* [PATCH]tty: Incorrect use tty_ldisc_flush() in TTY drivers.
@ 2012-11-23 18:42 Ilya Zykov
0 siblings, 0 replies; 2+ messages in thread
From: Ilya Zykov @ 2012-11-23 18:42 UTC (permalink / raw)
To: Alan Cox; +Cc: Greg Kroah-Hartman, Andrew McGregor, linux-kernel, ilya
This patch correct tty serial drivers.
Unfortunately, many drivers indirectly call ldisc's flush_buffer() function
in own callback function close(). Рarticularly, by the use of tty_ldisc_flush(),
before TTY LAYER calls ldisc's flush_buffer() in the right moment.
1. It disturb the logic of work ldisc's layer.
2. It is simple overhead because we call ldisc's flush_buffer() at least two times.
Signed-off-by: Ilya Zykov <ilya@ilyx.ru>
---
diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c
index a21dc8e..c4a0a6d 100644
--- a/drivers/tty/serial/serial_core.c
+++ b/drivers/tty/serial/serial_core.c
@@ -1288,7 +1288,7 @@ static void uart_close(struct tty_struct *tty, struct file *filp)
uart_shutdown(tty, state);
uart_flush_buffer(tty);
- tty_ldisc_flush(tty);
+ tty_buffer_flush(tty);
tty_port_tty_set(port, NULL);
spin_lock_irqsave(&port->lock, flags);
diff --git a/drivers/tty/tty_port.c b/drivers/tty/tty_port.c
index bf6e238..c23dd30 100644
--- a/drivers/tty/tty_port.c
+++ b/drivers/tty/tty_port.c
@@ -364,8 +364,8 @@ int tty_port_close_start(struct tty_port *port,
timeout = 2 * HZ;
schedule_timeout_interruptible(timeout);
}
- /* Flush the ldisc buffering */
- tty_ldisc_flush(tty);
+ /* Flush the driver buffering */
+ tty_buffer_flush(tty);
/* Drop DTR/RTS if HUPCL is set. This causes any attached modem to
hang up the line */
^ permalink raw reply related [flat|nested] 2+ messages in thread* [PATCH]tty: Incorrect use tty_ldisc_flush() in TTY drivers.
@ 2012-11-27 0:06 Ilya Zykov
0 siblings, 0 replies; 2+ messages in thread
From: Ilya Zykov @ 2012-11-27 0:06 UTC (permalink / raw)
To: Alan Cox
Cc: Andrew McGregor, Greg Kroah-Hartman, linux-serial, linux-kernel,
ilya
Sorry. Correct patch format.
Related bug 'tty: fix "IRQ45: nobody cared"'
Related commit 7b292b4bf9a9d6098440d85616d6ca4c608b8304
Unfortunately, some drivers indirectly call ldisc's flush_buffer() function
in own callback function close(). In particularly, by the use of tty_ldisc_flush(),
before TTY LAYER calls ldisc's flush_buffer() in the right moment.
1. It disturb the logic of work ldisc's layer.
2. It is simple overhead because we call ldisc's flush_buffer() at least two times.
Please, tell me what do you think about this? Can I make the adjustment for other drivers?
Signed-off-by: Ilya Zykov <ilya@ilyx.ru>
---
diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c
index 0fcfd98..d87f353 100644
--- a/drivers/tty/serial/serial_core.c
+++ b/drivers/tty/serial/serial_core.c
@@ -1304,7 +1304,7 @@ static void uart_close(struct tty_struct *tty, struct file *filp)
uart_shutdown(tty, state);
uart_flush_buffer(tty);
- tty_ldisc_flush(tty);
+ tty_buffer_flush(tty);
tty_port_tty_set(port, NULL);
spin_lock_irqsave(&port->lock, flags);
diff --git a/drivers/tty/tty_port.c b/drivers/tty/tty_port.c
index d7bdd8d..8a02996 100644
--- a/drivers/tty/tty_port.c
+++ b/drivers/tty/tty_port.c
@@ -428,8 +428,8 @@ int tty_port_close_start(struct tty_port *port,
timeout = 2 * HZ;
schedule_timeout_interruptible(timeout);
}
- /* Flush the ldisc buffering */
- tty_ldisc_flush(tty);
+ /* Flush the driver buffering */
+ tty_buffer_flush(tty);
/* Drop DTR/RTS if HUPCL is set. This causes any attached modem to
hang up the line */
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2012-11-27 0:07 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-11-23 18:42 [PATCH]tty: Incorrect use tty_ldisc_flush() in TTY drivers Ilya Zykov
-- strict thread matches above, loose matches on Subject: below --
2012-11-27 0:06 Ilya Zykov
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox