Date: Thu, 17 Apr 2014 11:00:44 -0700 From: Adrian Chadd <adrian@freebsd.org> To: Da Rock <freebsd-questions@herveybayaustralia.com.au> Cc: FreeBSD Questions <freebsd-questions@freebsd.org> Subject: Re: FBSD10 Atheros wifi not working Message-ID: <CAJ-Vmoksxv3skF21fhJkYUKpitcadfd9d-L2WAXBDaQvFawo3g@mail.gmail.com> In-Reply-To: <534E80AB.4000007@herveybayaustralia.com.au> References: <534B32FA.30500@herveybayaustralia.com.au> <CAJ-Vmok9Y2Dkgh=x2kP8su7LjrVo==ijSJ5GYrc=wzZv0uwNgQ@mail.gmail.com> <534BC380.90608@herveybayaustralia.com.au> <CAJ-Vmomb2d7jdPRmruaYK1LJ6gZ62_=4N3Ko9Z2eANLJ02c6Rw@mail.gmail.com> <534E80AB.4000007@herveybayaustralia.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
So, try eliminating the whole rc script mess for wpa_supplicant. * Don't put ath0/wlan0 in your startup config (rc.conf) * Do it manually: # ifconfig wlan0 create wlandev ath0 # wpa_supplicant -i wlan0 -c /etc/wpa_supplicant.conf & see if that's more stable. -a On 16 April 2014 06:07, Da Rock <freebsd-questions@herveybayaustralia.com.au> wrote: > On 04/15/14 12:16, Adrian Chadd wrote: >> >> Wait, you don't have this problem on the install image when running from >> USB? > > I thought I had answered this, but it doesn't appear to be showing in the > list (but that could just be my dying machine here with a slow email > client). > > Thats the weird part of all this: the install disk could scan for APs and > setup the network, but the actual installed system is having a coronary and > can't do anything. > > I've now managed to build a new kernel and install it (with debug), and > still no go. I can send the error logs if needed (privately), but I do > notice a lot of messages involving more than just the 9285 chipset (5416, > 5212) - is this normal? > > The kernel picks the right product, as the logs show ath0: <Atheros 9285> > and 'AR9285E_20 detected'. The errors remain unchanged though. This new > kernel also seems to crash quicker (or lock, might be a better term - I need > to hold the power button down and reboot for a response). The other gave > more time - might be the debug options? This happens despite the sysctls > turned off; with them on it turns in seconds rather than around 2-3 minutes. > > I could speculate further, but I reckon I'd hinder more than help at this > point. > >> >> >> -a >> >> >> On 14 April 2014 04:16, Da Rock >> <freebsd-questions@herveybayaustralia.com.au> wrote: >>> >>> Finally got the email client working again on the walking dead system :) >>> >>> On 04/14/14 13:57, Adrian Chadd wrote: >>>> >>>> Hi! >>> >>> I was hoping you were tuned in. If you can get this thing sorted again >>> I'll >>> owe you one! >>> >>>> You never said exactly what hardware it is Can you include a dmesg of >>>> the relevant bits during boot? >>> >>> Sorry, that was a bit remiss wasn't it? I had the info to put in too... >>> its >>> the AR9285 atheros chipset. >>> >>> dmesg is a bit harder being on a different system. I'll try and get the >>> info >>> in order though, at the time I was reading off the scrolling output from >>> vt1. The error(s) are repetitive though, so I'll just offer one cycle. It >>> doesn't seem to necessarily follow any real order though, from my >>> observation. >>> >>> ath0: ath_reset_grablock: didn't finish after 10 iterations >>> ath0: ath_reset_grablock: warning, recursive reset path! >>> ath0: ath_reset: concurrent reset! danger! >>> ath0: ath_reset: unable to reset hardware; hal status 14 >>> ath0: ath_chanset: unable to reset channel <channel>, hal status 14 >>> ... >>> ath0: ath_reset: