On Mon, Aug 21, 2023 at 2:43 PM Bartosz Golaszewski <brgl@xxxxxxxx> wrote: > > On Fri, Aug 18, 2023 at 12:31 PM Andy Shevchenko > <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote: > > > > On Thu, Aug 17, 2023 at 08:49:56PM +0200, Bartosz Golaszewski wrote: > > > From: Bartosz Golaszewski <bartosz.golaszewski@xxxxxxxxxx> > > > > > > Add a notifier block to the gpio_chardev_data structure and register it > > > with the gpio_device's device notifier. Upon reception of an event, wake > > > up the wait queue so that the user-space be forced out of poll() and need > > > to go into a new system call which will then fail due to the chip being > > > gone. > > > > ... > > > > > + struct gpio_chardev_data *cdev = container_of(nb, > > > + struct gpio_chardev_data, > > > + device_unregistered_nb); > > > > struct gpio_chardev_data *cdev = > > container_of(nb, struct gpio_chardev_data, device_unregistered_nb); > > > > ? > > I could live with the other version but sure, why not. > > I will send a v3 with a helper wrapper around > blocking_notifier_call_chain() for more brevity. > Scratch that, I need coffee. This was supposed to go under the line state notifications patch. Bart