Re: [PATCH] brcmfmac: support AP isolation to restrict reachability between stations

[Arend van Spriel <arend.vanspriel@broadcom.com>]

On 4/24/2025 12:22 PM, Johannes Berg wrote:
> On Thu, 2025-04-24 at 11:50 +0200, Arend van Spriel wrote:
>>
>> Looked at other drivers implementing this callback and here are the results:
>>
>> [wil6210] wil_cfg80211_change_bss(): does exactly the same thing.
>> [wilc1000] change_bss(): worse! it accepts everything and does nothing.
>> [rtl8723bs] cfg80211_rtw_change_bss(): same. just an empty callback.
> 
> OK, though I guess other drivers being bad doesn't mean this one should
> be :)

Sure. I am on your team in this. Can you recommend a plan of attack 
here? Should we add a mechanism to expose what BSS parameter changes the 
driver can handle similar to what is used for struct 
station_info::bss_params?

>> If this patch means Infineon is (mildly) regaining interest in upstream
>> wifi development let's not discourage them.
> 
> Fair. I didn't mean to discourage. I just think to meaningfully
> contribute upstream people should follow the list. And even review other
> people's patches. I've been meaning to make that more of a requirement,
> since I can't possibly meaningfully review everything I now need to
> merge.
> 
> https://lore.kernel.org/linux-wireless/21896d2788b8bc6c7fcb534cd43e75671a57f494.camel@sipsolutions.net/

Yeah. It's a bit of a tough spot I reckon.

>> I do watch the
>> linux-wireless list on occasion but I am a bit lost on your remark. What
>> do you mean by tree tags. You mean the "wifi:" prefix? But then I am
>> confused about the "subject prefix" remark.
> 
> Oh, well I guess terminology:

Ack.

Regards,
Arend