netdev - Re: [PATCH net-next v3 2/3] net: sched: Make tc-related drop reason more flexible for remaining qdiscs

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAM0EoMk9cA0qCGNa181QkGjRHr=4oZhvfMGEWoTRS-kHXFWt7g@mail.gmail.com>
Date: Wed, 13 Dec 2023 13:36:31 -0500
From: Jamal Hadi Salim <jhs@...atatu.com>
To: Jamal Hadi Salim <hadi@...atatu.com>
Cc: Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, 
	Victor Nogueira <victor@...atatu.com>, xiyou.wangcong@...il.com, jiri@...nulli.us, 
	davem@...emloft.net, pabeni@...hat.com, daniel@...earbox.net, 
	dcaratti@...hat.com, netdev@...r.kernel.org, kernel@...atatu.com
Subject: Re: [PATCH net-next v3 2/3] net: sched: Make tc-related drop reason
 more flexible for remaining qdiscs

On Tue, Dec 12, 2023 at 12:17 PM Jamal Hadi Salim <hadi@...atatu.com> wrote:
>
> On Tue, Dec 12, 2023 at 11:57 AM 'Eric Dumazet' via kernel issues
> <kernel@...atatu.com> wrote:
> >
> > On Tue, Dec 12, 2023 at 5:28 PM Jamal Hadi Salim <jhs@...atatu.com> wrote:
> > >
> >
> > >
> > > So when we looked at this code there was some mystery. It wasnt clear
> > > how to_free could have more than one skb.
> >
> > Some qdisc can free skbs at enqueue() time in  a batch for performance
> > reason (fq_codel_drop() is such an instance)
>
> Ok, makes more sense, should've caught that (there are others like taprio).
>
> > I agree that all skbs in this list have probably the same drop_reason ?
>
> It seems that way. We will review all the qdiscs to see if there's any
> exceptions.

Putting this to rest:
Other than fq codel, the others that deal with multiple skbs due to
gso segments. So the conclusion is:  if we have a bunch in the list
then they all suffer the same fate. So a single reason for the list is
sufficient.

cheers,
jamal