Skip to contentSkip to conversation

i2c: imx: remove unnecessary delay at startup

PATCH In review Ian Dannapel 4 participants 4 messages Mar 14, 2022 – Mar 15, 2022

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 i2c: imx: remove unnecessary delay at startup

a delay on the startup of the i2c imx controller is not required or defined in the specs. By removing it, the user can see a latency decrease from up to 150μs in communication. Additional info: https://lore.kernel.org/all/[email protected]/

---
 drivers/i2c/busses/i2c-imx.c | 6 ------
 1 file changed, 6 deletions(-)
drivers/i2c/busses/i2c-imx.c +0 -6
index 3576b63a6c03..019dda5301df 100644
@@ -602,12 +602,6 @@ static int i2c_imx_start(struct imx_i2c_struct *i2c_imx, bool atomic)
 	imx_i2c_write_reg(i2c_imx->hwdata->i2sr_clr_opcode, i2c_imx, IMX_I2C_I2SR);
 	imx_i2c_write_reg(i2c_imx->hwdata->i2cr_ien_opcode, i2c_imx, IMX_I2C_I2CR);
 
-	/* Wait controller to be stable */
-	if (atomic)
-		udelay(50);
-	else
-		usleep_range(50, 150);
-
 	/* Start I2C transaction */
 	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
 	temp |= I2CR_MSTA;
Signature · 1 line
2.25.1
#1 lore raw

Hello,

Quoted 7 lines · Ian Dannapel
drivers/i2c/busses/i2c-imx.c +0 -0
index 3576b63a6c03..019dda5301df 100644
@@ -602,12 +602,6 @@ static int i2c_imx_start(struct imx_i2c_struct *i2c_imx, bool atomic)
 	imx_i2c_write_reg(i2c_imx->hwdata->i2sr_clr_opcode, i2c_imx, IMX_I2C_I2SR);
 	imx_i2c_write_reg(i2c_imx->hwdata->i2cr_ien_opcode, i2c_imx, IMX_I2C_I2CR);

drivers/i2c/busses/i2c-imx.c static int i2c_imx_start(struct imx_i2c_struct *i2c_imx, bool atomic) · line 607

Quoted 10 lines
drivers/i2c/busses/i2c-imx.c +0 -6
 
-	/* Wait controller to be stable */
-	if (atomic)
-		udelay(50);
-	else
-		usleep_range(50, 150);
-
 	/* Start I2C transaction */
 	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
 	temp |= I2CR_MSTA;

This contradicts statements made in
43309f3b521302bb66c4c9e66704dd3675e4d725.

Maybe the sleep/delay should be done conditionally on the busy bit?

Best regards
Uwe

Signature · 2 lines
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |
maintainer
#2 lore raw
Quoted 24 lines · Uwe Kleine-König

Hello,

Quoted 7 lines · Ian Dannapel
drivers/i2c/busses/i2c-imx.c +0 -0
index 3576b63a6c03..019dda5301df 100644
@@ -602,12 +602,6 @@ static int i2c_imx_start(struct imx_i2c_struct *i2c_imx, bool atomic)
 	imx_i2c_write_reg(i2c_imx->hwdata->i2sr_clr_opcode, i2c_imx, IMX_I2C_I2SR);
 	imx_i2c_write_reg(i2c_imx->hwdata->i2cr_ien_opcode, i2c_imx, IMX_I2C_I2CR);

drivers/i2c/busses/i2c-imx.c static int i2c_imx_start(struct imx_i2c_struct *i2c_imx, bool atomic) · line 607

Quoted 10 lines
drivers/i2c/busses/i2c-imx.c +0 -6
 
-	/* Wait controller to be stable */
-	if (atomic)
-		udelay(50);
-	else
-		usleep_range(50, 150);
-
 	/* Start I2C transaction */
 	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
 	temp |= I2CR_MSTA;

This contradicts statements made in
43309f3b521302bb66c4c9e66704dd3675e4d725.

Quoted 1 line

Maybe the sleep/delay should be done conditionally on the busy bit?

Maybe. I do not see what exact problem is this sleep addressing. How exact the can get stable? Is it addressing some clock issue or wading until something happening on the bus?

Regards,
Oleskij

Signature · 4 lines
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
#3 lore raw
Quoted 4 lines

Maybe. I do not see what exact problem is this sleep addressing. How exact the can get stable? Is it addressing some clock issue or wading until something happening on the bus?

Yeah, it is really hard to see why it is there. And given that the introducing commit is from 2009, we can't really expect an active userbase reporting breakage. So, although it may be a useless delay, I'd like to see an opt-in approach for not using it.