Skip to contentSkip to conversation

serial: 8250: revert UART_CAP_NOTEMT changes

PATCH In review Ilpo Järvinen 2 participants 3 messages Apr 19, 2022 – Apr 20, 2022

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 serial: 8250: revert UART_CAP_NOTEMT changes

This reverts UART_CAP_NOTEMT commit and driver changes depending on it:
f6f586102ad1 (serial: 8250: Handle UART without interrupt on TEMT

                using em485)

296385fe1275 (serial: 8250: Add UART_CAP_NOTEMT on PORT_16550A_FSL64) bec1f1b66a6a (serial: 8250: add compatible for fsl,16550-FIFO64)

The UART_CAP_NOTEMT code added in f6f586102add1 (serial: 8250:
Handle UART without interrupt on TEMT using em485) containts math overflow for 32-bit archs. In addition, the approach used in it is unnecessarily complicated requiring a dedicated timer just for notemt. A simpler approach for providing UART_CAP_NOTEMT already exists (patches 1-2):
https://lore.kernel.org/linux-serial/[email protected]/T/#u
Thus, simply revert the UART_CAP_NOTEMT changes for now.

---
 drivers/tty/serial/8250/8250.h      |  1 -
 drivers/tty/serial/8250/8250_of.c   |  2 -
 drivers/tty/serial/8250/8250_port.c | 78 ++---------------------------
 include/linux/serial_8250.h         |  2 -
 4 files changed, 3 insertions(+), 80 deletions(-)
drivers/tty/serial/8250/8250.h +0 -1
index 39ffeb37786f..db784ace25d8 100644
@@ -83,7 +83,6 @@ struct serial8250_config {
 #define UART_CAP_MINI	BIT(17)	/* Mini UART on BCM283X family lacks:
 					 * STOP PARITY EPAR SPAR WLEN5 WLEN6
 					 */
-#define UART_CAP_NOTEMT	BIT(18)	/* UART without interrupt on TEMT available */
 
 #define UART_BUG_QUOT	BIT(0)	/* UART has buggy quot LSB */
 #define UART_BUG_TXEN	BIT(1)	/* UART has buggy TX IIR status */
drivers/tty/serial/8250/8250_of.c +0 -2
index 5a699a1aa79c..be8626234627 100644
@@ -326,8 +326,6 @@ static const struct of_device_id of_platform_serial_table[] = {
 		.data = (void *)PORT_ALTR_16550_F64, },
 	{ .compatible = "altr,16550-FIFO128",
 		.data = (void *)PORT_ALTR_16550_F128, },
-	{ .compatible = "fsl,16550-FIFO64",
-		.data = (void *)PORT_16550A_FSL64, },
 	{ .compatible = "mediatek,mtk-btif",
 		.data = (void *)PORT_MTK_BTIF, },
 	{ .compatible = "mrvl,mmp-uart",
drivers/tty/serial/8250/8250_port.c +3 -75

134 lines not on this page · raw message

include/linux/serial_8250.h +0 -2
index de135852107c..ff84a3ed10ea 100644
@@ -79,9 +79,7 @@ struct uart_8250_ops {
 struct uart_8250_em485 {
 	struct hrtimer		start_tx_timer; /* "rs485 start tx" timer */
 	struct hrtimer		stop_tx_timer;  /* "rs485 stop tx" timer */
-	struct hrtimer		no_temt_timer;  /* "rs485 no TEMT interrupt" timer */
 	struct hrtimer		*active_timer;  /* pointer to active timer */
-	unsigned long		no_temt_delay;  /* Delay for no_temt_timer */
 	struct uart_8250_port	*port;          /* for hrtimer callbacks */
 	unsigned int		tx_stopped:1;	/* tx is currently stopped */
 };
Signature · 1 line
2.30.2
#1 lore raw

Hello,

Quoted 17 lines · Ilpo Järvinen

This reverts UART_CAP_NOTEMT commit and driver changes depending on it:
f6f586102ad1 (serial: 8250: Handle UART without interrupt on TEMT

                using em485)

296385fe1275 (serial: 8250: Add UART_CAP_NOTEMT on PORT_16550A_FSL64) bec1f1b66a6a (serial: 8250: add compatible for fsl,16550-FIFO64)

The UART_CAP_NOTEMT code added in f6f586102add1 (serial: 8250:
Handle UART without interrupt on TEMT using em485) containts math overflow for 32-bit archs. In addition, the approach used in it is unnecessarily complicated requiring a dedicated timer just for notemt. A simpler approach for providing UART_CAP_NOTEMT already exists (patches 1-2):
https://lore.kernel.org/linux-serial/[email protected]/T/#u
Thus, simply revert the UART_CAP_NOTEMT changes for now.

Quoted 1 line

Oh I wasn't aware that Greg picked that up. OK for me to revert.

I wonder however if it's nice to revert three patches in one commit. I would have just reverted f6f586102ad1 and kept the define UART_CAP_NOTEMT such that the other two patches are noops until your fixed series comes in. Just my 0.02€.

Best regards
Uwe

Signature · 2 lines
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |
#2 lore raw
Quoted 20 lines · Uwe Kleine-König

Hello,

Quoted 17 lines · Ilpo Järvinen

This reverts UART_CAP_NOTEMT commit and driver changes depending on it:
f6f586102ad1 (serial: 8250: Handle UART without interrupt on TEMT

                using em485)

296385fe1275 (serial: 8250: Add UART_CAP_NOTEMT on PORT_16550A_FSL64) bec1f1b66a6a (serial: 8250: add compatible for fsl,16550-FIFO64)

The UART_CAP_NOTEMT code added in f6f586102add1 (serial: 8250:
Handle UART without interrupt on TEMT using em485) containts math overflow for 32-bit archs. In addition, the approach used in it is unnecessarily complicated requiring a dedicated timer just for notemt. A simpler approach for providing UART_CAP_NOTEMT already exists (patches 1-2):
https://lore.kernel.org/linux-serial/[email protected]/T/#u
Thus, simply revert the UART_CAP_NOTEMT changes for now.

Quoted 8 lines
Quoted 1 line

Oh I wasn't aware that Greg picked that up. OK for me to revert.

I wonder however if it's nice to revert three patches in one commit. I would have just reverted f6f586102ad1 and kept the define UART_CAP_NOTEMT such that the other two patches are noops until your fixed series comes in. Just my 0.02€.

I was thinking along the same lines but was a little worried Greg would be against such a solution. ...I'll send v2 with only that single commit reverted.

Signature · 1 line
 i.