[PATCH net-next] tcp: remove some holes in struct tcp_sock

[PATCH net-next] tcp: remove some holes in struct tcp_sock

[Eric Dumazet]

By moving some fields around, this patch shrinks
holes size from 56 to 32, saving 24 bytes on 64bit arches.

After the patch pahole gives the following for 'struct tcp_sock':

	/* size: 2304, cachelines: 36, members: 162 */
	/* sum members: 2234, holes: 6, sum holes: 32 */
	/* sum bitfield members: 34 bits, bit holes: 5, sum bit holes: 14 bits */
	/* padding: 32 */
	/* paddings: 3, sum paddings: 10 */
	/* forced alignments: 1, forced holes: 1, sum forced holes: 12 */

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/linux/tcp.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index a1c47a6d69b0efd7e62765fbd873c848da22aaec..988a30ef6bfe956fa573f1f18c8284aa382dc1cc 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -264,10 +264,10 @@ struct tcp_sock {
 	u32	pushed_seq;	/* Last pushed seq, required to talk to windows */
 	u32	lsndtime;
 	u32	mdev_us;	/* medium deviation			*/
+	u32	rtt_seq;	/* sequence number to update rttvar	*/
 	u64	tcp_wstamp_ns;	/* departure time for next sent data packet */
 	u64	tcp_clock_cache; /* cache last tcp_clock_ns() (see tcp_mstamp_refresh()) */
 	u64	tcp_mstamp;	/* most recent packet received/sent */
-	u32	rtt_seq;	/* sequence number to update rttvar	*/
 	struct list_head tsorted_sent_queue; /* time-sorted sent but un-SACKed skbs */
 	struct sk_buff *highest_sack;   /* skb just after the highest
 					 * skb with SACKed bit set
@@ -350,7 +350,6 @@ struct tcp_sock {
 	u32	dsack_dups;	/* RFC4898 tcpEStatsStackDSACKDups
 				 * total number of DSACK blocks received
 				 */
-	u32	last_oow_ack_time;  /* timestamp of last out-of-window ACK */
 	u32	compressed_ack_rcv_nxt;
 	struct list_head tsq_node; /* anchor in tsq_tasklet.head list */
 
@@ -384,12 +383,12 @@ struct tcp_sock {
 		syn_fastopen_ch:1, /* Active TFO re-enabling probe */
 		syn_data_acked:1;/* data in SYN is acked by SYN-ACK */
 
+	u8	keepalive_probes; /* num of allowed keep alive probes	*/
 	u32	tcp_tx_delay;	/* delay (in usec) added to TX packets */
 
 /* RTT measurement */
 	u32	mdev_max_us;	/* maximal mdev for the last rtt period	*/
 
-	u8	keepalive_probes; /* num of allowed keep alive probes	*/
 	u32	reord_seen;	/* number of data packet reordering events */
 
 /*
@@ -402,6 +401,7 @@ struct tcp_sock {
 	u32	prior_cwnd;	/* cwnd right before starting loss recovery */
 	u32	prr_delivered;	/* Number of newly delivered packets to
 				 * receiver in Recovery. */
+	u32	last_oow_ack_time;  /* timestamp of last out-of-window ACK */
 
 	struct hrtimer	pacing_timer;
 	struct hrtimer	compressed_ack_timer;
@@ -477,8 +477,8 @@ struct tcp_sock {
 	bool	is_mptcp;
 #endif
 #if IS_ENABLED(CONFIG_SMC)
-	bool	(*smc_hs_congested)(const struct sock *sk);
 	bool	syn_smc;	/* SYN includes SMC */
+	bool	(*smc_hs_congested)(const struct sock *sk);
 #endif
 
 #if defined(CONFIG_TCP_MD5SIG) || defined(CONFIG_TCP_AO)
-- 
2.44.0.rc1.240.g4c46232300-goog

Re: [PATCH net-next] tcp: remove some holes in struct tcp_sock

[Jiri Pirko]

Tue, Feb 27, 2024 at 08:27:21PM CET, edumazet@google.com wrote:
>By moving some fields around, this patch shrinks
>holes size from 56 to 32, saving 24 bytes on 64bit arches.
>
>After the patch pahole gives the following for 'struct tcp_sock':
>
>	/* size: 2304, cachelines: 36, members: 162 */
>	/* sum members: 2234, holes: 6, sum holes: 32 */
>	/* sum bitfield members: 34 bits, bit holes: 5, sum bit holes: 14 bits */
>	/* padding: 32 */
>	/* paddings: 3, sum paddings: 10 */
>	/* forced alignments: 1, forced holes: 1, sum forced holes: 12 */
>
>Signed-off-by: Eric Dumazet <edumazet@google.com>

Reviewed-by: Jiri Pirko <jiri@nvidia.com>

Re: [PATCH net-next] tcp: remove some holes in struct tcp_sock

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Tue, 27 Feb 2024 19:27:21 +0000 you wrote:
> By moving some fields around, this patch shrinks
> holes size from 56 to 32, saving 24 bytes on 64bit arches.
> 
> After the patch pahole gives the following for 'struct tcp_sock':
> 
> 	/* size: 2304, cachelines: 36, members: 162 */
> 	/* sum members: 2234, holes: 6, sum holes: 32 */
> 	/* sum bitfield members: 34 bits, bit holes: 5, sum bit holes: 14 bits */
> 	/* padding: 32 */
> 	/* paddings: 3, sum paddings: 10 */
> 	/* forced alignments: 1, forced holes: 1, sum forced holes: 12 */
> 
> [...]

Here is the summary with links:
  - [net-next] tcp: remove some holes in struct tcp_sock
    https://git.kernel.org/netdev/net-next/c/99123622050f

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html