From: Jakub Kicinski <kuba@kernel.org>
To: Eric Dumazet <edumazet@google.com>
Cc: "David S . Miller" <davem@davemloft.net>,
"Paolo Abeni" <pabeni@redhat.com>,
"Simon Horman" <horms@kernel.org>,
"Jamal Hadi Salim" <jhs@mojatatu.com>,
"Cong Wang" <xiyou.wangcong@gmail.com>,
"Jiri Pirko" <jiri@resnulli.us>,
"Toke Høiland-Jørgensen" <toke@redhat.com>,
"Kuniyuki Iwashima" <kuniyu@google.com>,
"Willem de Bruijn" <willemb@google.com>,
netdev@vger.kernel.org, eric.dumazet@gmail.com
Subject: Re: [PATCH net-next 00/10] net_sched: speedup qdisc dequeue
Date: Mon, 10 Nov 2025 08:44:32 -0800 [thread overview]
Message-ID: <20251110084432.7fdf647b@kernel.org> (raw)
In-Reply-To: <20251110094505.3335073-1-edumazet@google.com>
On Mon, 10 Nov 2025 09:44:55 +0000 Eric Dumazet wrote:
> Avoid up to two cache line misses in qdisc dequeue() to fetch
> skb_shinfo(skb)->gso_segs/gso_size while qdisc spinlock is held.
>
> Idea is to cache gso_segs at enqueue time before spinlock is
> acquired, in the first skb cache line, where we already
> have qdisc_skb_cb(skb)->pkt_len.
>
> This series gives a 8 % improvement in a TX intensive workload.
>
> (120 Mpps -> 130 Mpps on a Turin host, IDPF with 32 TX queues)
According to CI this breaks a bunch of tests.
https://netdev.bots.linux.dev/contest.html?branch=net-next-2025-11-10--12-00
I think they all hit:
[ 20.682474][ T231] WARNING: CPU: 3 PID: 231 at ./include/net/sch_generic.h:843 __dev_xmit_skb+0x786/0x1550
next prev parent reply other threads:[~2025-11-10 16:44 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-10 9:44 [PATCH net-next 00/10] net_sched: speedup qdisc dequeue Eric Dumazet
2025-11-10 9:44 ` [PATCH net-next 01/10] net_sched: make room for (struct qdisc_skb_cb)->pkt_segs Eric Dumazet
2025-11-10 9:44 ` [PATCH net-next 02/10] net: init shinfo->gso_segs from qdisc_pkt_len_init() Eric Dumazet
2025-11-10 9:44 ` [PATCH net-next 03/10] net_sched: initialize qdisc_skb_cb(skb)->pkt_segs in qdisc_pkt_len_init() Eric Dumazet
2025-11-10 9:44 ` [PATCH net-next 04/10] net_sched: use qdisc_skb_cb(skb)->pkt_segs in bstats_update() Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 05/10] net_sched: cake: use qdisc_pkt_segs() Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 06/10] net_sched: add Qdisc_read_mostly and Qdisc_write groups Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 07/10] net_sched: sch_fq: move qdisc_bstats_update() to fq_dequeue_skb() Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 08/10] net_sched: sch_fq: prefetch one skb ahead in dequeue() Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 09/10] net: prefech skb->priority in __dev_xmit_skb() Eric Dumazet
2025-11-10 9:45 ` [PATCH net-next 10/10] net: annotate a data-race " Eric Dumazet
2025-11-10 16:44 ` Jakub Kicinski [this message]
2025-11-10 17:15 ` [PATCH net-next 00/10] net_sched: speedup qdisc dequeue Eric Dumazet
2025-11-10 17:27 ` Jakub Kicinski
2025-11-10 18:05 ` Eric Dumazet
2025-11-10 18:22 ` Eric Dumazet
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20251110084432.7fdf647b@kernel.org \
--to=kuba@kernel.org \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=eric.dumazet@gmail.com \
--cc=horms@kernel.org \
--cc=jhs@mojatatu.com \
--cc=jiri@resnulli.us \
--cc=kuniyu@google.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=toke@redhat.com \
--cc=willemb@google.com \
--cc=xiyou.wangcong@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.