All of lore.kernel.org
 help / color / mirror / Atom feed
From: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: =?utf-8?q?netdev=40vger=2Ekernel=2Eorg=2C?=@vger.kernel.org,
	=?utf-8?q?_linyunsheng=40huawei=2Ecom=2C?=@vger.kernel.org,
	=?utf-8?q?_David_S=2E_Miller_=3Cdavem=40davemloft=2Enet=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Eric_Dumazet_=3Cedumazet=40google=2Ecom=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Daniel_Borkmann_=3Cdaniel=40iogearbox=2Enet=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Antoine_Tenart_=3Catenart=40kernel=2Eorg=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Alexander_Lobakin_=3Calobakin=40pm=2Eme=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Wei_Wang_=3Cweiwan=40google=2Ecom=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Taehee_Yoo_=3Cap420073=40gmail=2Ecom=3E=2C?=@vger.kernel.org,
	=?utf-8?b?IEJqw7ZybiBUw7ZwZWwgPGJqb3JuQGtlcm5lbC5vcmc+LA==?=@vger.kernel.org,
	=?utf-8?q?_Arnd_Bergmann_=3Carnd=40arndb=2Ede=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Kumar_Kartikeya_Dwivedi_=3Cmemxor=40gmail=2Ecom=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Neil_Horman_=3Cnhorman=40redhat=2Ecom=3E=2C?=@vger.kernel.org,
	=?utf-8?q?_Dust_Li_=3Cdust=2Eli=40linux=2Ealibaba=2Ecom=3E?=@vger.kernel.org
Subject: Re: [PATCH net v2] napi: fix race inside napi_enable
Date: Wed, 22 Sep 2021 14:47:47 +0800	[thread overview]
Message-ID: <1632293267.9421082-1-xuanzhuo@linux.alibaba.com> (raw)
In-Reply-To: <20210920122024.283fe8b2@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>

On Mon, 20 Sep 2021 12:20:24 -0700, Jakub Kicinski <kuba@kernel.org> wrote:
> On Sat, 18 Sep 2021 16:52:32 +0800 Xuan Zhuo wrote:
> > The process will cause napi.state to contain NAPI_STATE_SCHED and
> > not in the poll_list, which will cause napi_disable() to get stuck.
> >
> > The prefix "NAPI_STATE_" is removed in the figure below, and
> > NAPI_STATE_HASHED is ignored in napi.state.
> >
> >                       CPU0       |                   CPU1       | napi.state
> > ===============================================================================
> > napi_disable()                   |                              | SCHED | NPSVC
> > napi_enable()                    |                              |
> > {                                |                              |
> >     smp_mb__before_atomic();     |                              |
> >     clear_bit(SCHED, &n->state); |                              | NPSVC
> >                                  | napi_schedule_prep()         | SCHED | NPSVC
> >                                  | napi_poll()                  |
> >                                  |   napi_complete_done()       |
> >                                  |   {                          |
> >                                  |      if (n->state & (NPSVC | | (1)
> >                                  |               _BUSY_POLL)))  |
> >                                  |           return false;      |
> >                                  |     ................         |
> >                                  |   }                          | SCHED | NPSVC
> >                                  |                              |
> >     clear_bit(NPSVC, &n->state); |                              | SCHED
> > }                                |                              |
> >                                  |                              |
> > napi_schedule_prep()             |                              | SCHED | MISSED (2)
> >
> > (1) Here return direct. Because of NAPI_STATE_NPSVC exists.
> > (2) NAPI_STATE_SCHED exists. So not add napi.poll_list to sd->poll_list
> >
> > Since NAPI_STATE_SCHED already exists and napi is not in the
> > sd->poll_list queue, NAPI_STATE_SCHED cannot be cleared and will always
> > exist.
> >
> > 1. This will cause this queue to no longer receive packets.
> > 2. If you encounter napi_disable under the protection of rtnl_lock, it
> >    will cause the entire rtnl_lock to be locked, affecting the overall
> >    system.
> >
> > This patch uses cmpxchg to implement napi_enable(), which ensures that
> > there will be no race due to the separation of clear two bits.
> >
> > Fixes: 2d8bff12699abc ("netpoll: Close race condition between poll_one_napi and napi_disable")
> > Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> > Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
>
> Why don't you just invert the order of clearing the bits:

I think it should be an atomic operation. The original two-step clear itself is
problematic. So from this perspective, it is not just a solution to this
problem.

