Re: [PATCH] mfd: rohm-bdXXX - switch to use i2c probe_new

From: Lee Jones
Date: Thu Apr 16 2020 - 02:31:03 EST


On Wed, 15 Apr 2020, Vaittinen, Matti wrote:

>
> On Wed, 2020-04-15 at 12:04 +0100, Lee Jones wrote:
> > On Wed, 15 Apr 2020, Vaittinen, Matti wrote:
> > > Hello Lee,
> > >
> > > On Wed, 2020-04-15 at 10:50 +0100, Lee Jones wrote:
> > > > On Thu, 26 Mar 2020, Matti Vaittinen wrote:
> > > >
> > > > > ROHM BD70528 and BD718x7 drivers do not utilize the I2C id.
> > > > > Do the trivial conversion and make them to use probe_new
> > > > > instead of probe.
> > > >
> > > > Not sure I understand the purpose of the patch.
> > > >
> > > > The only reason to switch to probe_new is to aid the removal of
> > > > the
> > > > compulsory I2C tables. However, neither of these drivers have
> > > > them.
> > >
> > > Maybe I have misunderstood the probe_new. My understanding was that
> > > the
> > > probe_new is what should become the standard - Eg. eventually all
> > > I2C
> > > drivers would use probe which is not getting the ID tables as
> > > argument
> > > - and the old probe could be removed. Thus I thought conversion of
> > > probe to probe_new would be required in order to get rid of the old
> > > probe. I think I got this understanding when I submitted driver for
> > > BD71828 - and I think it was you who suggested me to switch to
> > > probe_new as 'id' was unused. But if please just drop this patch if
> > > this change is not needed!
> >
> > Some I2C drivers do make use of 'id' though. So the standard probe
> > will probably never go away. I wrote probe_new for drivers that
> > wished to omit the I2C table. A better use of your time would be to
> > go through the original probe users and figure out which I2C tables
> > can be removed.
>
> Right. Unfortunately I don't have problems with excessive amounts of
> free time xD I did this probe-conversion for bd718x7 because that
> driver is originally written by me - and I try to contribute on
> maintenance of stuff I have pushed on your and other maintainer's
> shoulders :)
>
> And as I said - if the usage of probe is just as fine as usage of
> probe_new - then please just drop this patch :)

I should revisit the I2C situation at one point. It does seem silly
to have a 4 year old *new* API.

--
Lee Jones [æçæ]
Linaro Services Technical Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog