Re: [RFCv4 02/16] Bluetooth: Revert to mutexes from RCU list

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Andrei,

On Mon, Feb 13, 2012 at 11:49 AM, Emeltchenko Andrei
<Andrei.Emeltchenko.news@xxxxxxxxx> wrote:
>
> Hi Ulisses,
>
> On Mon, Feb 13, 2012 at 10:58:43AM +0200, Emeltchenko Andrei wrote:
> > Hi Ulisses,
> >
> > On Fri, Feb 10, 2012 at 04:24:57PM -0200, Ulisses Furquim wrote:
> > > Hi Andrei,
> > >
> > > On Fri, Feb 10, 2012 at 11:54 AM, Emeltchenko Andrei
> > > <Andrei.Emeltchenko.news@xxxxxxxxx> wrote:
> > > > From: Andrei Emeltchenko <andrei.emeltchenko@xxxxxxxxx>
> > > >
> > > > Usage of RCU list looks not reasonalbe for a number of reasons:
> > > > our code sleep and we have to use socket spinlocks, some parts
> > > > of code are updaters thus we need to use mutexes anyway.
> > > >
> > > > Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@xxxxxxxxx>
> > >
> > > No need to lock and unlock conn->chan_lock in l2cap_disconnect_rsp()?
> > > This change from RCU to mutexes really should be just one commit IMO.
> >
> > I try to add chunks which are not in different patches but then this patch
> > would several hundreds lines long. If this OK I just merge them.
>
> I've picked chunks which might come after this patch that are dealing with
> locking conn->chan_lock. Please check it below. Do you think it needs to
> be merged with "RCU remove" patch?
>
> diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c
> index c0a35c5..356ce6e 100644
> --- a/net/bluetooth/l2cap_core.c
> +++ b/net/bluetooth/l2cap_core.c
> @@ -240,11 +240,13 @@ static void l2cap_chan_timeout(struct work_struct
> *work)
>  {
>        struct l2cap_chan *chan = container_of(work, struct l2cap_chan,
>                                                        chan_timer.work);
> +       struct l2cap_conn *conn = chan->conn;
>        struct sock *sk = chan->sk;
>        int reason;
>
>        BT_DBG("chan %p state %d", chan, chan->state);
>
> +       mutex_lock(&conn->chan_lock);
>        lock_sock(sk);
>
>        if (chan->state == BT_CONNECTED || chan->state == BT_CONFIG)
> @@ -258,6 +260,7 @@ static void l2cap_chan_timeout(struct work_struct
> *work)
>        l2cap_chan_close(chan, reason);
>
>        release_sock(sk);
> +       mutex_unlock(&conn->chan_lock);
>
>        chan->ops->close(chan->data);
>        l2cap_chan_put(chan);
> @@ -2619,6 +2622,7 @@ static inline int l2cap_connect_req(struct
> l2cap_conn *conn, struct l2cap_cmd_hd
>
>        parent = pchan->sk;
>
> +       mutex_lock(&conn->chan_lock);
>        lock_sock(parent);
>
>        /* Check if the ACL is secure enough (if not SDP) */
> @@ -2692,6 +2696,7 @@ static inline int l2cap_connect_req(struct
> l2cap_conn *conn, struct l2cap_cmd_hd
>
>  response:
>        release_sock(parent);
> +       mutex_unlock(&conn->chan_lock);
>
>  sendresp:
>        rsp.scid   = cpu_to_le16(scid);
> @@ -2733,6 +2738,7 @@ static inline int l2cap_connect_rsp(struct
> l2cap_conn *conn, struct l2cap_cmd_hd
>        struct l2cap_chan *chan;
>        struct sock *sk;
>        u8 req[128];
> +       int err;
>
>        scid   = __le16_to_cpu(rsp->scid);
>        dcid   = __le16_to_cpu(rsp->dcid);
> @@ -2742,16 +2748,24 @@ static inline int l2cap_connect_rsp(struct
> l2cap_conn *conn, struct l2cap_cmd_hd
>        BT_DBG("dcid 0x%4.4x scid 0x%4.4x result 0x%2.2x status 0x%2.2x",
