netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Willy Tarreau <w@1wt.eu>
To: yangzhuorao <alex000young@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
	xiyou.wangcong@gmail.com, jiri@resnulli.us, davem@davemloft.net,
	xkaneiki@gmail.com, hackerzheng666@gmail.com
Subject: Re: [PATCH] net: sched: use-after-free in tcf_action_destroy
Date: Fri, 16 Aug 2024 07:03:20 +0200	[thread overview]
Message-ID: <20240816050320.GA15186@1wt.eu> (raw)
In-Reply-To: <CAM0EoMmAcgbQWG7kQoe335079Y2UY_BmoYErL=44-itJ=p-B-Q@mail.gmail.com>

On Fri, Aug 16, 2024 at 12:06:25AM -0400, Jamal Hadi Salim wrote:
> On Thu, Aug 15, 2024 at 9:54 PM yangzhuorao <alex000young@gmail.com> wrote:
> >
> > There is a uaf bug in net/sched/act_api.c.
> > When Thread1 call [1] tcf_action_init_1 to alloc act which saves
> > in actions array. If allocation failed, it will go to err path.
> > Meanwhile thread2 call tcf_del_walker to delete action in idr.
> > So thread 1 in err path [3] tcf_action_destroy will cause
> > use-after-free read bug.
> >
> > Thread1                            Thread2
> >  tcf_action_init
> >   for(i;i<TCA_ACT_MAX_PRIO;i++)
> >    act=tcf_action_init_1 //[1]
> >    if(IS_ERR(act))
> >     goto err
> >    actions[i] = act
> >                                    tcf_del_walker
> >                                     idr_for_each_entry_ul(idr,p,id)
> >                                      __tcf_idr_release(p,false,true)
> >                                       free_tcf(p) //[2]
> >   err:
> >    tcf_action_destroy
> >     a=actions[i]
> >     ops = a->ops //[3]
> >
> > We add lock and unlock in tcf_action_init and tcf_del_walker function
> > to aviod race condition.
> >
> > ==================================================================
> > BUG: KASAN: use-after-free in tcf_action_destroy+0x138/0x150
> > Read of size 8 at addr ffff88806543e100 by task syz-executor156/295
> >
> 
> Since this is syzkaller induced, do you have a repro?
> Also what kernel (trying to see if it was before/after Eric's netlink changes).
> 
> cheers,
> jamal

In addition, please note that since this is public, there's no point in
CCing security@k.o (which I've dropped from the CC list) since there's
no coordination needed anymore.

Thanks!
Willy

  reply	other threads:[~2024-08-16  5:03 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-16  1:53 [PATCH] net: sched: use-after-free in tcf_action_destroy yangzhuorao
2024-08-16  4:06 ` Jamal Hadi Salim
2024-08-16  5:03   ` Willy Tarreau [this message]
2024-08-16 15:04   ` Jamal Hadi Salim
2024-08-17  9:27     ` Alex Young
2024-08-17  9:35       ` Greg KH
2024-08-17 12:11         ` Jamal Hadi Salim
2024-08-18 10:40           ` Greg KH
2024-08-19  1:10             ` Alex Young
2024-08-19  3:08               ` Greg KH
2024-08-16  7:20 ` Jiri Pirko

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=20240816050320.GA15186@1wt.eu \
    --to=w@1wt.eu \
    --cc=alex000young@gmail.com \
    --cc=davem@davemloft.net \
    --cc=hackerzheng666@gmail.com \
    --cc=jhs@mojatatu.com \
    --cc=jiri@resnulli.us \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=xiyou.wangcong@gmail.com \
    --cc=xkaneiki@gmail.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).