Hi, On Mon, Jan 30, 2023 at 4:55 AM Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > > Hi Alexander, > > aahringo@xxxxxxxxxx wrote on Thu, 26 Jan 2023 20:48:02 -0500: > > > Hi, > > > > On Thu, Jan 26, 2023 at 8:45 PM Alexander Aring <aahringo@xxxxxxxxxx> wrote: > > > > > > Hi, > > > > > > On Wed, Jan 25, 2023 at 5:31 AM Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > > > > > > > > Scanning being now supported, we can eg. play with hwsim to verify > > > > everything works as soon as this series including beaconing support gets > > > > merged. > > > > > > > > Thanks, > > > > Miquèl > > > > > > > > Changes in v2: > > > > * Clearly state in the commit log llsec is not supported yet. > > > > * Do not use mlme transmission helpers because we don't really need to > > > > stop the queue when sending a beacon, as we don't expect any feedback > > > > from the PHY nor from the peers. However, we don't want to go through > > > > the whole net stack either, so we bypass it calling the subif helper > > > > directly. > > > > > > > > moment, we use the mlme helpers to stop tx > > No, we no longer use the mlme helpers to stop tx when sending beacons > (but true MLME transmissions, we ack handling and return codes will be > used for other purposes). > then we run into an issue overwriting the framebuffer while the normal transmit path is active? > > but we use the > > ieee802154_subif_start_xmit() because of the possibility to invoke > > current 802.15.4 hooks like llsec? That's how I understand it. > > We go through llsec (see ieee802154_subif_start_xmit() implementation) > when we send data or beacons. When we send beacons, for now, we just > discard the llsec logic. This needs of course to be improved. We will > probably need some llsec handling in the mlme case as well in the near > future. > i agree, thanks. - Alex
