Re: [PATCH wpan-next v2 14/14] net: mac802154: Introduce a synchronous API for MLME commands

[Alexander Aring <alex.aring@gmail.com>]

Hi,

On Sun, Feb 20, 2022 at 6:52 PM Alexander Aring <alex.aring@gmail.com> wrote:
>
> Hi,
>
> On Mon, Feb 7, 2022 at 9:48 AM Miquel Raynal <miquel.raynal@bootlin.com> wrote:
> >
> > This is the slow path, we need to wait for each command to be processed
> > before continuing so let's introduce an helper which does the
> > transmission and blocks until it gets notified of its asynchronous
> > completion. This helper is going to be used when introducing scan
> > support.
> >
> > Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
> > ---
> >  net/mac802154/ieee802154_i.h | 1 +
> >  net/mac802154/tx.c           | 6 ++++++
> >  2 files changed, 7 insertions(+)
> >
> > diff --git a/net/mac802154/ieee802154_i.h b/net/mac802154/ieee802154_i.h
> > index 295c9ce091e1..ad76a60af087 100644
> > --- a/net/mac802154/ieee802154_i.h
> > +++ b/net/mac802154/ieee802154_i.h
> > @@ -123,6 +123,7 @@ extern struct ieee802154_mlme_ops mac802154_mlme_wpan;
> >  void ieee802154_rx(struct ieee802154_local *local, struct sk_buff *skb);
> >  void ieee802154_xmit_sync_worker(struct work_struct *work);
> >  void ieee802154_sync_and_stop_tx(struct ieee802154_local *local);
> > +void ieee802154_mlme_tx(struct ieee802154_local *local, struct sk_buff *skb);
> >  netdev_tx_t
> >  ieee802154_monitor_start_xmit(struct sk_buff *skb, struct net_device *dev);
> >  netdev_tx_t
> > diff --git a/net/mac802154/tx.c b/net/mac802154/tx.c
> > index 06ae2e6cea43..7c281458942e 100644
> > --- a/net/mac802154/tx.c
> > +++ b/net/mac802154/tx.c
> > @@ -126,6 +126,12 @@ void ieee802154_sync_and_stop_tx(struct ieee802154_local *local)
> >         atomic_dec(&local->phy->hold_txs);
> >  }
> >
> > +void ieee802154_mlme_tx(struct ieee802154_local *local, struct sk_buff *skb)
> > +{
> > +       ieee802154_tx(local, skb);
> > +       ieee802154_sync_and_stop_tx(local);
>
> Some of those functions can fail, in async case we can do some stats
> but here we can deliver the caller an error. Please do so.
>

to more specify what I mean here is also to get an error if the async
transmit path ends with the error case and not success. You need to
put a tx result int a global pointer to return it here and fetch it
before wake_up() "ieee802154_sync_and_stop_tx()".

Also I think we ignore here the case that the interface goes down
waiting for tx completion which means we are calling stop() callback
on the driver layer. I am worried that if stop() happens during
ieee802154_mlme_tx() we are running in a deadlock because a "tx
complete" will never occur?

- Alex