netdev - Re: [PATCH net-next 1/4] net: fix possible race in skb_attempt_defer

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]

Date:   Mon, 16 May 2022 11:24:40 -0700
From:   Eric Dumazet <edumazet@...gle.com>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     Eric Dumazet <eric.dumazet@...il.com>,
        "David S . Miller" <davem@...emloft.net>,
        Paolo Abeni <pabeni@...hat.com>,
        netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next 1/4] net: fix possible race in skb_attempt_defer_free()

On Mon, May 16, 2022 at 11:16 AM Jakub Kicinski <kuba@...nel.org> wrote:
>
> On Sun, 15 May 2022 21:24:53 -0700 Eric Dumazet wrote:
> > A cpu can observe sd->defer_count reaching 128,
> > and call smp_call_function_single_async()
> >
> > Problem is that the remote CPU can clear sd->defer_count
> > before the IPI is run/acknowledged.
> >
> > Other cpus can queue more packets and also decide
> > to call smp_call_function_single_async() while the pending
> > IPI was not yet delivered.
> >
> > This is a common issue with smp_call_function_single_async().
> > Callers must ensure correct synchronization and serialization.
> >
> > I triggered this issue while experimenting smaller threshold.
> > Performing the call to smp_call_function_single_async()
> > under sd->defer_lock protection did not solve the problem.
> >
> > Commit 5a18ceca6350 ("smp: Allow smp_call_function_single_async()
> > to insert locked csd") replaced an informative WARN_ON_ONCE()
> > with a return of -EBUSY, which is often ignored.
> > Test of CSD_FLAG_LOCK presence is racy anyway.
>
> If I'm reading this right this is useful for backports but in net-next
> it really is a noop? The -EBUSY would be perfectly safe to ignore?
> Just checking.

Not sure I understand the question.

trigger_rx_softirq() and friends were only in net-next, so there is no
backport needed.

Are you talking of calls from net_rps_send_ipi() ?
These are fine, because we own an atomic bit (NAPI_STATE_SCHED).