unable to reset hardware; hal status 14 >>> >>>> Anything like the above kinda indicates there's some bigger issue going >>>> on. >>> >>> I can't imagine its a hardware issue as it seems to work with other >>> systems >>> (like the install memdisk). I thought with the hal debug sysctl being >>> available it would offer more info, but it seems it may not even be in >>> the >>> kernel although there is a sysctl. I tried compiling a new kernel with >>> the >>> debug options, but the backup battery is not working properly so it >>> depends >>> on ntpd to set the correct time, and the make requires the correct time >>> set; >>> so I'll have to address that before I can continue that course. >>> >>> Argh! The joys of Murphy and his damned laws! >>> >>>> >>>> -a >>>> >>>> >>>> >>>> On 13 April 2014 17:59, Da Rock >>>> <freebsd-questions@herveybayaustralia.com.au> wrote: >>>>> >>>>> The subject line may not be very descriptive, but I'm not sure exactly >>>>> where >>>>> to point as to the problem. All other similar issues are a few years >>>>> old >>>>> now, and on 8,9, some 7's. >>>>> >>>>> Long story short I'm just installing 10 (finally) on my laptop, and I >>>>> need >>>>> to get it back up and running asap as I am swamped in work. I didn't >>>>> think >>>>> it would be an issue as 9 had a decent working ath driver and I thought >>>>> it >>>>> could only get better. I'm now emailing using yet another system with >>>>> the >>>>> same device on 9, but a dying hdd which obviously also needs a >>>>> reinstall >>>>> on >>>>> new hdd. Wired is no good as we aren't equipped. >>>>> >>>>> So in bsdinstall, all works well and it sets up wifi easily. Reboot >>>>> into >>>>> the >>>>> new system an it has a coronary with messages from ath_reset_grablock, >>>>> ath_reset, ath_raw_xmit, ath_chan_set, ath_legacy_rx_tasklet: >>>>> >>>>> ath_reset: concurrent reset! danger! >>>>> >>>>> ath_reset_grablock: didn't finish after 10 iterations >>>>> ath_reset_grablock: warning, recursive reset path! >>>>> >>>>> ath_chan_set: concurrent reset! danger! >>>>> >>>>> ath_raw_xmit: sc_inreset_cnt > 0; bailing >>>>> >>>>> ath_legacy_rx_tasklet: sc_inreset_cnt > 0; bailing >>>>> >>>>> Scanning doesn't work; ath0 says it is associated, but wlan0 says no >>>>> channel. >>>>> >>>>> I've set dev.ath.0.hal.debug=1, but there is no ath.0.debug either >>>>> under >>>>> dev >>>>> or hw. I tried compiling the tools but it failed for lack of gcc. >>>>> >>>>> TIA guys, but I am really under the pump here as I thought this would >>>>> be >>>>> a >>>>> quick up and go now with the pkgng system as well in play. Now I'm up >>>>> the >>>>> proverbial creek without a paddle. >>>>> >>>>> Cheers >>>>> _______________________________________________ >>>>> freebsd-questions@freebsd.org mailing list >>>>> http://lists.freebsd.org/mailman/listinfo/freebsd-questions >>>>> To unsubscribe, send any mail to >>>>> "freebsd-questions-unsubscribe@freebsd.org" >>>> >>>> _______________________________________________ >>>> freebsd-questions@freebsd.org mailing list >>>> http://lists.freebsd.org/mailman/listinfo/freebsd-questions >>>> To unsubscribe, send any mail to >>>> "freebsd-questions-unsubscribe@freebsd.org" >>> >>> >>> _______________________________________________ >>> freebsd-questions@freebsd.org mailing list >>> http://lists.freebsd.org/mailman/listinfo/freebsd-questions >>> To unsubscribe, send any mail to >>> "freebsd-questions-unsubscribe@freebsd.org" > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmoksxv3skF21fhJkYUKpitcadfd9d-L2WAXBDaQvFawo3g>