tcp: remove tcp_rate.c
Subsystems (3)
cover tcp: remove tcp_rate.c
Move tcp_rate_gen() to tcp_input.c and tcp_rate_check_app_limited() to tcp.c for better code generation.
tcp_rate.c was interesting from code maintenance perspective but was adding cpu costs.
Eric Dumazet (2):
tcp: move tcp_rate_gen to tcp_input.c
tcp: move tcp_rate_check_app_limited() to tcp.c
include/net/tcp.h | 5 +- net/ipv4/Makefile | 2 +- net/ipv4/tcp.c | 18 ++++++ net/ipv4/tcp_input.c | 110 ++++++++++++++++++++++++++++++++++++ net/ipv4/tcp_rate.c | 130 ------------------------------------------- 5 files changed, 130 insertions(+), 135 deletions(-) delete mode 100644 net/ipv4/tcp_rate.c
base-commit: d8f87aa5fa0a4276491fa8ef436cd22605a3f9ba
Signature · 1 line
2.52.0.457.g6b5491de43-goog
PATCH 1/2 tcp: move tcp_rate_gen to tcp_input.c
This function is called from one caller only, in TCP fast path.
Move it to tcp_input.c so that compiler can inline it.
$ scripts/bloat-o-meter -t vmlinux.old vmlinux.new add/remove: 0/2 grow/shrink: 1/0 up/down: 226/-300 (-74) Function old new delta tcp_ack 5405 5631 +226 __pfx_tcp_rate_gen 16 - -16 tcp_rate_gen 284 - -284 Total: Before=22566536, After=22566462, chg -0.00%
--- include/net/tcp.h | 2 - net/ipv4/tcp_input.c | 110 +++++++++++++++++++++++++++++++++++++++++++ net/ipv4/tcp_rate.c | 110 ------------------------------------------- 3 files changed, 110 insertions(+), 112 deletions(-)
include/net/tcp.h +0 -2
@@ -1356,8 +1356,6 @@ static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)void tcp_set_ca_state (struct sock *sk,const u8 ca_state);/* From tcp_rate.c */ -void tcp_rate_gen (struct sock *sk,u32 delivered,u32 lost, -bool is_sack_reneg,struct rate_sample *rs);void tcp_rate_check_app_limited (struct sock *sk);static inline bool tcp_skb_sent_after (u64 t1,u64 t2,u32 seq1,u32 seq2)
net/ipv4/tcp_input.c +110 -0
116 lines not on this page · raw message
net/ipv4/tcp_rate.c +0 -110
116 lines not on this page · raw message
Signature · 1 line
2.52.0.457.g6b5491de43-goog
Re: [PATCH net-next 1/2] tcp: move tcp_rate_gen to tcp_input.c
Quoted 13 lines · Eric Dumazet
This function is called from one caller only, in TCP fast path.
Move it to
tcp_input.cso that compiler can inline it.$ scripts/bloat-o-meter -t vmlinux.old vmlinux.new add/remove: 0/2 grow/shrink: 1/0 up/down: 226/-300 (-74) Function old new delta tcp_ack 5405 5631 +226 __pfx_tcp_rate_gen 16 - -16 tcp_rate_gen 284 - -284 Total: Before=22566536, After=22566462, chg -0.00%
Quoted 2 lines
Signed-off-by Eric Dumazet <[email protected]>---
Thanks, Eric!
neal
PATCH 2/2 tcp: move tcp_rate_check_app_limited() to tcp.c
In reply to Eric Dumazettcp_rate_check_app_limited() is used from tcp_sendmsg_locked() fast path and from other callers.
Move it to tcp.c so that it can be inlined in tcp_sendmsg_locked().
Small increase of code, for better TCP performance.
