On 01/28/2016 02:13 PM, Greg Kroah-Hartman wrote:

On Sun, Jan 10, 2016 at 08:36:16PM -0800, Peter Hurley wrote:

quoted

tty->flags are handled with atomic bit ops; use test_bit() treewide.

Signed-off-by: Peter Hurley <redacted>
---
 arch/ia64/hp/sim/simserial.c       | 2 +-
 drivers/char/pcmcia/synclink_cs.c  | 2 +-
 drivers/isdn/i4l/isdn_tty.c        | 6 +++---
 drivers/s390/char/tty3270.c        | 4 ++--
 drivers/staging/dgap/dgap.c        | 2 +-
 drivers/staging/dgnc/dgnc_tty.c    | 2 +-
 drivers/tty/amiserial.c            | 6 +++---
 drivers/tty/serial/68328serial.c   | 2 +-
 drivers/tty/serial/crisv10.c       | 4 ++--
 drivers/tty/serial/serial_core.c   | 6 +++---
 drivers/tty/synclink.c             | 4 ++--
 drivers/tty/synclink_gt.c          | 4 ++--
 drivers/tty/synclinkmp.c           | 4 ++--
 drivers/tty/tty_port.c             | 2 +-
 net/irda/ircomm/ircomm_tty_ioctl.c | 6 +++---
 15 files changed, 28 insertions(+), 28 deletions(-)

diff --git a/arch/ia64/hp/sim/simserial.c b/arch/ia64/hp/sim/simserial.c
index e70cade..2ab9ba6 100644
--- a/arch/ia64/hp/sim/simserial.c
+++ b/arch/ia64/hp/sim/simserial.c

@@ -300,7 +300,7 @@ static int rs_ioctl(struct tty_struct *tty, unsigned int cmd, unsigned long arg)
 	if ((cmd != TIOCGSERIAL) && (cmd != TIOCSSERIAL) &&
 	    (cmd != TIOCSERCONFIG) && (cmd != TIOCSERGSTRUCT) &&
 	    (cmd != TIOCMIWAIT)) {
-		if (tty->flags & (1 << TTY_IO_ERROR))
+		if (test_bit(TTY_IO_ERROR, &tty->flags))

I thought you were going to wrap these up in a tty_*() macro instead of
open-coding it everywhere?  That way you can change how those
bits/features are represented in the future, which is what I thought you
wanted to do?

I wasn't going to do that for tty->flags, just for tty_port->flags.

Even if you don't want to do that, something like:

static inline int tty_is_io_error(struct tty_struct *tty)
{
	return test_bit(TTY_IO_ERROR, &tty->flags);
}

and then this patch would just be:
		if (tty_is_io_error(tty))

Or something like that, I'm not sold on the name, I'm sure you can pick
a better one :)

Sound reasonable?

But I agree that seems better; I can certainly do something like that.
I think just tty_io_error().
 

I'll not apply this patch in the series for now.

Regards,
Peter Hurley