[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAM0EoM=Cen-0ctMkBvDL-jsuwPKGetz4yTG+RpmW7dXjjeVaQg@mail.gmail.com>
Date: Thu, 18 Apr 2024 19:05:08 -0400
From: Jamal Hadi Salim <jhs@...atatu.com>
To: Simon Horman <horms@...nel.org>
Cc: Eric Dumazet <edumazet@...gle.com>, "David S . Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
Toke Høiland-Jørgensen <toke@...hat.com>,
Cong Wang <xiyou.wangcong@...il.com>, Jiri Pirko <jiri@...nulli.us>, netdev@...r.kernel.org,
eric.dumazet@...il.com
Subject: Re: tdc [Was: Re: [PATCH v2 net-next 00/14] net_sched: first series
for RTNL-less] qdisc dumps
On Thu, Apr 18, 2024 at 11:08 AM Simon Horman <horms@...nel.org> wrote:
>
> On Thu, Apr 18, 2024 at 06:23:27AM -0400, Jamal Hadi Salim wrote:
> > On Thu, Apr 18, 2024 at 3:32 AM Eric Dumazet <edumazet@...gle.com> wrote:
> > >
> > > Medium term goal is to implement "tc qdisc show" without needing
> > > to acquire RTNL.
> > >
> > > This first series makes the requested changes in 14 qdisc.
> > >
> > > Notes :
> > >
> > > - RTNL is still held in "tc qdisc show", more changes are needed.
> > >
> > > - Qdisc returning many attributes might want/need to provide
> > > a consistent set of attributes. If that is the case, their
> > > dump() method could acquire the qdisc spinlock, to pair the
> > > spinlock acquision in their change() method.
> > >
> >
> > For the series:
> > Reviewed-by: Jamal Hadi Salim<jhs@...atatu.com>
> >
> > Not a show-stopper, we'll run the tdc tests after (and use this as an
> > opportunity to add more tests if needed).
> > For your next series we'll try to do that after you post.
>
> Hi Jamal,
>
> On the topic of tdc, I noticed the following both
> with and without this series applied. Is this something
> you are aware of?
>
> not ok 990 ce7d - Add mq Qdisc to multi-queue device (4 queues)
>
Since you said it also happens before Eric's patch, I took a look in
the test and nothing seems to stand out. Which iproute2 version are
you using?
We are running tdc in tandem with net-next (and iproute2-next) via
nipa for a while now and didn't see this problem pop up. So I am
guessing something in your setup?
> I'm not sure if it is valid, but I tried running tdc like this:
>
> $ ng --build --config tools/testing/selftests/tc-testing/config
> $ vng -v --run . --user root --cpus 4 -- \
> "cd ./tools/testing/selftests/tc-testing; ./tdc.py;"
This looks reasonable...
cheers,
jamal
Powered by blists - more mailing lists