i2c: fix bit algorithm timeout for low values.

PATCH Applied Dave Airlie 3 participants 4 messages Apr 23, 2009 – May 5, 2009 I2C subsystem host drivers

lore

Share & tools

lore

Dave Airlie

author

2009-04-23 23:09 UTC #0 lore raw

PATCH 1/1 i2c: fix bit algorithm timeout for low values.

From: Dave Airlie <[email protected]>

When fetching DDC using i2c algo bit, we were often seeing timeouts before getting valid EDID on a retry. The VESA spec states 2ms is the DDC timeout, so when this translates into 1 jiffie and we are close to the end of the time period, it could return with a timeout less than 2ms.

Change this code to use time_after instead of time_after_eq.

Signed-off-by Dave Airlie <[email protected]>

---
 drivers/i2c/algos/i2c-algo-bit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

drivers/i2c/algos/i2c-algo-bit.c +1 -1

index d420cc5..e25e139 100644
@@ -104,7 +104,7 @@ static int sclhi(struct i2c_algo_bit_data *adap)
 		 * chips may hold it low ("clock stretching") while they
 		 * are processing data internally.
 		 */
-		if (time_after_eq(jiffies, start + adap->timeout))
+		if (time_after(jiffies, start + adap->timeout))
 			return -ETIMEDOUT;
 		cond_resched();
 	}

Signature · 1 line

1.6.0.6

Jean Delvare

Applied

2009-04-24 09:42 UTC #1 lore raw

Hi Dave,

Quoted 20 lines · Dave Airlie

From: Dave Airlie <[email protected]>
When fetching DDC using i2c algo bit, we were often seeing timeouts before getting valid EDID on a retry. The VESA spec states 2ms is the DDC timeout, so when this translates into 1 jiffie and we are close to the end of the time period, it could return with a timeout less than 2ms.
Change this code to use time_after instead of time_after_eq.
Signed-off-by Dave Airlie <[email protected]>
---
 drivers/i2c/algos/i2c-algo-bit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
drivers/i2c/algos/i2c-algo-bit.c +0 -0
index d420cc5..e25e139 100644

drivers/i2c/algos/i2c-algo-bit.c static int sclhi(struct i2c_algo_bit_data *adap) · line 110

Quoted 8 lines

drivers/i2c/algos/i2c-algo-bit.c +1 -1

@@ -104,7 +104,7 @@ static int sclhi(struct i2c_algo_bit_data *adap)
 		 * chips may hold it low ("clock stretching") while they
 		 * are processing data internally.
 		 */
-		if (time_after_eq(jiffies, start + adap->timeout))
+		if (time_after(jiffies, start + adap->timeout))
 			return -ETIMEDOUT;
 		cond_resched();
 	}

Applied, thanks. I agree this should go in stable.

Signature · 1 line

Jean Delvare

Greg KH

2009-05-04 21:53 UTC #2 lore raw

Quoted 18 lines · Jean Delvare

Hi Dave,
Quoted 15 lines · Dave Airlie
From: Dave Airlie <[email protected]>
When fetching DDC using i2c algo bit, we were often seeing timeouts before getting valid EDID on a retry. The VESA spec states 2ms is the DDC timeout, so when this translates into 1 jiffie and we are close to the end of the time period, it could return with a timeout less than 2ms.
Change this code to use time_after instead of time_after_eq.
Signed-off-by Dave Airlie <[email protected]>
---
 drivers/i2c/algos/i2c-algo-bit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

Quoted 15 lines

drivers/i2c/algos/i2c-algo-bit.c static int sclhi(struct i2c_algo_bit_data *adap) · line 110
Quoted 13 lines
drivers/i2c/algos/i2c-algo-bit.c +1 -1
index d420cc5..e25e139 100644
@@ -104,7 +104,7 @@ static int sclhi(struct i2c_algo_bit_data *adap)
 		 * chips may hold it low ("clock stretching") while they
 		 * are processing data internally.
 		 */
-		if (time_after_eq(jiffies, start + adap->timeout))
+		if (time_after(jiffies, start + adap->timeout))
 			return -ETIMEDOUT;
 		cond_resched();
 	}
Applied, thanks. I agree this should go in stable.

I don't see this in Linus's tree anywhere, am I just missing it?

thanks,

greg k-h

Jean Delvare

2009-05-05 06:05 UTC #3 lore raw

Hi Greg,

Quoted 35 lines · Greg KH

Quoted 18 lines · Jean Delvare
Hi Dave,
Quoted 15 lines · Dave Airlie
From: Dave Airlie <[email protected]>
When fetching DDC using i2c algo bit, we were often seeing timeouts before getting valid EDID on a retry. The VESA spec states 2ms is the DDC timeout, so when this translates into 1 jiffie and we are close to the end of the time period, it could return with a timeout less than 2ms.
Change this code to use time_after instead of time_after_eq.
Signed-off-by Dave Airlie <[email protected]>
---
 drivers/i2c/algos/i2c-algo-bit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
Quoted 15 lines
drivers/i2c/algos/i2c-algo-bit.c static int sclhi(struct i2c_algo_bit_data *adap) · line 110
Quoted 13 lines
drivers/i2c/algos/i2c-algo-bit.c +1 -1
index d420cc5..e25e139 100644
@@ -104,7 +104,7 @@ static int sclhi(struct i2c_algo_bit_data *adap)
 		 * chips may hold it low ("clock stretching") while they
 		 * are processing data internally.
 		 */
-		if (time_after_eq(jiffies, start + adap->timeout))
+		if (time_after(jiffies, start + adap->timeout))
 			return -ETIMEDOUT;
 		cond_resched();
 	}
Applied, thanks. I agree this should go in stable.

Quoted 1 line

I don't see this in Linus's tree anywhere, am I just missing it?

You're correct, it's not there yet, I'll push it to Linus today. Busy busy busy.

Signature · 1 line

Jean Delvare