>                                                dcid, scid, result,
> status);
>
> +       mutex_lock(&conn->chan_lock);
> +
>        if (scid) {
> -               chan = l2cap_get_chan_by_scid(conn, scid);
> -               if (!chan)
> -                       return -EFAULT;
> +               chan = __l2cap_get_chan_by_scid(conn, scid);
> +               if (!chan) {
> +                       err = -EFAULT;
> +                       goto unlock;
> +               }
>        } else {
> -               chan = l2cap_get_chan_by_ident(conn, cmd->ident);
> -               if (!chan)
> -                       return -EFAULT;
> +               chan = __l2cap_get_chan_by_ident(conn, cmd->ident);
> +               if (!chan) {
> +                       err = -EFAULT;
> +                       goto unlock;
> +               }
>        }
>
> +       err = 0;
> +
>        sk = chan->sk;
>
>        switch (result) {
> @@ -2779,7 +2793,10 @@ static inline int l2cap_connect_rsp(struct
> l2cap_conn *conn, struct l2cap_cmd_hd
>        }
>
>        release_sock(sk);
> -       return 0;
> +unlock:
> +       mutex_unlock(&conn->chan_lock);
> +
> +       return err;
>  }
>
>  static inline void set_default_fcs(struct l2cap_chan *chan)
> @@ -3025,9 +3042,13 @@ static inline int l2cap_disconnect_req(struct
> l2cap_conn *conn, struct l2cap_cmd
>
>        BT_DBG("scid 0x%4.4x dcid 0x%4.4x", scid, dcid);
>
> -       chan = l2cap_get_chan_by_scid(conn, dcid);
> -       if (!chan)
> +       mutex_lock(&conn->chan_lock);
> +
> +       chan = __l2cap_get_chan_by_scid(conn, dcid);
> +       if (!chan) {
> +               mutex_unlock(&conn->chan_lock);
>                return 0;
> +       }
>
>        sk = chan->sk;
>
> @@ -3039,6 +3060,7 @@ static inline int l2cap_disconnect_req(struct
> l2cap_conn *conn, struct l2cap_cmd
>
>        l2cap_chan_del(chan, ECONNRESET);
>        release_sock(sk);
> +       mutex_unlock(&conn->chan_lock);
>
>        chan->ops->close(chan->data);
>        return 0;
> @@ -3056,14 +3078,19 @@ static inline int l2cap_disconnect_rsp(struct
> l2cap_conn *conn, struct l2cap_cmd
>
>        BT_DBG("dcid 0x%4.4x scid 0x%4.4x", dcid, scid);
>
> -       chan = l2cap_get_chan_by_scid(conn, scid);
> -       if (!chan)
> +       mutex_lock(&conn->chan_lock);
> +
> +       chan = __l2cap_get_chan_by_scid(conn, scid);
> +       if (!chan) {
> +               mutex_unlock(&conn->chan_lock);
>                return 0;
> +       }
>
>        sk = chan->sk;
>
>        l2cap_chan_del(chan, 0);
>        release_sock(sk);
> +       mutex_unlock(&conn->chan_lock);
>
>        chan->ops->close(chan->data);
>        return 0;
> --
> 1.7.8.3


Yes, I do think they belong together. And please, check l2cap_sock.c
where l2cap_chan_close() seems to be called without locking
conn->chan_lock in l2cap_sock_shutdown(). And please remove the bogus
comment below from l2cap_conn_start, ok?

  /* l2cap_chan_close() calls list_del(chan)
   *  so release the lock */

Thanks,
Regards,

--
Ulisses Furquim
ProFUSION embedded systems
http://profusion.mobi
Mobile: +55 19 9250 0942
Skype: ulissesffs
--
To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Bluez Devel]     [Linux Wireless Networking]     [Linux Wireless Personal Area Networking]     [Linux ATH6KL]     [Linux USB Devel]     [Linux Media Drivers]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Big List of Linux Books]

  Powered by Linux