Re: rn5t618 vs. rn5t568

[Andreas Kemnade <andreas@kemnade.info>]

Hi,

On Tue, 02 May 2023 10:37:08 +0200
Juergen Borleis <jbe@pengutronix.de> wrote:

> Hi,
> 
> Am Donnerstag, dem 27.04.2023 um 09:36 +0200 schrieb Sascha Hauer:
> > On Wed, Apr 26, 2023 at 09:09:36PM +0200, Andreas Kemnade wrote:  
> > > Hi,
> > > 
> > > just looked a bit around in the source and stumbled across drivers/mfd/rn5t568.c
> > > 
> > > It contains
> > >  * MFD core driver for Ricoh RN5T618 PMIC
> > > 
> > > but the filename is rn5t568.c That is very confusing.
> > >  a compatible ricoh,rn5t568 (but 567) does not exist in the kernel.  
> 
> 567 and 568 are pin compatible. As far as I know the 568 has some fixed OTP
> settings which are used after reset. Depending on the type extension (C, S, AD…)
> the I²C address differs and the default output voltages.
> 
> > > So which chip this is really about? Did not analyse things deeper yet.
> > > 
> > > I am also curious because I wrote most stuff in the kernel about the rn5t618/rc5t619.  
> > 
> > Jürgen had to support the rn5t568 which is not yet supported in Linux.
> > He took the Kernel rn5t618.c driver and adjusted it accordingly. Jürgen
> > can't remember exactly, but he said the chips are quite similar.  
> 
> Due to the support of the 567 variant it is already supported in the kernel.
> 
hmm, but if you do not share dt compatibles, you need to maintain different
devicetrees in kernel vs. barebox. I thought you try to keep them in sync as
much a possible. Also in mainline kernel the watchdog does not have its own
device tree node.

Regards,
Andreas