[v1] staging: av7110: Replace msleep() with usleep_range() in tuner function

[PATCH] staging: av7110: Replace msleep() with usleep_range() in tuner function

Posted by Pratibimba Khadka 3 months, 1 week ago

Replace msleep(10) with usleep_range(10000, 12000) in the PLL lock
waiting loop of nexusca_stv0297_tuner_set_params().

usleep_range() is preferred over msleep() for short delays (< 20ms) as
it provides better accuracy and allows the scheduler flexibility for
power management optimizations. The 10-12ms range maintains the same
timing behavior while giving the kernel scheduler room to optimize.

This change improves timing precision during the PLL lock detection
phase without affecting the overall functionality of the tuner
configuration process.

Signed-off-by: Pratibimba Khadka <pratibimbakhadka@gmail.com>
---
 drivers/staging/media/av7110/av7110.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/staging/media/av7110/av7110.c b/drivers/staging/media/av7110/av7110.c
index bc9a2a40afcb..7f6c610579c8 100644
--- a/drivers/staging/media/av7110/av7110.c
+++ b/drivers/staging/media/av7110/av7110.c
@@ -1827,7 +1827,7 @@ static int nexusca_stv0297_tuner_set_params(struct dvb_frontend *fe)
 		if (i2c_transfer(&av7110->i2c_adap, &readmsg, 1) == 1)
 			if (data[0] & 0x40)
 				break;
-		msleep(10);
+		usleep_range(10000, 12000);
 	}
 
 	return 0;
-- 
2.49.0

Re: [PATCH] staging: av7110: Replace msleep() with usleep_range() in tuner function

Posted by Dan Carpenter 3 months, 1 week ago

On Thu, Jun 26, 2025 at 09:53:50PM +0545, Pratibimba Khadka wrote:
> Replace msleep(10) with usleep_range(10000, 12000) in the PLL lock
> waiting loop of nexusca_stv0297_tuner_set_params().
> 
> usleep_range() is preferred over msleep() for short delays (< 20ms) as
> it provides better accuracy and allows the scheduler flexibility for
> power management optimizations. The 10-12ms range maintains the same
> timing behavior while giving the kernel scheduler room to optimize.
> 
> This change improves timing precision during the PLL lock detection
> phase without affecting the overall functionality of the tuner
> configuration process.
> 
> Signed-off-by: Pratibimba Khadka <pratibimbakhadka@gmail.com>

Sorry, these msleep() need to be tested on real hardware for us to
accept them.

regards,
dan carpenter