tty: n_gsm: fix mux cleanup after unregister tty device

Internally, we manage the alive state of the mux channels and mux itself
with the field member 'dead'. This makes it possible to notify the user
if the accessed underlying link is already gone. On the other hand,
however, removing the virtual ttys before terminating the channels may
result in peer messages being received without any internal target. Move
the mux cleanup procedure from gsmld_detach_gsm() to gsmld_close() to fix
this by keeping the virtual ttys open until the mux has been cleaned up.

Fixes: e1eaea46bb40 ("tty: n_gsm line discipline")
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Starke <daniel.starke@siemens.com>
Link: https://lore.kernel.org/r/20220414094225.4527-4-daniel.starke@siemens.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/drivers/tty/n_gsm.c b/drivers/tty/n_gsm.c
index f546dfe..de97a38 100644 (file)
--- a/drivers/tty/n_gsm.c
+++ b/drivers/tty/n_gsm.c
@@ -2479,7 +2479,6 @@ static void gsmld_detach_gsm(struct tty_struct *tty, struct gsm_mux *gsm)
                for (i = 1; i < NUM_DLCI; i++)
                        tty_unregister_device(gsm_tty_driver, base + i);
        }
-       gsm_cleanup_mux(gsm, false);
        tty_kref_put(gsm->tty);
        gsm->tty = NULL;
 }
@@ -2544,6 +2543,12 @@ static void gsmld_close(struct tty_struct *tty)
 {
        struct gsm_mux *gsm = tty->disc_data;
 
+       /* The ldisc locks and closes the port before calling our close. This
+        * means we have no way to do a proper disconnect. We will not bother
+        * to do one.
+        */
+       gsm_cleanup_mux(gsm, false);
+
        gsmld_detach_gsm(tty, gsm);
 
        gsmld_flush_buffer(tty);