Re: [PATCH] staging: iio: adt7316: replace sprintf() with sysfs_emit()

["Nuno Sá" <noname.nuno@gmail.com>]

On Wed, 2025-11-19 at 14:57 +0000, Jonathan Cameron wrote:
> On Tue, 18 Nov 2025 14:10:25 +0200
> Andy Shevchenko <andy.shevchenko@gmail.com> wrote:
> 
> > On Tue, Nov 18, 2025 at 10:17 AM Nuno Sá <noname.nuno@gmail.com> wrote:
> > > On Sun, 2025-11-16 at 15:28 +0000, Jonathan Cameron wrote:  
> > > > On Sun, 16 Nov 2025 13:08:07 +0200
> > > > Andy Shevchenko <andy.shevchenko@gmail.com> wrote:  
> > 
> > ...
> > 
> > > > Applied, but I will note that this driver is a long way from suitable
> > > > for moving out of staging and I would be surprised if more than
> > > > one or two of the lines changed in this patch make it through the
> > > > necessary refactors (should anyone actually have another go at
> > > > doing them).  Anyhow, I still think this is worth taking just to
> > > > reduce the noise of remaining instances of this.
> > > > 
> > > > I'd have deleted this long ago except I actually have one somewhere
> > > > and it one of the supported parts was still listed as suitable for
> > > > new designs when I checked not long ago.  
> > > 
> > > Seems to be now "just" in production. So no longer advised for new designs. I gave a
> > > quick look and this is far from being a simple driver. It would require a fair amount
> > > of work to bring it out of staging. So, are there users relying on staging drivers?  
> > 
> > If there is no choice, yes. For example FBTFT is famous for being used
> > by IoT / DIY _a lot_. Currently a bit less since we got tinyDRM
> > subsystem with most popular drivers reimplemented there.
> > 
> > > But I would assume that for staging we are still free to drop support it?  
> > 
> > At any level, we are still free to drop :-) Just different conditions
> > applied. For staging is basically an "effective immediately" case.
> > 
> 
> If no one replies to this thread in next few weeks to say otherwise, let us see
> if anyone objects to patches. I'll spin a series in a few weeks and they can sit
> in linux-next for most of the cycle to see anyone notices.  It is easy to bring
> drivers back if turns out anyone cares and then there is some motivation to
> finally clean them up.
> 
> staging/iio/
> 	accel/adis16203 is production
> 	adc/ad7816 is production (7817, 7818) - the 7816 itself is obsolete.
> 	addac/adt7316 is production (what we are discussing here)

I checked internally and we still have customers actually buying some of the chips supported by the
above driver. Not sure if it is being used with linux though... And more importantly, I cannot also
commit that someone will jump in to clean it up.

- Nuno Sá