[PATCH v2 libinput 1/2] evdev: Ensure the libevdev object receives the new fd on resume

[Hans de Goede]

Hi,

On 10-12-14 02:47, Peter Hutterer wrote:
> From: Carlos Garnacho <carlosg at gnome.org>
>
> Otherwise, input_events will be attempted to read from the wrong place,
> which also leaves the right/current fd with pending data to be read,
> making the epoll fd wake up constantly.
>
> Signed-off-by: Carlos Garnacho <carlosg at gnome.org>
> Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
> Signed-off-by: Peter Hutterer <peter.hutterer at who-t.net>

Series looks good and is:

Reviewed-by: Hans de Goede <hdegoede at redhat.com>

Regards,

Hans

> ---
> Changes to v1:
> - sync libevdev's state after change_fd
>
>   src/evdev.c | 16 ++++++++++++++++
>   1 file changed, 16 insertions(+)
>
> diff --git a/src/evdev.c b/src/evdev.c
> index 0873c99..3f15560 100644
> --- a/src/evdev.c
> +++ b/src/evdev.c
> @@ -1985,6 +1985,8 @@ evdev_device_resume(struct evdev_device *device)
>   	struct libinput *libinput = device->base.seat->libinput;
>   	int fd;
>   	const char *devnode;
> +	struct input_event ev;
> +	enum libevdev_read_status status;
>
>   	if (device->fd != -1)
>   		return 0;
> @@ -2011,6 +2013,20 @@ evdev_device_resume(struct evdev_device *device)
>   			return -ENODEV;
>   	}
>
> +	libevdev_change_fd(device->evdev, fd);
> +	libevdev_set_clock_id(device->evdev, CLOCK_MONOTONIC);
> +
> +	/* re-sync libevdev's view of the device, but discard the actual
> +	   events. Our device is in a neutral state already */
> +	libevdev_next_event(device->evdev,
> +			    LIBEVDEV_READ_FLAG_FORCE_SYNC,
> +			    &ev);
> +	do {
> +		status = libevdev_next_event(device->evdev,
> +					     LIBEVDEV_READ_FLAG_SYNC,
> +					     &ev);
> +	} while (status == LIBEVDEV_READ_STATUS_SYNC);
> +
>   	device->source =
>   		libinput_add_fd(libinput, fd, evdev_device_dispatch, device);
>   	if (!device->source) {
>