni
Cc: Simon Horman, Neal Cardwell, Kuniyuki Iwashima, netdev,
eric.dumazet, Eric Dumazet
tp->bytes_acked is touched in TX path only.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
Documentation/networking/net_cachelines/tcp_sock.rst | 2 +-
include/linux/tcp.h | 8 ++++----
net/ipv4/tcp.c | 2 +-
3 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/Documentation/networking/net_cachelines/tcp_sock.rst b/Documentation/networking/net_cachelines/tcp_sock.rst
index b914382cb8fd7ae739297180ff5101acb8dc8a14..0f6088c4ab8bb872e7fc86f02479592e84c0247a 100644
--- a/Documentation/networking/net_cachelines/tcp_sock.rst
+++ b/Documentation/networking/net_cachelines/tcp_sock.rst
@@ -22,7 +22,7 @@ u32 snd_nxt read_write read_m
u32 segs_out read_write read_write __tcp_transmit_skb
u32 data_segs_out read_write __tcp_transmit_skb,tcp_update_skb_after_send
u64 bytes_sent read_write __tcp_transmit_skb
-u64 bytes_acked read_write tcp_snd_una_update/tcp_ack
+u64 bytes_acked read_write tcp_snd_una_update/tcp_ack
u32 dsack_dups
u32 snd_una read_mostly read_write tcp_wnd_end,tcp_urg_mode,tcp_minshall_check,tcp_cwnd_validate(tx);tcp_ack,tcp_may_update_window,tcp_clean_rtx_queue(write),tcp_ack_tstamp(rx)
u32 snd_sml read_write tcp_minshall_check,tcp_minshall_update
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index e9adc88b73b43c39c734361da149ad29ef48e922..d3650f04d94251ebef6f1bae296adc33a5491768 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -255,6 +255,10 @@ struct tcp_sock {
__cacheline_group_begin(tcp_sock_write_tx) ____cacheline_aligned;
u32 delivered; /* Total data packets delivered incl. rexmits */
u32 delivered_ce; /* Like the above but only ECE marked packets */
+ u64 bytes_acked; /* RFC4898 tcpEStatsAppHCThruOctetsAcked
+ * sum(delta(snd_una)), or how many bytes
+ * were acked.
+ */
u64 bytes_sent; /* RFC4898 tcpEStatsPerfHCDataOctetsOut
* total number of data bytes sent.
*/
@@ -349,10 +353,6 @@ struct tcp_sock {
u32 rcv_rtt_last_tsecr;
u32 delivered_ecn_bytes[3];
u16 pkts_acked_ewma;/* Pkts acked EWMA for AccECN cep heuristic */
- u64 bytes_acked; /* RFC4898 tcpEStatsAppHCThruOctetsAcked
- * sum(delta(snd_una)), or how many bytes
- * were acked.
- */
struct {
u32 rtt_us;
u32 seq;
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index d25698a0e6819cb4d3563461bb1ae89d5104669f..7211d2c669b8e4ac15b3a6fcfc572f7199e310cf 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -5233,6 +5233,7 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, data_segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_ce);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, bytes_acked);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, bytes_sent);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, first_tx_mstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, delivered_mstamp);
@@ -5282,7 +5283,6 @@ static void __init tcp_struct_check(void)
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, bytes_acked);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_rtt_est);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcvq_space);
}
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH net-next 5/5] tcp: move max_packets_out, cwnd_usage_seq, rate_delivered and rate_interval_us to tcp_sock_write_tx group
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
` (3 preceding siblings ...)