Thanks.

>
> diff --git a/net/core/dev.c b/net/core/dev.c
> index a796754f75cc..706eca8112c1 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -6953,8 +6953,8 @@ void napi_enable(struct napi_struct *n)
>  {
>         BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
>         smp_mb__before_atomic();
> -       clear_bit(NAPI_STATE_SCHED, &n->state);
>         clear_bit(NAPI_STATE_NPSVC, &n->state);
> +       clear_bit(NAPI_STATE_SCHED, &n->state);
>         if (n->dev->threaded && n->thread)
>                 set_bit(NAPI_STATE_THREADED, &n->state);
>  }
>
> That's simpler and symmetric with the disable path.

  reply	other threads:[~2021-09-22  6:48 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-18  8:52 Xuan Zhuo
2021-09-20  8:50 ` patchwork-bot+netdevbpf
2021-09-20 19:20 ` Jakub Kicinski
2021-09-22  6:47   ` Xuan Zhuo [this message]
2021-09-23 13:14     ` Jakub Kicinski
     [not found]       ` <1632404456.506512-1-xuanzhuo@linux.alibaba.com>
2021-09-23 14:54         ` Jakub Kicinski
2021-10-18 21:58 ` Sukadev Bhattiprolu
2021-10-18 22:55   ` Jakub Kicinski
2021-10-18 23:36     ` Dany Madden
2021-10-18 23:47       ` Jakub Kicinski
2021-10-19  0:01         ` Sukadev Bhattiprolu
2021-10-22  3:16       ` Sukadev Bhattiprolu
2021-10-25 17:36         ` Jakub Kicinski

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

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

  git send-email \
    --in-reply-to=1632293267.9421082-1-xuanzhuo@linux.alibaba.com \
    --to=xuanzhuo@linux.alibaba.com \
    --cc==?utf-8?b?IEJqw7ZybiBUw7ZwZWwgPGJqb3JuQGtlcm5lbC5vcmc+LA==?=@vger.kernel.org \
    --cc==?utf-8?q?_Alexander_Lobakin_=3Calobakin=40pm=2Eme=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Antoine_Tenart_=3Catenart=40kernel=2Eorg=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Arnd_Bergmann_=3Carnd=40arndb=2Ede=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Daniel_Borkmann_=3Cdaniel=40iogearbox=2Enet=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_David_S=2E_Miller_=3Cdavem=40davemloft=2Enet=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Dust_Li_=3Cdust=2Eli=40linux=2Ealibaba=2Ecom=3E?=@vger.kernel.org \
    --cc==?utf-8?q?_Eric_Dumazet_=3Cedumazet=40google=2Ecom=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Kumar_Kartikeya_Dwivedi_=3Cmemxor=40gmail=2Ecom=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Neil_Horman_=3Cnhorman=40redhat=2Ecom=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Taehee_Yoo_=3Cap420073=40gmail=2Ecom=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_Wei_Wang_=3Cweiwan=40google=2Ecom=3E=2C?=@vger.kernel.org \
    --cc==?utf-8?q?_linyunsheng=40huawei=2Ecom=2C?=@vger.kernel.org \
    --cc==?utf-8?q?netdev=40vger=2Ekernel=2Eorg=2C?=@vger.kernel.org \
    --cc=kuba@kernel.org \
    --subject='Re: [PATCH net v2] napi: fix race inside napi_enable' \
    /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

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.