[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAFXKEHaaYXpcQ92ahWwuBbsCJYEzCi=tUSkpjWMkzhfrCOT_Tg@mail.gmail.com>
Date: Tue, 4 Feb 2025 15:13:34 +0100
From: Lothar Rubusch <l.rubusch@...il.com>
To: Jonathan Cameron <jic23@...nel.org>
Cc: lars@...afoo.de, Michael.Hennerich@...log.com, linux-iio@...r.kernel.org,
linux-kernel@...r.kernel.org, eraretuya@...il.com
Subject: Re: [PATCH v1 01/12] iio: accel: adxl345: migrate constants to core
Hi Jonathan,
On Sat, Feb 1, 2025 at 5:36 PM Jonathan Cameron <jic23@...nel.org> wrote:
>
> On Tue, 28 Jan 2025 12:00:49 +0000
> Lothar Rubusch <l.rubusch@...il.com> wrote:
>
> > The set of constants does not need to be exposed. Move constants to core
> > to reduce namespace polution.
> >
> > Signed-off-by: Lothar Rubusch <l.rubusch@...il.com>
> Hi Lothar,
>
>
> > -#define ADXL345_REG_INT_MAP 0x2F
> > -#define ADXL345_REG_INT_SOURCE 0x30
> > -#define ADXL345_REG_INT_SOURCE_MSK 0xFF
> > #define ADXL345_REG_DATA_FORMAT 0x31
>
> Normally I'd be entirely in favour of this, but...
> I'm not sure we want to leave one random register here
> and move the rest.
>
> Se can move the stuff that isn't register related, but
> for the registers I'd prefer to keep them in one place
> and I can't see a clean way to move them all to the core.c
> file. Even separating reg address and fields within it
> makes for a harder check against a datasheet etc.
>
> So I think all we can move is the fifo size :(
>
I understood that it could be one of the first follow up patches to move those
defines (parts of them?) over to core, as here in this mail:
https://lore.kernel.org/linux-iio/20241214123926.0b42ea59@jic23-huawei/
Anyway, I already had presented moving the constants before, when you
had decided to keep them in the header. I thought you changed your mind
on that, but I don't want to bother you with the same issue over and over
again, probably I missunderstood that here.
I leave the constants in the .h file then, no problem. :) I can understand the
intention to keep the things rather together in one place. There seem to be
pros & cons for both.
>
> > -#define ADXL345_REG_XYZ_BASE 0x32
> > -#define ADXL345_REG_DATA_AXIS(index) \
> > - (ADXL345_REG_XYZ_BASE + (index) * sizeof(__le16))
>
Powered by blists - more mailing lists