Hi Jesper,

	the following patch fixes the usage of the delay in RS485 mode
on the Cris architecture.

Very recently, the RS485 interface has been fixed by adding two further
fields (see commit 1b6331848b69d1ed165a6bdc75c4046d68767563).

This patch checks the value of the flag SER_RS485_RTS_BEFORE_SEND before
delaying.

Best regards,

	Claudio


CRIS RS485: Check SER_RS485_RTS_BEFORE_SEND before delaying.

Signed-off-by: Claudio Scordino <claudio@evidence.eu.com>
---
 drivers/serial/crisv10.c |   21 ++++++++++++++-------
 1 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/drivers/serial/crisv10.c b/drivers/serial/crisv10.c
index c856905..a2237e7 100644
--- a/drivers/serial/crisv10.c
+++ b/drivers/serial/crisv10.c
@@ -1411,11 +1411,12 @@ e100_enable_rs485(struct tty_struct *tty, struct serial_rs485 *r)
 		       CONFIG_ETRAX_RS485_LTC1387_RXEN_PORT_G_BIT, 1);
 #endif
 
-	info->rs485.flags = r->flags;
-	if (r->delay_rts_before_send >= 1000)
+	info->rs485 = *r;
+
+	/* Maximum delay before RTS equal to 1000 */
+	if (info->rs485.delay_rts_before_send >= 1000)
 		info->rs485.delay_rts_before_send = 1000;
-	else
-		info->rs485.delay_rts_before_send = r->delay_rts_before_send;
+
 /*	printk("rts: on send = %i, after = %i, enabled = %i",
 		    info->rs485.rts_on_send,
 		    info->rs485.rts_after_sent,
@@ -3234,9 +3235,9 @@ rs_write(struct tty_struct *tty,
 		e100_disable_rx(info);
 		e100_enable_rx_irq(info);
 #endif
-
-		if (info->rs485.delay_rts_before_send > 0)
-			msleep(info->rs485.delay_rts_before_send);
+		if ((info->rs485.flags & SER_RS485_RTS_BEFORE_SEND) &&
+			(info->rs485.delay_rts_before_send > 0))
+				msleep(info->rs485.delay_rts_before_send);
 	}
 #endif /* CONFIG_ETRAX_RS485 */
 
@@ -3694,6 +3695,11 @@ rs_ioctl(struct tty_struct *tty, struct file * file,
 
 		rs485data.delay_rts_before_send = rs485ctrl.delay_rts_before_send;
 		rs485data.flags = 0;
+		if (rs485data.delay_rts_before_send != 0)
+			rs485data.flags |= ...
[ message continues ]

Hi all,

	so far I didn't receive any acknowledgement about this patch.

As I said, recently (see commit 1b6331848b69d1ed165a6bdc75c4046d68767563) the
Linux RS485 interface has been fixed by adding two missing fields.

With the "new" interface, you have to check the SER_RS485_RTS_BEFORE_SEND flag
before using the delay_rts_before_send field.

The Atmel serial driver has been already fixed (with commit
1b6331848b69d1ed165a6bdc75c4046d68767563 itself).

This patch fixes the usage of the interface in the Cris driver too.

Best regards,

		Claudio
--

[ message continues ]

Hi!

Jesper is on vacation but I find this patch ok.

Acked-by: Mikael Starvik<Starvik@axis.com>

/Mikael

-----Original Message-----
From: Claudio Scordino [mailto:claudio@evidence.eu.com] 
Sent: den 30 augusti 2010 10:32
To: Jesper Nilsson
Cc: Mikael Starvik; linux-cris-kernel; Alan Cox; Andrew Morton; Hinko Kocevar; Janez Cufer; Andy Whitcroft; Linux Kernel; dev-etrax; linux-serial@vger.kernel.org
Subject: Re: [PATCH] CRIS RS485: Check SER_RS485_RTS_BEFORE_SEND before delaying.



Hi all,

	so far I didn't receive any acknowledgement about this patch.

As I said, recently (see commit 1b6331848b69d1ed165a6bdc75c4046d68767563) the
Linux RS485 interface has been fixed by adding two missing fields.

With the "new" interface, you have to check the SER_RS485_RTS_BEFORE_SEND flag
before using the delay_rts_before_send field.

The Atmel serial driver has been already fixed (with commit
1b6331848b69d1ed165a6bdc75c4046d68767563 itself).

This patch fixes the usage of the interface in the Cris driver too.

Best regards,

		Claudio
--

[ message continues ]

Hi Mikael, Hi Jesper,

	the merge window will be opened very soon. So, please remember
this patch.

I already fixed the atmel_serial driver (commit number
1b6331848b69d1ed165a6bdc75c4046d68767563)
	
Now, it's quite important to fix the Cris driver as well.

Cheers,

       Claudio


CRIS RS485: Check SER_RS485_RTS_BEFORE_SEND before delaying.

Signed-off-by: Claudio Scordino <claudio@evidence.eu.com>
Acked-by: Mikael Starvik <Starvik@axis.com>
---
 drivers/serial/crisv10.c |   21 ++++++++++++++-------
 1 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/drivers/serial/crisv10.c b/drivers/serial/crisv10.c
index c856905..a2237e7 100644
--- a/drivers/serial/crisv10.c
+++ b/drivers/serial/crisv10.c
@@ -1411,11 +1411,12 @@ e100_enable_rs485(struct tty_struct *tty, struct serial_rs485 *r)
 		       CONFIG_ETRAX_RS485_LTC1387_RXEN_PORT_G_BIT, 1);
 #endif
 
-	info->rs485.flags = r->flags;
-	if (r->delay_rts_before_send >= 1000)
+	info->rs485 = *r;
+
+	/* Maximum delay before RTS equal to 1000 */
+	if (info->rs485.delay_rts_before_send >= 1000)
 		info->rs485.delay_rts_before_send = 1000;
-	else
-		info->rs485.delay_rts_before_send = r->delay_rts_before_send;
+
 /*	printk("rts: on send = %i, after = %i, enabled = %i",
 		    info->rs485.rts_on_send,
 		    info->rs485.rts_after_sent,
@@ -3234,9 +3235,9 @@ rs_write(struct tty_struct *tty,
 		e100_disable_rx(info);
 		e100_enable_rx_irq(info);
 #endif
-
-		if (info->rs485.delay_rts_before_send > 0)
-			msleep(info->rs485.delay_rts_before_send);
+		if ((info->rs485.flags & SER_RS485_RTS_BEFORE_SEND) &&
+			(info->rs485.delay_rts_before_send > 0))
+				msleep(info->rs485.delay_rts_before_send);
 	}
 #endif /* CONFIG_ETRAX_RS485 */
 
@@ -3694,6 +3695,11 @@ rs_ioctl(struct tty_struct *tty, struct file * file,
 
 		rs485data.delay_rts_before_send = rs485ctrl.delay_rts_before_send;
 		rs485data.flags = 0;
+		if (rs485data.delay_rts_before_send != 0)
+			rs485data.flags |= ...
[ message continues ]