TTY: serial, stop accessing potential NULLs
authorJiri Slaby <jslaby@suse.cz>
Wed, 13 Mar 2013 23:30:34 +0000 (00:30 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 19 Mar 2013 00:09:37 +0000 (17:09 -0700)
The following commits:
6732c8bb8671acbdac6cdc93dd72ddd581dd5e25 (TTY: switch
  tty_schedule_flip)
2e124b4a390ca85325fae75764bef92f0547fa25 (TTY: switch
  tty_flip_buffer_push)
05c7cd39907184328f48d3e7899f9cdd653ad336 (TTY: switch
  tty_insert_flip_string)
92a19f9cec9a80ad93c06e115822deb729e2c6ad (TTY: switch
  tty_insert_flip_char)
227434f8986c3827a1faedd1feb437acd6285315 (TTY: switch
  tty_buffer_request_room to tty_port)

introduced a potential NULL dereference to some drivers. In
particular, when the device is used as a console, incoming bytes can
kill the box. This is caused by removed checks for TTY against NULL.

It happened because it was unclear to me why the checks were there. I
assumed them superfluous because the interrupts were unbound or
otherwise stopped. But this is not the case for consoles for these
drivers, as was pointed out by David Miller.

Now, this patch re-introduces the checks (at this point we check
port->state, not the tty proper, as we do not care about tty pointers
anymore). For both of the drivers, we place the check below the
handling of break signal so that sysrq can actually work. (One needs
to issue a break and then sysrq key within the following 5 seconds.)

We do not change sc26xx, sunhv, and sunsu here because they behave the
same as before.  People having that hardware should fix the driver
eventually, however. They always could unconditionally dereference tty
in receive_chars, port->state in uart_handle_dcd_change, and
up->port.state->port.tty.

There is perhaps more to fix in all those drivers, but they are at
least in a state they were before.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Rob Herring <rob.herring@calxeda.com>
Cc: sparclinux@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/tty/serial/sunsab.c
drivers/tty/serial/sunzilog.c

index 8de2213664e0c8bd655fcd2ce5d19d6cd23de05b..a422c8b55a47b427b34ec09a3bda5b563ea6d8a0 100644 (file)
@@ -203,7 +203,7 @@ receive_chars(struct uart_sunsab_port *up,
                                flag = TTY_FRAME;
                }
 
-               if (uart_handle_sysrq_char(&up->port, ch))
+               if (uart_handle_sysrq_char(&up->port, ch) || !port)
                        continue;
 
                if ((stat->sreg.isr0 & (up->port.ignore_status_mask & 0xff)) == 0 &&
index 27669ff3d4463b0c023f7eb1120b8ed96792e39f..813ef8eb8effaa5b6b2d9116dc3c2a4c43cba4f5 100644 (file)
@@ -388,7 +388,7 @@ sunzilog_receive_chars(struct uart_sunzilog_port *up,
                        else if (r1 & CRC_ERR)
                                flag = TTY_FRAME;
                }
-               if (uart_handle_sysrq_char(&up->port, ch))
+               if (uart_handle_sysrq_char(&up->port, ch) || !port)
                        continue;
 
                if (up->port.ignore_status_mask == 0xff ||