Re: [alsa-devel] [PATCH v3 1/1] ASoC: mxs-saif: add record function

...

I think we should use usleep_range for both udelays here? Having a rate of 8000, we'd burn 250us here.

Yes, I agree that it's a bit long for 8000. I tried sleep way but I found the trigger function is called with spin_lock held, so it seems we may not be able to sleep here.

I think the way of dynamically calculate delay suggested by Liam has Already minimize the affection, especially for high sample rate, it may work more efficiency than sleep (context switch cost).

Do you think if it's reasonable to accept it?

Yes, it can be fixed when it becomes necessary

Reviewed-by: Wolfram Sang w.sang@pengutronix.de