netdev - Re: iproute2: tc deletion freezes whole server

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: <CAM_iQpWLK8ZKShdsWNQrbhFa2B9V8e+OSNRQ_06zyNmDToq5ew@mail.gmail.com>
Date:   Thu, 26 Mar 2020 10:38:45 -0700
From:   Cong Wang <xiyou.wangcong@...il.com>
To:     Václav Zindulka <vaclav.zindulka@...pnet.cz>
Cc:     Linux Kernel Network Developers <netdev@...r.kernel.org>
Subject: Re: iproute2: tc deletion freezes whole server

On Thu, Mar 26, 2020 at 10:07 AM Cong Wang <xiyou.wangcong@...il.com> wrote:
>
> On Thu, Mar 26, 2020 at 7:24 AM Václav Zindulka
> <vaclav.zindulka@...pnet.cz> wrote:
> >
> > > On Wed, Mar 25, 2020 at 6:43 PM Cong Wang <xiyou.wangcong@...il.com> wrote:
> > > > Are you able to test an experimental patch attached in this email?
> > >
> > > Sure. I'll compile new kernel tomorrow. Thank you for quick patch.
> > > I'll let you know as soon as I have anything.
> >
> > I've compiled kernel with the patch you provided and tested it.
> > However the problem is not solved. It behaves exactly the same way.
> > I'm trying to put some printk into the fq_codel_reset() to test it
> > more.
>
> Are the stack traces captured by perf any different with unpatched?

Wait, it seems my assumption of refcnt==0 is wrong even for deletion,
in qdisc_graft() the last refcnt is put after dev_deactivate()...

Let me think about how we could distinguish this case from other
reset cases.

Thanks.