On 14/02/2023 19.38, Arend van Spriel wrote: > > > On 2/14/2023 11:30 AM, Hector Martin wrote: >> On 14/02/2023 19.07, Arend van Spriel wrote: >>> + Double Lo >>> >>> On 2/14/2023 10:33 AM, Hector Martin wrote: >>>> Using the WSEC command instead of sae_password seems to be the supported >>>> mechanism on newer firmware, and also how the brcmdhd driver does it. >>> >>> The SAE code in brcmfmac was added by Cypress/Infineon. For my BCA >>> devices that did not work, but this change should be verified on Cypress >>> hardware. >> >> Do you mean the existing SAE code does not work on BCA, or this version >> doesn't? > > I meant the existing SAE code. I will give your patches a spin on the > devices I have. > >> I assume/hope this version works for WCC in general, since that is what >> the Apple-relevant chips are tagged as. If so it sounds like we need a >> firmware type conditional on this, if CYW needs the existing behavior. > > Right. Let's hope we get some feedback from them. Any news on this? Nothing from the Cypress guys (nor to any of my previous emails about other stuff, for that matter), so if you can confirm this works on your chips I'd rather just blindly add the CYW/not firmware variant check and call it a day. We can't wait forever for them to show up. If they expect their chips to continue work with mainline they need to actually interact on the MLs, otherwise they should expect us to possibly accidentally break things even if we try not to. As far as I can tell they seem completely disinterested in talking about anything, and we can't let that block progress for everyone else. - Hector