2026-04-30 10:00 ` [PATCH net-next 4/5] tcp: move tp->bytes_acked to tcp_sock_write_tx group Eric Dumazet
@ 2026-04-30 10:00 ` Eric Dumazet
2026-05-02 1:00 ` [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Eric Dumazet @ 2026-04-30 10:00 UTC (permalink / raw)
To: David S . Miller, Jakub Kicinski, Paolo Abeni
Cc: Simon Horman, Neal Cardwell, Kuniyuki Iwashima, netdev,
eric.dumazet, Eric Dumazet
These fields are used in TX path.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
include/linux/tcp.h | 8 ++++----
net/ipv4/tcp.c | 8 ++++----
2 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index d3650f04d94251ebef6f1bae296adc33a5491768..8a6807082672a48a62292fc9ba9776297c7cc670 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -276,6 +276,10 @@ struct tcp_sock {
u32 lsndtime; /* timestamp of last sent data packet (for restart window) */
u32 mdev_us; /* medium deviation */
u32 rtt_seq; /* sequence number to update rttvar */
+ u32 max_packets_out; /* max packets_out in last window */
+ u32 cwnd_usage_seq; /* right edge of cwnd usage tracking flight */
+ u32 rate_delivered; /* saved rate sample: packets delivered */
+ u32 rate_interval_us; /* saved rate sample: time elapsed */
u64 tcp_wstamp_ns; /* departure time for next sent data packet */
u64 accecn_opt_tstamp; /* Last AccECN option sent timestamp */
struct list_head tsorted_sent_queue; /* time-sorted sent but un-SACKed skbs */
@@ -346,10 +350,6 @@ struct tcp_sock {
* total number of data segments in.
*/
u32 rcv_wup; /* rcv_nxt on last window update sent */
- u32 max_packets_out; /* max packets_out in last window */
- u32 cwnd_usage_seq; /* right edge of cwnd usage tracking flight */
- u32 rate_delivered; /* saved rate sample: packets delivered */
- 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 */
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 7211d2c669b8e4ac15b3a6fcfc572f7199e310cf..21ece4c7161216093e15b295b0a79db23ad406ce 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -5247,6 +5247,10 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tcp_wstamp_ns);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, accecn_opt_tstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, rtt_seq);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, max_packets_out);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, cwnd_usage_seq);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, rate_delivered);
+ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, rate_interval_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tsorted_sent_queue);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, highest_sack);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, ecn_flags);
@@ -5276,10 +5280,6 @@ static void __init tcp_struct_check(void)
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_received);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, data_segs_in);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_wup);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, max_packets_out);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, cwnd_usage_seq);
- CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rate_delivered);
- 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);
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related [flat|nested] 7+ messages in thread* Re: [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups
2026-04-30 10:00 [PATCH net-next 0/5] tcp: move some fastpath fields to appropriate groups Eric Dumazet
` (4 preceding siblings ...)
2026-04-30 10:00 ` [PATCH net-next 5/5] tcp: move max_packets_out, cwnd_usage_seq, rate_delivered and rate_interval_us " Eric Dumazet
@ 2026-05-02 1:00 ` patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: patchwork-bot+netdevbpf @ 2026-05-02 1:00 UTC (permalink / raw)
To: Eric Dumazet
Cc: davem, kuba, pabeni, horms, ncardwell, kuniyu, netdev,
eric.dumazet
Hello:
This series was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:
On Thu, 30 Apr 2026 10:00:16 +0000 you wrote:
> Move following fields to better groups to increase data locality.
>
> - delivered
> - delivered_ce
> - segs_in
> - segs_out
> - first_tx_mstamp
> - delivered_mstamp
> - max_packets_out
> - cwnd_usage_seq
> - rate_delivered
> - rate_interval_us
>
> [...]
Here is the summary with links:
- [net-next,1/5] tcp: move tp->delivered and tp->delivered_ce to tcp_sock_write_tx group
https://git.kernel.org/netdev/net-next/c/9f810527a343
- [net-next,2/5] tcp: move tp->segs_in and tp->segs_out to tcp_sock_write_txrx group
https://git.kernel.org/netdev/net-next/c/2b28dd212a8a
- [net-next,3/5] tcp: move tp->first_tx_mstamp and tp->delivered_mstamp to tcp_sock_write_tx
https://git.kernel.org/netdev/net-next/c/07db42c4b3eb
- [net-next,4/5] tcp: move tp->bytes_acked to tcp_sock_write_tx group
https://git.kernel.org/netdev/net-next/c/dd033ec406b4
- [net-next,5/5] tcp: move max_packets_out, cwnd_usage_seq, rate_delivered and rate_interval_us to tcp_sock_write_tx group
https://git.kernel.org/netdev/net-next/c/affe6c27651a
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 7+ messages in thread