[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAHp75VfnVzrVLUpXdspw1H-e4qxVzia+hf_2uJ5xgSm6Q0C7ug@mail.gmail.com>
Date: Thu, 23 Mar 2023 20:50:41 +0200
From: Andy Shevchenko <andy.shevchenko@...il.com>
To: Chester Lin <clin@...e.com>
Cc: Linus Walleij <linus.walleij@...aro.org>,
NXP S32 Linux Team <s32@....com>,
linux-gpio@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org,
Ghennadi Procopciuc <Ghennadi.Procopciuc@....nxp.com>,
Andrei Stefanescu <andrei.stefanescu@....com>,
Radu Pirea <radu-nicolae.pirea@....com>,
Andreas Färber <afaerber@...e.de>,
Matthias Brugger <mbrugger@...e.com>
Subject: Re: [PATCH v3 6/6] pinctrl: s32: separate const device data from
struct s32_pinctrl_soc_info
On Thu, Mar 23, 2023 at 4:49 PM Chester Lin <clin@...e.com> wrote:
>
> The .data field in struct of_device_id is used as a const member so it's
> inappropriate to attach struct s32_pinctrl_soc_info with of_device_id
> because some members in s32_pinctrl_soc_info need to be filled by
> pinctrl-s32cc at runtime.
>
> For this reason, struct s32_pinctrl_soc_info must be allocated in
> pinctrl-s32cc and then create a new struct s32_pinctrl_soc_data in order
> to represent const .data in of_device_id. To combine these two structures,
> a s32_pinctrl_soc_data pointer is introduced in s32_pinctrl_soc_info.
So, the first patch has to be embedded in this one, correct?
Don't forget to compile and test your contributions beforehand.
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists