Re: [PATCHv3 1/3] mac80211: move mesh sync beacon handler into neighbour_update

[Marco Porsch <marco@cozybit.com>]

Am 26.02.2013 22:31, schrieb Thomas Pedersen:
> Hi Marco,
>
> On Mon, Feb 25, 2013 at 8:00 AM, Marco Porsch<marco@cozybit.com>  wrote:
>> Move the beacon handler into mesh_neighbour_update where the STA
>> pointer is already available. This avoids additional overhead
>> and simplifies the handler.
>> The repositioning will also benefit mesh PS which uses T_r and
>> the updated T_offset value.
>>
>> Check frame type before calling the handler. Rename it to better
>> reflect its purpose.
>>
>> Signed-off-by: Marco Porsch<marco@cozybit.com>
>> ---
>>
>> v3:
>>          check frame type before rx_bcn handler (Johannes)
>>
>>   net/mac80211/ieee80211_i.h |    9 ++++-----
>>   net/mac80211/mesh.c        |    7 +------
>>   net/mac80211/mesh.h        |    4 +++-
>>   net/mac80211/mesh_plink.c  |   38 ++++++++++++++++++++++++++++-------
>>   net/mac80211/mesh_sync.c   |   47 +++++++++-----------------------------------
>>   5 files changed, 48 insertions(+), 57 deletions(-)
>>
>> diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h
>> index 388580a..631760f 100644
>> --- a/net/mac80211/ieee80211_i.h
>> +++ b/net/mac80211/ieee80211_i.h
>> @@ -529,11 +529,10 @@ struct ieee80211_if_ibss {
>>    */
>>   struct ieee802_11_elems;
>>   struct ieee80211_mesh_sync_ops {
>> -       void (*rx_bcn_presp)(struct ieee80211_sub_if_data *sdata,
>> -                            u16 stype,
>> -                            struct ieee80211_mgmt *mgmt,
>> -                            struct ieee802_11_elems *elems,
>> -                            struct ieee80211_rx_status *rx_status);
>> +       void (*rx_bcn)(struct sta_info *sta,
>> +                      struct ieee80211_mgmt *mgmt,
>> +                      struct ieee802_11_elems *elems,
>> +                      u64 t_r);
>
> Why not continue to sync against probe responses?
> mesh_neighbour_update() gets these as well.

Because there was a comment about the standard mentioning only beacon 
frames. Fooled me here =/
Now I see that the standard explicitly mentions probe responses as well.
Is the sync code currently ready to handle probe responses? I am not 
perfectly sure about the sta->t_offset_setpoint stuff.

--Marco