Re: [PATCH] w1-gpio: handle of_get_gpio() returning -EPROBE_DEFER better

["Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>]

On Tue, Feb 18, 2014 at 05:26:20AM +0400, zbr@ioremap.net wrote:
> Hi
> 
> 14.02.2014, 02:05, "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>:
> > of_get_gpio() might return -EPROBE_DEFER meaning that the driver
> > providing the gpio isn't ready yet. If that happens for the first gpio
> > the resulting kernel output without this patch is:
> >
> >         w1-gpio somename: Failed to parse DT
> >         platform somename: Driver w1-gpio requests probe deferral
> >
> > The first message is misleading and so is suppressed with this patch.
> >
> > Further if determining the gpio to switch the external pullup yields
> > -EPROBE_DEFER this error should be passed back to the caller instead of
> > just continuing without pullup.
> >
> > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> 
> I'm ok with this patch, but I virtually do not understand what it does :)
> Since I know nothing about device trees.
The thing is that the w1-gpio device might be probed before the gpio's
controller is available. In this case of_get_gpio returns -EPROBE_DEFER
and this makes the device core retry probing the w1-gpio device later
again. So this is not a dt-parsing problem but just dependency stuff.

The 2nd issue addressed didn't happen to me, but it might occur that the
first of_get_gpio for the data pin succeeds, but the getting the pullup
gpio fails with -EPROBE_DEFER (e.g. because it sits on a different
controller). In this case you still want to retry probing later instead
of ignoring the problem.
 
Does that make it clearer?

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |