Cake - FQ_codel the next generation
 help / color / mirror / Atom feed
From: "Toke Høiland-Jørgensen" <toke@redhat.com>
To: "Jonas Köppeler" <j.koeppeler@tu-berlin.de>,
	"Jamal Hadi Salim" <jhs@mojatatu.com>,
	"Jiri Pirko" <jiri@resnulli.us>,
	"David S. Miller" <davem@davemloft.net>,
	"Eric Dumazet" <edumazet@kernel.org>,
	"Jakub Kicinski" <kuba@kernel.org>,
	"Paolo Abeni" <pabeni@redhat.com>,
	"Simon Horman" <horms@kernel.org>,
	"Willem de Bruijn" <willemb@google.com>
Cc: cake@lists.bufferbloat.net, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: [Cake] Re: [PATCH net] net/sched: sch_cake: fix unintended tin override with cake_mq
Date: Thu, 19 Nov 2026 03:49:58 -0600	[thread overview]
Message-ID: <CAPB2EApXULDiMGXMS6qg17NJRyEruybWMfu_d35tjBPdk4ftSQ@mail.gmail.com> (raw)
In-Reply-To: <20261119-cake_mq_fix_incorrect_tin_override-v1-1-3238122d09ca@tu-berlin.de>

Jonas Köppeler <j.koeppeler@tu-berlin.de> writes:

> CAKE allows overriding the tin selection through skb->priority: if the
> major number of skb->priority matches the qdisc handle, the minor number
> selects the tin. With cake_mq, the per-queue CAKE instances are created
> through qdisc_create_dflt() and never get a handle assigned, so
> sch->handle is 0 for all of them.
>
> For forwarded IPv4 traffic, ip_forward() sets skb->priority from the TOS
> byte via rt_tos2priority() when net.ipv4.ip_forward_update_priority is
> enabled, which is the default. rt_tos2priority() returns 0, 2, 4 or 6,
> all with a major number of 0, which matches the handle of the cake_mq
> children. The minor number is then taken as a tin override and the DSCP
> mapping is skipped.
>
> With diffserv4 and diffserv8 this puts packets into the wrong tins. For
> example, AF13 (TOS byte 0x38) under diffserv4:
>
>   ip_forward():
>     skb->priority = ip_tos2prio[(0x38 & 0x1E) >> 1]
>                   = ip_tos2prio[12] = TC_PRIO_INTERACTIVE_BULK = 4
>   cake_select_tin():
>     TC_H_MAJ(4) == 0 == sch->handle
>     tin = bulk_order[4 - 1] = 3                     -> Voice
>
> whereas the DSCP table gives diffserv4[0x38 >> 2] = diffserv4[14] = 0,
> i.e. Best Effort. Likewise, AF21, AF31 and AF41 are demoted from Video
> to Best Effort. diffserv3 happens to be unaffected, because the only
> reachable override (priority 2) agrees with its DSCP table.
>
> Fix this by matching skb->priority against the major number of the
> parent cake_mq qdisc when the instance is part of cake_mq.
>
> Fixes: ebc65a873eff ("net/sched: sch_cake: Add cake_mq qdisc for using cake on mq devices")
> Signed-off-by: Jonas Köppeler <j.koeppeler@tu-berlin.de>

Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>


      reply	other threads:[~2026-11-19  9:50 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-19  9:43 [Cake] [PATCH net] net/sched: sch_cake: fix unintended tin override with cake_mq Jonas Köppeler
2026-11-19  9:49 ` Toke Høiland-Jørgensen [this message]

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

  List information: https://lists.bufferbloat.net/postorius/lists/cake.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAPB2EApXULDiMGXMS6qg17NJRyEruybWMfu_d35tjBPdk4ftSQ@mail.gmail.com \
    --to=toke@redhat.com \
    --cc=cake@lists.bufferbloat.net \
    --cc=davem@davemloft.net \
    --cc=edumazet@kernel.org \
    --cc=horms@kernel.org \
    --cc=j.koeppeler@tu-berlin.de \
    --cc=jhs@mojatatu.com \
    --cc=jiri@resnulli.us \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=willemb@google.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox