On Thu, 2023-04-06 at 18:47 +0530, Veerendranath Jakkam wrote:
> On 3/7/2023 3:52 PM, Johannes Berg wrote:
> > ---
> > So this is the patch I ended up with after some cleanups, but
> > then at the end I noticed you didn't unset the temp address
> > when the remain-on-channel expires, and wasn't sure exactly
> > how that should be handled, and you probably have the better
> > test setup right now too ...
> Apologies for the delayed response. Thanks for the cleaning up the patch.
>
> Regarding the comment on clearing temp address, I think the temp address
> should get cleared with below changes added in
> ieee80211_roc_notify_destroy() when remain-on-channel expires. Please
> let me know if I am missing something.
>
>
> > diff --git a/net/mac80211/offchannel.c b/net/mac80211/offchannel.c
> > index d78c82d6b696..74d2b826916a 100644
> > --- a/net/mac80211/offchannel.c
> > +++ b/net/mac80211/offchannel.c
> > @@ -187,6 +187,12 @@ static void ieee80211_roc_notify_destroy(struct ieee80211_roc_work *roc)
> > roc->mgmt_tx_cookie,
> > roc->chan, GFP_KERNEL);
> >
> > + if (!is_zero_ether_addr(roc->sdata->vif.cfg.temp_addr)) {
> > + eth_zero_addr(roc->sdata->vif.cfg.temp_addr);
> > + ieee80211_vif_cfg_change_notify(roc->sdata,
> > + BSS_CHANGED_TEMP_ADDR);
> > + }
> > +
> > list_del(&roc->list);
> > kfree(roc);
> > }
>
Huh, yeah, looks like I just missed that? Sorry about that.
Seeing that though, I remembered another thing - don't we have to adjust
the merging logic in ieee80211_start_roc_work() and maybe also
ieee80211_coalesce_hw_started_roc()?
johannes
