On Wed, Oct 08, 2014 at 10:50:00AM +0200, Linus Walleij wrote:

On Wed, Oct 8, 2014 at 9:05 AM, Alexandre Courbot [off-list ref] wrote:

quoted

But transition to gpiod is the way to ultimately fix this issue, as
well as many others. Not to mention that renumbering GPIOs will
certainly make a few users of the GPIO sysfs (another abomination,
agreed) unhappy. I can only recommend switching drivers to gpiod when
such issues are spotted.

Yeah that is another issue ... we end up in catch 22 situations like
that, renumber the GPIOs, OK, then we break the ABI.

Admittedly that "ABI" is something people break all the time,
/sys/*gpioN just isnt what it should be, not stable at all.

I'm not against renumbering GPIO if it's minor effort, but if it
start to consume hundreds of hours and regressions and what not,
that time is better spent focusing on the gpiod transition.

My guess is that it's relatively little work for most platforms - with
the systems I've done enough work on to notice everything is keyed off a
few defines in the header file.  Things tend to be worse in out of tree
code but mainline's generally been pretty good.