$ scripts/bloat-o-meter -t vmlinux.old vmlinux.new add/remove: 0/0 grow/shrink: 1/0 up/down: 87/0 (87) Function old new delta tcp_sendmsg_locked 4217 4304 +87 Total: Before=22566462, After=22566549, chg +0.00%
--- include/net/tcp.h | 3 +-- net/ipv4/Makefile | 2 +- net/ipv4/tcp.c | 18 ++++++++++++++++++ net/ipv4/tcp_rate.c | 20 -------------------- 4 files changed, 20 insertions(+), 23 deletions(-) delete mode 100644 net/ipv4/tcp_rate.c
include/net/tcp.h +1 -2
@@ -809,6 +809,7 @@ static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize)/* tcp.c */ void tcp_get_info (struct sock *,struct tcp_info *); +void tcp_rate_check_app_limited (struct sock *sk);/* Read 'sendfile()'-style from a TCP socket */ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
@@ -1355,8 +1356,6 @@ static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)/* From tcp_cong.c */ void tcp_set_ca_state (struct sock *sk,const u8 ca_state); -/* From tcp_rate.c */ -void tcp_rate_check_app_limited (struct sock *sk);static inline bool tcp_skb_sent_after (u64 t1,u64 t2,u32 seq1,u32 seq2) {
net/ipv4/Makefile +1 -1
@@ -9,7 +9,7 @@ obj-y := route.o inetpeer.o protocol.o \ inet_timewait_sock.o inet_connection_sock.o \ tcp.o tcp_input.o tcp_output.o tcp_timer.o tcp_ipv4.o \ tcp_minisocks.o tcp_cong.o tcp_metrics.o tcp_fastopen.o \ - tcp_rate.o tcp_recovery.o tcp_ulp.o \ + tcp_recovery.o tcp_ulp.o \ tcp_offload.o tcp_plb.o datagram.o raw.o udp.o udplite.o \ udp_offload.o arp.o icmp.o devinet.o af_inet.o igmp.o \ fib_frontend.o fib_semantics.o fib_trie.o fib_notifier.o \
net/ipv4/tcp.c +18 -0
@@ -1074,6 +1074,24 @@ int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, int *copied,return err; } +/* If a gap is detected between sends, mark the socket application-limited. */ +void tcp_rate_check_app_limited (struct sock *sk) +{ +struct tcp_sock *tp =tcp_sk (sk); + +if (/* We have less than one packet to send. */ + tp->write_seq - tp->snd_nxt < tp->mss_cache && +/* Nothing in sending host's qdisc queues or NIC tx queue. */ +sk_wmem_alloc_get (sk) <SKB_TRUESIZE (1 ) && +/* We are not limited by CWND. */ +tcp_packets_in_flight (tp) <tcp_snd_cwnd (tp) && +/* All lost packets have been retransmitted. */ + tp->lost_out <= tp->retrans_out) + tp->app_limited = + (tp->delivered +tcp_packets_in_flight (tp)) ? :1 ; +} +EXPORT_SYMBOL_GPL (tcp_rate_check_app_limited); +int tcp_sendmsg_locked (struct sock *sk,struct msghdr *msg,size_t size) {struct net_devmem_dmabuf_binding *binding =NULL ;
net/ipv4/tcp_rate.c deleted +0 -20
@@ -1,20 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -#include <net/tcp.h> - -/* If a gap is detected between sends, mark the socket application-limited. */ -void tcp_rate_check_app_limited (struct sock *sk) -{ -struct tcp_sock *tp =tcp_sk (sk); - -if (/* We have less than one packet to send. */ - tp->write_seq - tp->snd_nxt < tp->mss_cache && -/* Nothing in sending host's qdisc queues or NIC tx queue. */ -sk_wmem_alloc_get (sk) <SKB_TRUESIZE (1 ) && -/* We are not limited by CWND. */ -tcp_packets_in_flight (tp) <tcp_snd_cwnd (tp) && -/* All lost packets have been retransmitted. */ - tp->lost_out <= tp->retrans_out) - tp->app_limited = - (tp->delivered +tcp_packets_in_flight (tp)) ? :1 ; -} -EXPORT_SYMBOL_GPL (tcp_rate_check_app_limited);
Signature · 1 line
2.52.0.457.g6b5491de43-goog
Re: [PATCH net-next 2/2] tcp: move tcp_rate_check_app_limited() to tcp.c
Quoted 8 lines · Eric Dumazet
tcp_rate_check_app_limited()is used fromtcp_sendmsg_locked()fast path and from other callers.Move it to
tcp.cso that it can be inlined intcp_sendmsg_locked().Small increase of code, for better TCP performance.
Quoted 8 lines
$ scripts/bloat-o-meter -t vmlinux.old vmlinux.new add/remove: 0/0 grow/shrink: 1/0 up/down: 87/0 (87) Function old new delta tcp_sendmsg_locked 4217 4304 +87 Total: Before=22566462, After=22566549, chg +0.00%Signed-off-by Eric Dumazet <[email protected]>---
Thanks, Eric!
neal
This series was applied to netdev/net-next.git (main) by Jakub Kicinski <[email protected]>:
Hello:
This series was applied to netdev/net-next.git (main) by Jakub Kicinski <[email protected]>:
Quoted 11 lines · On Wed, 21 Jan 2026 09:59:21 +0000 you wrote:
Move
tcp_rate_gen()totcp_input.candtcp_rate_check_app_limited()totcp.cfor better code generation.
tcp_rate.cwas interesting from code maintenance perspective but was adding cpu costs.Eric Dumazet (2):
tcp: movetcp_rate_gentotcp_input.c
tcp: movetcp_rate_check_app_limited()totcp.c[...]
Here is the summary with links:
- [net-next,1/2] tcp: move tcp_rate_gen to tcp_input.c
https://git.kernel.org/netdev/net-next/c/b814bdcecd79 - [net-next,2/2] tcp: move tcp_rate_check_app_limited() to tcp.c https://git.kernel.org/netdev/net-next/c/bc1f0b1c98f5
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html