On Monday, March 01, 2010 6:14 AM Mark Brown wrote:
> On Fri, Feb 26, 2010 at 06:22:34PM -0600, Olaya, Margarita wrote:
>
>> Do you mean something like this?
>> time_left = wait_for_completion_timeout(&priv->ready,
>> msecs_to_jiffies(48));
>> if(!time_left) { twl_i2c_read_u8(TWL4030_MODULE_AUDIO_VOICE,
>> &intid, TWL6040_REG_INTID);
>> if (!(intid & TWL6040_READYINT))
>> goto error;
>> }
>>
>> return 0;
>>
>> error:
>> dev_err(codec->dev, "timeout waiting for READYINT\n");
>> return -ETIMEDOUT;
>
> Yes, or wrapped in a for loop with shorter timeouts on the
> individual waits.
>
>> but in this case will it not take unnecessarily 48ms when the
>> interruption line is not valid?
>
> You're always going to get some additional delay when polling
> unless you
> busy wait for completion, which obviously has its own problems.
Ok, thanks for the comment I'll re-write the loop for next version of
patches
_______________________________________________
Alsa-devel mailing list
Alsa-devel@xxxxxxxxxxxxxxxx
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
