linux-kernel - Re: [PATCH v4 4/5] mfd: rn5t618: add more subdevices

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 17 Dec 2019 20:26:45 +0100
From:   Andreas Kemnade <andreas@...nade.info>
To:     Lee Jones <lee.jones@...aro.org>
Cc:     Discussions about the Letux Kernel <letux-kernel@...nphoenux.org>,
        mark.rutland@....com, a.zummo@...ertech.it,
        alexandre.belloni@...tlin.com, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org, stefan@...er.ch, b.galvani@...il.com,
        robh+dt@...nel.org, phh@....me, linux-rtc@...r.kernel.org
Subject: Re: [PATCH v4 4/5] mfd: rn5t618: add more subdevices

On Tue, 17 Dec 2019 12:25:26 +0100
Andreas Kemnade <andreas@...nade.info> wrote:

> On Tue, 17 Dec 2019 08:01:46 +0000
> Lee Jones <lee.jones@...aro.org> wrote:
> 
> > On Mon, 16 Dec 2019, Andreas Kemnade wrote:
> >   
> > > On Mon, 16 Dec 2019 15:31:06 +0000
> > > Lee Jones <lee.jones@...aro.org> wrote:
> > >     
> > > > On Wed, 11 Dec 2019, Andreas Kemnade wrote:
> > > >     
> > > > > The RC5T619 has a RTC which is missing in the
> > > > > RN5T618. Add it as subdevice to prepare for their implementation
> > > > > 
> > > > > Signed-off-by: Andreas Kemnade <andreas@...nade.info>
> > > > > ---
> > > > > Changes in v3:
> > > > > - alignment cleanup
> > > > >  drivers/mfd/rn5t618.c | 16 ++++++++++++++--
> > > > >  1 file changed, 14 insertions(+), 2 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/mfd/rn5t618.c b/drivers/mfd/rn5t618.c
> > > > > index d78eb29b94a4..18d56a732b20 100644
> > > > > --- a/drivers/mfd/rn5t618.c
> > > > > +++ b/drivers/mfd/rn5t618.c
> > > > > @@ -22,6 +22,12 @@ static const struct mfd_cell rn5t618_cells[] = {
> > > > >  	{ .name = "rn5t618-wdt" },
> > > > >  };
> > > > >  
> > > > > +static const struct mfd_cell rc5t619_cells[] = {
> > > > > +	{ .name = "rn5t618-regulator" },
> > > > > +	{ .name = "rc5t619-rtc" },
> > > > > +	{ .name = "rn5t618-wdt" },
> > > > > +};
> > > > > +
> > > > >  static bool rn5t618_volatile_reg(struct device *dev, unsigned int reg)
> > > > >  {
> > > > >  	switch (reg) {
> > > > > @@ -173,8 +179,14 @@ static int rn5t618_i2c_probe(struct i2c_client *i2c,
> > > > >  		return ret;
> > > > >  	}
> > > > >  
> > > > > -	ret = devm_mfd_add_devices(&i2c->dev, -1, rn5t618_cells,
> > > > > -				   ARRAY_SIZE(rn5t618_cells), NULL, 0, NULL);
> > > > > +	if (priv->variant == RC5T619)
> > > > > +		ret = devm_mfd_add_devices(&i2c->dev, -1, rc5t619_cells,      
> > > > 
> > > > Ref: The "-1", please use this as an opportunity to use the defines.
> > > >     
> BTW: of course I can clean that up.
> 
> > > > > +					   ARRAY_SIZE(rc5t619_cells),
> > > > > +					   NULL, 0, NULL);
> > > > > +	else      
> > > > 
> > > > Are you sure it's not possible for 'variant' to be an unsupported
> > > > device?
> > > >     
> > > Well, does it change the behavior for devices other than the rc5t619?
> > > I do not think so. If the mfd driver is bound to unsupported devices,
> > > rn5t618_of_match is wrong.    
> > 
> > Right, and can you catch that?
> >   
> Well, maybe we don't get each other. RC5T619 has an RTC.
> If I understand the code right, priv->variant is set to RC5T619.
> if there is compatible = "ricoh,rc5t619" in the device tree.
> So in that and only in that case I have an extended subdevice list,
> which includes the RTC subdevice instead of only regulator and wdt.
> For everything else I do not touch it.
> 
Hmm, if the driver might be probed without device tree node,
there is:
   of_id = of_match_device(rn5t618_of_match, &i2c->dev);
        if (!of_id) {
                dev_err(&i2c->dev, "Failed to find matching DT ID\n");
                return -EINVAL;
        }

so nothing will be registered at all. So the only reasons I see the
driver might be bound to unsupported devices could be:
- a bug in of_match_device() 
- nonsense in rn5t618_of_match (I think that are well-thought entries)

Regards,
Andreas