Re: [PATCH] net: sctp: fix memory leak in sctp_send_reset_streams

[Marcelo Ricardo Leitner]

On Wed, Jul 24, 2019 at 03:56:40PM +0800, Xin Long wrote:
> On Sun, Jun 2, 2019 at 9:36 PM Xin Long <lucien.xin@xxxxxxxxx> wrote:
> >
> > On Sun, Jun 2, 2019 at 6:52 PM Neil Horman <nhorman@xxxxxxxxxxxxx> wrote:
> > >
> > > On Sun, Jun 02, 2019 at 11:44:29AM +0800, Hillf Danton wrote:
> > > >
> > > > syzbot found the following crash on:
> > > >
> > > > HEAD commit:    036e3431 Merge git://git.kernel.org/pub/scm/linux/kernel/g..
> > > > git tree:       upstream
> > > > console output: https://syzkaller.appspot.com/x/log.txt?x=153cff12a00000
> > > > kernel config:  https://syzkaller.appspot.com/x/.config?x=8f0f63a62bb5b13c
> > > > dashboard link: https://syzkaller.appspot.com/bug?extid=6ad9c3bd0a218a2ab41d
> > > > compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
> > > > syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=12561c86a00000
> > > > C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=15b76fd8a00000
> > > >
> > > > executing program
> > > > executing program
> > > > executing program
> > > > executing program
> > > > executing program
> > > > BUG: memory leak
> > > > unreferenced object 0xffff888123894820 (size 32):
> > > >   comm "syz-executor045", pid 7267, jiffies 4294943559 (age 13.660s)
> > > >   hex dump (first 32 bytes):
> > > >     00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
> > > >     00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
> > > >   backtrace:
> > > >     [<00000000c7e71c69>] kmemleak_alloc_recursive
> > > > include/linux/kmemleak.h:55 [inline]
> > > >     [<00000000c7e71c69>] slab_post_alloc_hook mm/slab.h:439 [inline]
> > > >     [<00000000c7e71c69>] slab_alloc mm/slab.c:3326 [inline]
> > > >     [<00000000c7e71c69>] __do_kmalloc mm/slab.c:3658 [inline]
> > > >     [<00000000c7e71c69>] __kmalloc+0x161/0x2c0 mm/slab.c:3669
> > > >     [<000000003250ed8e>] kmalloc_array include/linux/slab.h:670 [inline]
> > > >     [<000000003250ed8e>] kcalloc include/linux/slab.h:681 [inline]
> > > >     [<000000003250ed8e>] sctp_send_reset_streams+0x1ab/0x5a0 net/sctp/stream.c:302
> > > >     [<00000000cd899c6e>] sctp_setsockopt_reset_streams net/sctp/socket.c:4314 [inline]
> > > >     [<00000000cd899c6e>] sctp_setsockopt net/sctp/socket.c:4765 [inline]
> > > >     [<00000000cd899c6e>] sctp_setsockopt+0xc23/0x2bf0 net/sctp/socket.c:4608
> > > >     [<00000000ff3a21a2>] sock_common_setsockopt+0x38/0x50 net/core/sock.c:3130
> > > >     [<000000009eb87ae7>] __sys_setsockopt+0x98/0x120 net/socket.c:2078
> > > >     [<00000000e0ede6ca>] __do_sys_setsockopt net/socket.c:2089 [inline]
> > > >     [<00000000e0ede6ca>] __se_sys_setsockopt net/socket.c:2086 [inline]
> > > >     [<00000000e0ede6ca>] __x64_sys_setsockopt+0x26/0x30 net/socket.c:2086
> > > >     [<00000000c61155f5>] do_syscall_64+0x76/0x1a0 arch/x86/entry/common.c:301
> > > >     [<00000000e540958c>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
> > > >
> > > >
> > > > It was introduced in commit d570a59c5b5f ("sctp: only allow the out stream
> > > > reset when the stream outq is empty"), in orde to check stream outqs before
> > > > sending SCTP_STRRESET_IN_PROGRESS back to the peer of the stream. EAGAIN is
> > > > returned, however, without the nstr_list slab released, if any outq is found
> > > > to be non empty.
> > > >
> > > > Freeing the slab in question before bailing out fixes it.
> > > >
> > > > Fixes: d570a59c5b5f ("sctp: only allow the out stream reset when the stream outq is empty")
> > > > Reported-by: syzbot <syzbot+6ad9c3bd0a218a2ab41d@xxxxxxxxxxxxxxxxxxxxxxxxx>
> > > > Reported-by: Marcelo Ricardo Leitner <marcelo.leitner@xxxxxxxxx>
> > > > Tested-by: Marcelo Ricardo Leitner <marcelo.leitner@xxxxxxxxx>
> > > > Cc: Xin Long <lucien.xin@xxxxxxxxx>
> > > > Cc: Neil Horman <nhorman@xxxxxxxxxxxxx>
> > > > Cc: Vlad Yasevich <vyasevich@xxxxxxxxx>
> > > > Cc: Eric Dumazet <edumazet@xxxxxxxxxx>
> > > > Signed-off-by: Hillf Danton <hdanton@xxxxxxxx>
> > > > ---
> > > > net/sctp/stream.c | 1 +
> > > > 1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/net/sctp/stream.c b/net/sctp/stream.c
> > > > index 93ed078..d3e2f03 100644
> > > > --- a/net/sctp/stream.c
> > > > +++ b/net/sctp/stream.c
> > > > @@ -310,6 +310,7 @@ int sctp_send_reset_streams(struct sctp_association *asoc,
> > > >
> > > >       if (out && !sctp_stream_outq_is_empty(stream, str_nums, nstr_list)) {
> > > >               retval = -EAGAIN;
> > > > +             kfree(nstr_list);
> > > >               goto out;
> > > >       }
> > > >
> > > > --
> > > >
> > > >
> > > Acked-by: Neil Horman <nhorman@xxxxxxxxxxxxx>
> > Reviewed-by: Xin Long <lucien.xin@xxxxxxxxx>
> This fix is not applied, pls resend it with:
> to = network dev <netdev@xxxxxxxxxxxxxxx>
> cc = davem@xxxxxxxxxxxxx
> to = linux-sctp@xxxxxxxxxxxxxxx
> cc = Neil Horman <nhorman@xxxxxxxxxxxxx>
> cc = Marcelo Ricardo Leitner <marcelo.leitner@xxxxxxxxx>

Good catch, thanks Xin. I don't know what happened but I never got
this patch via netdev@, just the direct delivery. If it didn't reach
netdev@, that explains it.

  Marcelo