On Wed, 02 Jun 2021, Robert Marko wrote:

On Tue, Jun 1, 2021 at 4:48 PM Lee Jones [off-list ref] wrote:

quoted

On Tue, 01 Jun 2021, Lee Jones wrote:

quoted

On Tue, 01 Jun 2021, Michael Walle wrote:

quoted

Am 2021-06-01 10:19, schrieb Lee Jones:

quoted

Why do you require one single Regmap anyway?  Are they register banks
not neatly separated on a per-function basis?

AFAIK you can only have one I2C device driver per device, hence the
simple-mfd-i2c.

Sorry, can you provide more detail.

I'd still like further explanation to be sure, but if you mean what I
think you mean then, no, I don't think that's correct.

The point of simple-mfd-i2c is to provide an I2C device offering
multiple functions, but does so via a non-separated/linear register-
set, with an entry point and an opportunity to register its interwoven
bank of registers via Regmap.

However, if you can get away with not registering your entire register
set as a single Regmap chunk, then all the better.  This will allow
you to use the OF provided 'simple-mfd' compatible instead.

Now, if you're talking about Regmap not supporting multiple
registrations with only a single I2C address, this *may* very well be
the case, but IIRC, I've spoken to Mark about this previously and he
said the extension to make this possible would be trivial.

This is my understanding, that you cannot have multiple regmap registrations
with on the same I2C address.
At least that is how it was the last time I tested.
That is why I went the MFD way.

I've just clarified with Mark.

There does not appear to be such a restriction.

-- 
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog