Hi Michael, On Mon, 2024-04-22 at 02:35 +0000, Michael-cy Lee (李峻宇) wrote: > On Wed, 2024-04-17 at 09:06 +0200, Benjamin Berg wrote: > > On Tue, 2024-04-09 at 14:50 +0800, Michael-CY Lee wrote: > > > MLD STA finds MLD AP's partner links by BSSID from the scan results. > > > However, if the scan results contain BSSs with same BSSID but different > > > BSS information, the MLD STA might assign a wrong BSS to one of the MLD > > > AP's partner links. > > > > As a clarification, why/how do we get into the situation that the > > same BSSIDs exists with the wrong SSID in the supplicant? > > > > Did the AP change its SSID and we still have the old one in the > > cache? > > Yes, that's exactly the situation we've faced. > To be more precise, after we changed the SSID and restarted the AP MLD, > we triggered the STA scan and saw the STA's scan results containing two > BSSs with the same BSSID but different SSIDs, then the STA incorrectly > matched one of the AP MLD's links to the wrong BSS (old SSID). > > > > > If that is the case, then maybe what we should be doing is to verify > > that the BSS parameter change count matches. If not, it makes sense to > > reject the BSS for now, which should trigger an ML probe request to > > update the information and the next attempt will work fine. > > We think the BSS parameter change count (BPCC) might not be helpful > here since a change in SSID will not trigger a critical update. > Additionally, An SSID change is usually followed by an AP restart, > which will reset the BPCC value to 0. Right, I guess that is fair then. I guess it feels a bit weird to me overall, because I would expect that we only ever have one BSS entry for a BSSID (except as a workaround for some broken legacy APs). And it seems to me like one could increase the BPCC across an AP restart. So, I don't have a good alternative to your solution, I think. That said, maybe we should still add an additional check for the BPCC? Benjamin _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
