Re: [PATCH] audit: fix memleak in auditd_send_unicast_skb.

[Shu Wang <shuwang@redhat.com>]

----- Original Message -----
> From: "Paul Moore" <paul@paul-moore.com>
> To: shuwang@redhat.com
> Cc: "Eric Paris" <eparis@redhat.com>, linux-audit@redhat.com, linux-kernel@vger.kernel.org, liwang@redhat.com,
> chuhu@redhat.com
> Sent: Tuesday, July 18, 2017 8:21:05 PM
> Subject: Re: [PATCH] audit: fix memleak in auditd_send_unicast_skb.
> 
> On Tue, Jul 18, 2017 at 2:37 AM,  <shuwang@redhat.com> wrote:
> > From: Shu Wang <shuwang@redhat.com>
> >
> > Found this issue by kmemleak report, auditd_send_unicast_skb
> > did not free skb if rcu_dereference(auditd_conn) returns null.
> >
> > unreferenced object 0xffff88082568ce00 (size 256):
> > comm "auditd", pid 1119, jiffies 4294708499
> > backtrace:
> > [<ffffffff8176166a>] kmemleak_alloc+0x4a/0xa0
> > [<ffffffff8121820c>] kmem_cache_alloc_node+0xcc/0x210
> > [<ffffffff8161b99d>] __alloc_skb+0x5d/0x290
> > [<ffffffff8113c614>] audit_make_reply+0x54/0xd0
> > [<ffffffff8113dfa7>] audit_receive_msg+0x967/0xd70
> > ----------------
> > (gdb) list *audit_receive_msg+0x967
> > 0xffffffff8113dff7 is in audit_receive_msg (kernel/audit.c:1133).
> > 1132    skb = audit_make_reply(0, AUDIT_REPLACE, 0,
> >                                 0, &pvnr, sizeof(pvnr));
> > ---------------
> > [<ffffffff8113e402>] audit_receive+0x52/0xa0
> > [<ffffffff8166c561>] netlink_unicast+0x181/0x240
> > [<ffffffff8166c8e2>] netlink_sendmsg+0x2c2/0x3b0
> > [<ffffffff816112e8>] sock_sendmsg+0x38/0x50
> > [<ffffffff816117a2>] SYSC_sendto+0x102/0x190
> > [<ffffffff81612f4e>] SyS_sendto+0xe/0x10
> > [<ffffffff8176d337>] entry_SYSCALL_64_fastpath+0x1a/0xa5
> > [<ffffffffffffffff>] 0xffffffffffffffff
> >
> > Signed-off-by: Shu Wang <shuwang@redhat.com>
> > ---
> >  kernel/audit.c | 1 +
> >  1 file changed, 1 insertion(+)
> 
> Hello and thank you for the problem report, it is appreciated.  This
> was also reported by Masami Ichikawa who provided a patch with the
> correct fix (your patch does not catch any error conditions from
> netlink_unicast()).

netlink_unicast has it's error handling, and is responsible for
releasing skb. so the Masami Ichikawa's fix may cause double
free problems.



1274 int netlink_unicast(struct sock *ssk, struct sk_buff *skb,
1275 >--->---    u32 portid, int nonblock)
1276 {
......
1285 >---sk = netlink_getsockbyportid(ssk, portid);
1286 >---if (IS_ERR(sk)) {
1287 >--->---kfree_skb(skb);
1288 >--->---return PTR_ERR(sk);
1289 >---}
......
1293 >---if (sk_filter(sk, skb)) {
1294 >--->---err = skb->len;
1295 >--->---kfree_skb(skb);
1296 >--->---sock_put(sk);
1297 >--->---return err;
1298 >---}




> 
> * https://www.redhat.com/archives/linux-audit/2017-July/msg00073.html
> 
> > diff --git a/kernel/audit.c b/kernel/audit.c
> > index 833267b..6dd5569 100644
> > --- a/kernel/audit.c
> > +++ b/kernel/audit.c
> > @@ -641,6 +641,7 @@ static int auditd_send_unicast_skb(struct sk_buff *skb)
> >         ac = rcu_dereference(auditd_conn);
> >         if (!ac) {
> >                 rcu_read_unlock();
> > +               kfree_skb(skb);
> >                 rc = -ECONNREFUSED;
> >                 goto err;
> >         }
> > --
> > 2.5.0
> 
> --
> paul moore
> www.paul-moore.com
>