[net-next,v2,0/6] netdev event handling + neighbour config

Martin Schiller (6):
  net/x25: handle additional netdev events
  net/x25: make neighbour params configurable
  net/x25: replace x25_kill_by_device with x25_kill_by_neigh
  net/x25: support NETDEV_CHANGE notifier
  net/lapb: support netdev events
  net/lapb: fix t1 timer handling

 include/net/x25.h        |  10 +-
 include/uapi/linux/x25.h |  56 ++++++-----
 net/lapb/lapb_iface.c    |  83 ++++++++++++++++
 net/lapb/lapb_timer.c    |  11 ++-
 net/x25/af_x25.c         | 206 +++++++++++++++++++++++++++++++--------
 net/x25/x25_facilities.c |   6 +-
 net/x25/x25_link.c       | 142 +++++++++++++++++++++++----
 net/x25/x25_subr.c       |  22 ++++-
 8 files changed, 445 insertions(+), 91 deletions(-)

On Mon, Nov 16, 2020 at 6:00 AM Martin Schiller <ms@dev.tdt.de> wrote:
>
> This makes it possible to handle carrier lost and detection.
> In case of carrier lost, we shutdown layer 3 and flush all sessions.
>
> @@ -275,6 +275,19 @@ static int x25_device_event(struct notifier_block *this, unsigned long event,
>                                  dev->name);
>                         x25_link_device_remove(dev);
>                         break;
> +               case NETDEV_CHANGE:
> +                       pr_debug("X.25: got event NETDEV_CHANGE for device: %s\n",
> +                                dev->name);
> +                       if (!netif_carrier_ok(dev)) {
> +                               pr_debug("X.25: Carrier lost -> set link state down: %s\n",
> +                                        dev->name);
> +                               nb = x25_get_neigh(dev);
> +                               if (nb) {
> +                                       x25_link_terminated(nb);
> +                                       x25_neigh_put(nb);
> +                               }
> +                       }
> +                       break;
>                 }
>         }

I think L2 will notify L3 if the L2 connection is terminated. Is this
patch necessary?

On 2020-11-17 12:41, Xie He wrote:
> On Mon, Nov 16, 2020 at 6:00 AM Martin Schiller <ms@dev.tdt.de> wrote:
>> 
>> This makes it possible to handle carrier lost and detection.
>> In case of carrier lost, we shutdown layer 3 and flush all sessions.
>> 
>> @@ -275,6 +275,19 @@ static int x25_device_event(struct notifier_block 
>> *this, unsigned long event,
>>                                  dev->name);
>>                         x25_link_device_remove(dev);
>>                         break;
>> +               case NETDEV_CHANGE:
>> +                       pr_debug("X.25: got event NETDEV_CHANGE for 
>> device: %s\n",
>> +                                dev->name);
>> +                       if (!netif_carrier_ok(dev)) {
>> +                               pr_debug("X.25: Carrier lost -> set 
>> link state down: %s\n",
>> +                                        dev->name);
>> +                               nb = x25_get_neigh(dev);
>> +                               if (nb) {
>> +                                       x25_link_terminated(nb);
>> +                                       x25_neigh_put(nb);
>> +                               }
>> +                       }
>> +                       break;
>>                 }
>>         }
> 
> I think L2 will notify L3 if the L2 connection is terminated. Is this
> patch necessary?

Hmm... well I guess you're right. Admittedly, these patches were made
about 7 - 8 years ago and I have to keep thinking about them.
But I can't think of any situation where this patch should be necessary
at the moment.

I will drop this patch from the patch-set.