Re: My iwn device can no longer scan for access points with stable/13
Date: Sun, 20 Feb 2022 18:28:53 UTC
On Thu, 17 Feb 2022, John Schneider wrote: > I did the normal upgrade process described in the handbook - buildworld, > buildkernel, installkernel, reboot, installworld, etc. iwn works fine > after stable/13 installkernel and reboot (which is at that point > releng/13.0 world with stable/13 kernel. In fact I am sending this email > via iwn while using stable/13 kernel. It is only after installing stable/13 > world and rebooting that iwn can no longer scan. > > I am happy to re-install stable/13 world and produce more debugging info. > Just let me know what would be helpful. Ok, that limits things; adding Cy to Cc: ; he may know better about the state on stable/13 when it comes to wpa_supplicant. Might also be intersting to try an ifconfig from stable/13 an manually trigger and list the scan result: ifconfig wlan0 scan ifconfig wlan0 list scan and see if that still works. Problem is that ifconfig these days needs a lot of libraries (on main) but the /rescue version is probably directly usable so finding that in the obj tree and trying it out directly would be helpful as a starter. /bz > Thanks, > John > > On Wed, Feb 16, 2022 at 7:46 PM Bjoern A. Zeeb < > bzeeb-lists@lists.zabbadoz.net> wrote: > >> On Wed, 16 Feb 2022, John Schneider wrote: >> >> Hi, >> >>> Today I upgraded from releng/13.0 to stable/13 but found my iwn device >>> could no longer scan for access points. I rolled back to releng/13.0 and >>> everything was fine again. No config changes were made during the >> upgrade >>> or rollback process. >>> >>> /var/log/messages contained messages such as the one below with each scan >>> attempt: >>> Feb 16 16:12:33 mobile1 wpa_supplicant[385]: wlan0: >> CTRL-EVENT-SCAN-FAILED >>> ret=-1 retry=1 >> >> -1 is nothing the kernel should return so I don't know what the actual >> error is. >> >> >>> I didn't see any other errors except ACPI stuff which looked normal to me >>> for this machine. For reference, here is the pciconf output for my >> device: >>> >>> iwn0@pci0:2:0:0: class=0x028000 rev=0x35 hdr=0x00 vendor=0x8086 >>> device=0x >>> 422b subvendor=0x8086 subdevice=0x1121 >>> vendor = 'Intel Corporation' >>> device = 'Centrino Ultimate-N 6300' >>> class = network >>> >>> Is it possible that work being done for iwlwifi could affect scanning >> with >>> iwn and other drivers? >> >> It shouldn't as there were very little functional changes to net80211 >> itself so far relating to this. >> >> I would more likely think of wpa_supplicant updates but even that we'd >> probably have heard a lot more noise. You could run that with >> >> wpa_supplicant_flags="-sdd" >> >> in rc.conf (from the top of my head) and see. >> >> Also did you update kernel and world or was that just a kernel update >> (which would rule out the above wpa_supplicant update theory)? >> >> >> I also only see 3 changes to iwn(4) in main and 2 of them in stable/13 >> since releng/13.0: >> >> >> https://cgit.freebsd.org/src/commit/sys/dev/iwn?h=stable/13&id=d61f10a8780ef94f82e384248ce40d24fc27ad2b >> >> https://cgit.freebsd.org/src/commit/sys/dev/iwn?h=stable/13&id=67b7e2de4fc5dc1165feee8ca281c44444425d0f >> >> and the one Adrian did not MFC: >> >> https://cgit.freebsd.org/src/commit/sys/dev/iwn?id=b6fd00791f2b9690b0a5d8670fc03f74eda96da2&h=stable%2F13 >> >> >> In case you are willing to go back to stable/13 and re-test things >> let me know so we can crank some debugging/logging up to see where the >> error comes from. >> >> >> >>> On a separate subject, I had tested iwlwifi on a >>> separate machine with stable/13 and an AX201 and it was quite stable >>> (though slow as has been reported by others). Keep up the great work! >> >> Thanks for testing. I'll try to make it more stable and then faster :) >> >> Best Regards, >> Bjoern >> >> -- >> Bjoern A. Zeeb r15:7 >> > -- Bjoern A. Zeeb r15:7