Re: [PATCH v2 0/6] gpio: cdev: bail out of poll() if the device goes down

[Bartosz Golaszewski <brgl@xxxxxxxx>]

On Fri, Aug 18, 2023 at 3:29 PM Andy Shevchenko
<andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
>
> On Fri, Aug 18, 2023 at 02:06:21PM +0200, Bartosz Golaszewski wrote:
> > On Fri, Aug 18, 2023 at 12:34 PM Andy Shevchenko
> > <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> > >
> > > On Thu, Aug 17, 2023 at 08:49:52PM +0200, Bartosz Golaszewski wrote:
> > > > From: Bartosz Golaszewski <bartosz.golaszewski@xxxxxxxxxx>
> > > >
> > > > Wake up all three wake queues (the one associated with the character
> > > > device file, the one for V1 line events and the V2 line request one)
> > > > when the underlying GPIO device is unregistered. This way we won't get
> > > > stuck in poll() after the chip is gone as user-space will be forced to
> > > > go back into a new system call and will see that gdev->chip is NULL.
> > >
> > > Why can't you use the global device unbind notifications and filter out
> > > what you are interested in?
> >
> > There's no truly global device unbind notification - only per-bus.
> > GPIO devices can reside on any bus, there are no limitations and so
> > we'd have to subscribe to all of them.
>
> We have, but it requires a bit of code patching.
> Look at device_platform_notify()/device_platform_notify_remove().
>
> I noticed, btw, that platform_notify() and Co is a dead code :-)
> Maybe it can be converted to a list and a manager of that list,
> so specific cases can utilize it.
>

That's not going to happen anytime soon and I doubt Greg would like
the idea without more users interested in receiving these events. :(

This GPIO notifier is an implementation detail - once we do have
global notifiers, we can switch to using them instead.

Bart