re: tty: Change tty lock order to master->slave

re: tty: Change tty lock order to master->slave

[Dan Carpenter]

Hello Peter Hurley,

The patch 2aff5e2bc62d: "tty: Change tty lock order to master->slave"
from Nov 5, 2014, leads to the following static checker warning:

	drivers/tty/tty_mutex.c:50 tty_lock_slave()
	warn: add some parenthesis here?

drivers/tty/tty_mutex.c
    47  void __lockfunc tty_lock_slave(struct tty_struct *tty)
    48  {
    49          if (tty && tty != tty->link) {
    50                  WARN_ON(!mutex_is_locked(&tty->link->legacy_mutex) ||
    51                          !tty->driver->type = TTY_DRIVER_TYPE_PTY ||
    52                          !tty->driver->type = PTY_TYPE_SLAVE);

! has high precedence than = so those tests are never true.  != doesn't
make sense either.  Probably the negate operatons are just copy-and-paste
errors?

    53                  tty_lock(tty);
    54          }
    55  }

regards,
dan carpenter

Re: tty: Change tty lock order to master->slave

[Peter Hurley]

On 11/11/2014 07:50 AM, Dan Carpenter wrote:
> Hello Peter Hurley,
>
> The patch 2aff5e2bc62d: "tty: Change tty lock order to master->slave"
> from Nov 5, 2014, leads to the following static checker warning:
>
> 	drivers/tty/tty_mutex.c:50 tty_lock_slave()
> 	warn: add some parenthesis here?
>
> drivers/tty/tty_mutex.c
>     47  void __lockfunc tty_lock_slave(struct tty_struct *tty)
>     48  {
>     49          if (tty && tty != tty->link) {
>     50                  WARN_ON(!mutex_is_locked(&tty->link->legacy_mutex) ||
>     51                          !tty->driver->type = TTY_DRIVER_TYPE_PTY ||
>     52                          !tty->driver->type = PTY_TYPE_SLAVE);
>
> ! has high precedence than = so those tests are never true.  != doesn't
> make sense either.  Probably the negate operatons are just copy-and-paste
> errors?

Thanks for catching this, Dan.

Unfortunately, this is not a copy-and-paste error; apparently, emacs cannot
code-what-I-mean rather than code-what-I-type.

The warning is not necessary now; when I respun this series with fixed lock
subclasses I should have ripped this out. I will send a patch forthwith.

Regards,
Peter Hurley


>     53                  tty_lock(tty);
>     54          }
>     55  }
>
> regards,
> dan carpenter