I think this patch will result in delays 10 times too long.


Nishanth Aravamudan wrote:
>  
> -/* helper function */
> -static void my_wait(struct saa7146_dev *dev, long ms)
> -{
> -	set_current_state(TASK_INTERRUPTIBLE);
> -	schedule_timeout((((ms+10)/10)*HZ)/1000);
> -}

ms seems to be a number that expresses an interval in units of 1/10 of a 
millisecond.

> -
>  u32 saa7146_i2c_func(struct i2c_adapter *adapter)
>  {
>  //fm	DEB_I2C(("'%s'.\n", adapter->name));
> @@ -136,12 +129,12 @@ static int saa7146_i2c_reset(struct saa7
>  		/* set "ABORT-OPERATION"-bit (bit 7)*/
>  		saa7146_write(dev, I2C_STATUS, (dev->i2c_bitrate | MASK_07));
>  		saa7146_write(dev, MC2, (MASK_00 | MASK_16));
> -		my_wait(dev,SAA7146_I2C_DELAY);
> +		msleep(SAA7146_I2C_DELAY);

But this would use it as a delay in milliseconds.



-- 
--------------------------
Mark Hollomon