linux-kernel - Re: [PATCH 2/2] mfd: add regulator cell to Cherry Trail Whiskey Cove PMIC

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:   Fri, 25 Oct 2019 11:16:22 +0200
From:   Andrey Zhizhikin <andrey.z@...il.com>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Cc:     lgirdwood@...il.com, broonie@...nel.org, lee.jones@...aro.org,
        linux-kernel@...r.kernel.org,
        Andrey Zhizhikin <andrey.zhizhikin@...ca-geosystems.com>
Subject: Re: [PATCH 2/2] mfd: add regulator cell to Cherry Trail Whiskey Cove PMIC

On Fri, Oct 25, 2019 at 10:07 AM Andy Shevchenko
<andriy.shevchenko@...ux.intel.com> wrote:
>
> On Thu, Oct 24, 2019 at 02:29:39PM +0000, Andrey Zhizhikin wrote:
> > Add a regulator mfd cell to Whiskey Cove PMIC driver, which is used to
> > supply various voltage rails.
> >
> > In addition, make the initialization of this mfd driver early enough in
> > order to provide regulator cell to mmc sub-system when it is initialized.
>
> Doesn't deferred probe mechanism work for you?
> MMC core returns that error till we have the driver initialized.

This would work for mmc sub-system, but my idea was that later when
more cells are added to this mfd - it might turn out that we would
require an early initialization anyway. So I decided to take an
opportunity to adjust it with this patch as well, and this is what I
roughly explained in the commit message. When I'm reading it now,
exactly this point was not mentioned in commit message at all, and I
rather coupled the early init with mmc sub-system, which creates a
source of confusion here. I guess if there would be no other
objections about early init - I'd go with v2 of this patch, where I
would clean-up the point below and adjust the commit description.

Thanks a lot for pointing this out!

>
> > +     }, {
> > +             .name = "cht_wcove_regulator",
> > +             .id = CHT_WC_REGULATOR_VSDIO + 1,
>
> > +             .num_resources = 0,
> > +             .resources = NULL,
>
> No need to put these.

Agreed, this was forgotten. Sorry for that.

>
> > -static struct i2c_driver cht_wc_driver = {
> > +static struct i2c_driver cht_wc_i2c_driver = {
>
> Renaming is not explained in the commit message.

True, this point I forgot to mention. Actually, this is tightly
coupled with the fact that mfd driver has been moved to an earlier
init stage and since it does belong to I2C sub-system (and represented
by i2c_device structure) - I decided to make a name sound more
logical.

>
> --
> With Best Regards,
> Andy Shevchenko
>
>

-- 
Regards,
Andrey.