Re: [wpan-next v2 18/27] net: mac802154: Handle scan requests

[Miquel Raynal <miquel.raynal@bootlin.com>]

Hi Alexander,

alex.aring@gmail.com wrote on Wed, 12 Jan 2022 17:44:02 -0500:

> Hi,
> 
> On Wed, 12 Jan 2022 at 12:33, Miquel Raynal <miquel.raynal@bootlin.com> wrote:
> ...
> > +       return 0;
> > +}
> > diff --git a/net/mac802154/tx.c b/net/mac802154/tx.c
> > index c829e4a75325..40656728c624 100644
> > --- a/net/mac802154/tx.c
> > +++ b/net/mac802154/tx.c
> > @@ -54,6 +54,9 @@ ieee802154_tx(struct ieee802154_local *local, struct sk_buff *skb)
> >         struct net_device *dev = skb->dev;
> >         int ret;
> >
> > +       if (unlikely(mac802154_scan_is_ongoing(local)))
> > +               return NETDEV_TX_BUSY;
> > +  
> 
> Please look into the functions "ieee802154_wake_queue()" and
> "ieee802154_stop_queue()" which prevent this function from being
> called. Call stop before starting scanning and wake after scanning is
> done or stopped.

Mmmh all this is already done, isn't it?
- mac802154_trigger_scan_locked() stops the queue before setting the
  promiscuous mode
- mac802154_end_of_scan() wakes the queue after resetting the
  promiscuous mode to its original state

Should I drop the check which stands for an extra precaution?


But overall I think I don't understand well this part. What is
a bit foggy to me is why the (async) tx implementation does:

*Core*                           *Driver*

stop_queue()
drv_async_xmit() -------
                        \------> do something
                         ------- calls ieee802154_xmit_complete()
wakeup_queue() <--------/

So we actually disable the queue for transmitting. Why??

> Also there exists a race which exists in your way and also the one
> mentioned above. There can still be some transmissions going on... We
> need to wait until "all possible" tx completions are done... to be
> sure there are really no transmissions going on. However we need to be
> sure that a wake cannot be done if a tx completion is done, we need to
> avoid it when the scan operation is ongoing as a workaround for this
> race.
> 
> This race exists and should be fixed in future work?

Yep, this is true, do you have any pointers? Because I looked at the
code and for now it appears quite unpractical to add some kind of
flushing mechanism on that net queue. I believe we cannot use the netif
interface for that so we would have to implement our own mechanism in
the ieee802154 core.

Thanks,
Miquèl