Re: [RFC PATCH v2 3/7] i2c: core: introduce callbacks for atomic transfers

[Simon Horman]

Hi Wolfram,

On Wed, Mar 27, 2019 at 02:47:51PM +0100, Wolfram Sang wrote:
> Hi Simon,
> 
> please delete unrelated text. I nearly missed the typo fix later.

Ack, will do.

> > > -	if (in_atomic() || irqs_disabled())
> > > +	if (in_atomic() || irqs_disabled()) {
> > > +		WARN(!adap->algo->master_xfer_atomic && !adap->algo->smbus_xfer_atomic,
> > > +		     "No atomic I2C transfer handler for '%s'\n", dev_name(&adap->dev));
> > 
> > 		Is WARN_ONCE more appropriate here?
> 
> Why? It could be multiple adapters or clients causing this?

Good point. I am concerned about the presence of a case which
causes a logging storm.

...