On 10/04/2026 03:53 pm, Johannes Berg wrote:
> On Thu, 2026-09-24 at 08:10 +0530, Pooventhiran G wrote:
>> A UHR Link Reconfiguration Request frame (ST Preparation or Execution)
>> triggers SMD BSS Transition on the current AP MLD. Userspace needs
>> reporting of the STA's dynamic context along with such frames so that
>> the same can be transported to the target AP MLD for setting up the STA
>> TX and RX queues.
>>
>> Reserve a field in ieee80211_rx_status that enables drivers to attach
>> the STA's dynamic context to the corresponding frame. Since the maximum
>> possible context can grow too big, attach the pointer to the context to
>> the frame. Add handling for UHR ST Preparation and Execution Request
>> frames so that the associated context is propagated through cfg80211 and
>> nl80211 for userspace reporting.
>>
>> Signed-off-by: Pooventhiran G <[email protected]>
>> ---
>> include/linux/ieee80211-uhr.h | 59 ++++++++++++++++++++++++++
>> include/net/cfg80211.h | 13 ++++++
>> include/net/mac80211.h | 10 ++++-
>> net/mac80211/ieee80211_i.h | 2 +
>> net/mac80211/rx.c | 96
>> +++++++++++++++++++++++++++++--------------
>
> This change is big due to argument refactoring - better do that
> separately.
>
Sure, I will split up the patch.
> Also, I think better to combine the whole series a bit differently with
> cfg80211 APIs more isolated and then mac80211 implementation(s)
>
I will reorder the patches this way.
> One thing I'm missing completely - which is why I was looking for the
> rx.c changes - is handling of 4a substate frame limitations?
>
I will add the sub-state frame limitations in v3.
> johannes
