Re: [PATCH] iio: imu: kmx61: Use guard(mutex)() family over manual locking

[Jonathan Cameron <jic23@kernel.org>]

> ...
> 
> > -	mutex_lock(&data->lock);
> > +	guard(mutex)(&data->lock);
> >  	iio_for_each_active_channel(indio_dev, bit) {
> >  		ret = kmx61_read_measurement(data, base, bit);
> >  		if (ret < 0) {
> > -			mutex_unlock(&data->lock);
> > -			goto err;
> > +			iio_trigger_notify_done(indio_dev->trig);
> > +			return IRQ_HANDLED;  
> 
> Hmm... Is the HANDLED a right choice?

Ah. Long running discussion point on what right thing to do if you
are sure it's your interrupt but your handling failed - in a fashion
that doesn't prevent the interrupt clearing.  Not obvious :(

> 
> >  		}
> >  		buffer[i++] = ret;
> >  	}
> > -	mutex_unlock(&data->lock);
> >  
> >  	iio_push_to_buffers(indio_dev, buffer);
> > -err:
> >  	iio_trigger_notify_done(indio_dev->trig);
> >  
> >  	return IRQ_HANDLED;  
> 
> If the answer is yes, I'm wondering if we may deduplicate that...

Would require a helper function to avoid gotos + guard() in the same
function or some ugly code flow. Might not be worth it.

Jonathan
p.s. I deleted some more stuff from Andy so look at his reply.