Potential Broadcast Issues After GTK Key Exchange on ath11k with IPQ8072A (QCN5024/QCN5054)

[Steffen Moser]

Hi,

I’m not a kernel developer myself, but from a formal perspective, it 
looks good to me. Unfortunately, I don't really know what the patch was 
for originally. Maybe reverting it brakes something.

Best regards,
Steffen


On 17.01.25 8:17 PM, Nicolas Escande wrote:
> On Thu Jan 9, 2025 at 2:25 PM CET, Kalle Valo wrote:
>> "Nicolas Escande" <nico.escande at gmail.com> writes:
>>
>> > On Sat Dec 28, 2024 at 11:13 AM CET, Steffen Moser wrote:
>> >> Dear Remi,
>> >>
>> >> thank you very much for the pointer to the patch. Sebastian integrated 
>> >> it into DD-WRT. Now the DynaLink DL-WRX36 runs absolutely smoothly, the 
>> >> WLAN links are stable, the packet loss is gone. No weird states anymore, 
>> >> independent from the group key exchange interval:
>> >>
>> >> https://data.saps.uni-ulm.de/index.php/s/NLNpWqjc8iGsaEM
>> >>
>> >> So your idea was a direct hit! Thank you very, very much. Several months 
>> >> of debugging have come to an end...
>> >
>> > So this is at least the second time this commit breaks a setup.
>> >  
>> > @ath11k why isn't this pushed to mainline ?
>> > This seems to be a clear regression, so even if there is no need to rush things
>> > in the long run this still needs to to reverted mainline right ?
>>
>> Good question, I don't have an answer to that. Could someone (also
>> outside of Qualcomm) send a proper patch ASAP so that we can solve this?
>> And it's good ot include the link to this discussion and describe the
>> symptoms the revert is fixing.
>>
>> This is the patch in question:
>>
>> https://git.codelinaro.org/clo/qsdk/oss/system/feeds/wlan-open/-/blob/win.wlan_host_opensource.3.0.r24/patches/ath11k/350-ath11k-Revert-clear-the-keys-properly-when-DISABLE_K.patch
> 
> Not sure if I did it the right way but it at least is here :
> https://lore.kernel.org/linux-wireless/20250117191455.3395145-1-nico.escande@gmail.com/
>