On 27/03/2023 12:19, Ping-Ke Shih wrote: > > >> -----Original Message----- >> From: Kalle Valo <kvalo@xxxxxxxxxx> >> Sent: Monday, March 27, 2023 4:42 PM >> To: Ping-Ke Shih <pkshih@xxxxxxxxxxx> >> Cc: Martin Kaistra <martin.kaistra@xxxxxxxxxxxxx>; linux-wireless@xxxxxxxxxxxxxxx; Jes Sorensen >> <Jes.Sorensen@xxxxxxxxx>; Bitterblue Smith <rtl8821cerfe2@xxxxxxxxx>; Sebastian Andrzej Siewior >> <bigeasy@xxxxxxxxxxxxx> >> Subject: Re: [RFC PATCH 07/14] wifi: rtl8xxxu: Add parameter macid to update_rate_mask >> >> Ping-Ke Shih <pkshih@xxxxxxxxxxx> writes: >> >>>> -----Original Message----- >>>> From: Martin Kaistra <martin.kaistra@xxxxxxxxxxxxx> >>>> Sent: Thursday, March 23, 2023 1:19 AM >>>> To: linux-wireless@xxxxxxxxxxxxxxx >>>> Cc: Jes Sorensen <Jes.Sorensen@xxxxxxxxx>; Kalle Valo <kvalo@xxxxxxxxxx>; Ping-Ke Shih >>>> <pkshih@xxxxxxxxxxx>; Bitterblue Smith <rtl8821cerfe2@xxxxxxxxx>; Sebastian Andrzej Siewior >>>> <bigeasy@xxxxxxxxxxxxx> >>>> Subject: [RFC PATCH 07/14] wifi: rtl8xxxu: Add parameter macid to update_rate_mask >>>> >>>> The HW maintains a rate_mask for each connection, referenced by the >>>> macid. Add a parameter to update_rate_mask and add the macid to the >>>> h2c call in the gen2 implementation. >>>> >>>> Also extend refresh_rate_mask to generate the macid in AP mode from >>>> sta->aid. >>> >>> Firmware can support 32 mac_id (station instance) at most, so it will be a >>> problem if hostapd assigns aid more than 32. Though I'm not clear how >>> hostpad assigns the aid, it would be always safe that rtl8xxxu maintains >>> mac_id by a bitmap in driver. >> >> Does rtlw8xxxu set struct wiphy::max_ap_assoc_sta? It would be good to >> advertise the user space the maximum number of stations. >> > > Thanks for this information, Kalle. > > Martin, please add this. I think we can preserve at least one mac_id for > broadcast/multicast frames. In fact, I'm not absolutely sure we can > support up to 32 mac_id, so set wiphy::max_ap_assoc_sta = 16 -1 or -2 > would be safer. > > Ping-Ke > > Indeed, the RTL8188FU driver has hal_spec->macid_num = 16. I think that's the maximum for this chip. RTL8710BU: 16 RTL8188EU: 64 RTL8192EU: 128 RTL8723BU: 128
