Re: [PATCH 1/7] mac80211: add tracking of temporary offchannel sdata

[Michal Kazior <michal.kazior@tieto.com>]

Johannes Berg wrote:
> Hi Michal,
>
>> This is necessary if we want to have a sdata-based
>> channel recalculation.
>>
>> Change-Id: I223e052146893b3ae1ca46de7d90c54ffc589f1b
>> Signed-off-by: Michal Kazior<michal.kazior@tieto.com>
>> ---
>>   net/mac80211/ieee80211_i.h |    1 +
>>   net/mac80211/work.c        |    2 ++
>>   2 files changed, 3 insertions(+), 0 deletions(-)
>>
>> diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h
>> index b5e491b..c109960 100644
>> --- a/net/mac80211/ieee80211_i.h
>> +++ b/net/mac80211/ieee80211_i.h
>> @@ -997,6 +997,7 @@ struct ieee80211_local {
>>   	struct ieee80211_channel *oper_channel, *csa_channel;
>>
>>   	/* Temporary remain-on-channel for off-channel operations */
>> +	struct ieee80211_sub_if_data *tmp_sdata;
>>   	struct ieee80211_channel *tmp_channel;
>>   	enum nl80211_channel_type tmp_channel_type;
>
> Do you actually need this? I'm still tempted to not worry about any of
> this and force drivers to implement remain-on-channel in the driver or
> device for multi-channel, and this temporary thing is only used for
> remain-on-channel operations now.

No I don't. But if it's not done we break the whole thing (sw 
offchannel/scan) with my next patch. We might as well just remove all 
the sw offchannel/scan code.

How many drivers are there that still depend on sw offchannel? Can we 
just go ahead and break them?


-- 
Pozdrawiam / Best regards, Michal Kazior.