[Solved] linux-3.3.1-1 breaks ath5k

Miblo · 2012-04-09 02:37:14

UPDATE: The situation as of 17th April 2012
Two patches – 1 and 2 – are reported to fix the issue, however they are not expected to be merged into the mainstream branch until linux-3.4. The choice, then, is between applying these patches (to 3.3.2) or waiting for 3.4 to be released.

UPDATE: 1st June 2012
As anticipated, linux-3.4.0-1 is being reported to fix the issue. It is currently in [testing], so the new choice is between enabling said repo and installing linux-3.4.0-1 or else waiting for it to arrive in [core].

UPDATE: 16th June 2012
linux-3.4 has since arrived in [core], so this problem may now be considered [Solved] for most users. The moral of the story: if you use ath5k, avoid linux-3.3 and skip straight to linux-3.4.

After completing an upgrade at 2012-04-08 21:21, during which the listed packages were upgraded…

automake-1.11.4-1
chromium-18.0.1025.151-3
dirmngr-1.1.0-4
sqlite-3.7.11-2
firefox-11.0-3
gawk-4.0.1-1
linux-3.3.1-1
linux-headers-3.3.1-1
lirc-utils-1:0.9.0-15
man-pages-3.38-1
ncmpcpp-0.5.10-1
ruby-1.9.3_p125-4

…I rebooted into the newly-installed 3.3.1 kernel and all was well. A subsequent upgrade at 2012-04-09 00:56 pulled in avahi-0.6.31-2 but immediately upon completion my Atheros AR2425 wireless fell silent. wicd-curses said 'No wireless networks found' and I rebooted to no avail. I downgraded back to avahi-0.6.31-1 but still wicd said the same. I found that Micah89 had a slightly different issue with wireless, tried nirvanix's suggestion to downgrade back to linux{,-headers}-3.2.14-1 but that hasn't worked either.

As it stands, at the last boot the ath5k module found my Atheros AR2425:

Apr  9 01:56:33 localhost kernel: [    8.874393] ath5k phy0: Atheros AR2425 chip found (MAC: 0xe2, PHY: 0x70)

…but then churned out the following messages:

Apr  9 02:14:54 localhost kernel: [   17.299370] ath5k phy0: gain calibration timeout (2412MHz)
Apr  9 02:14:55 localhost kernel: [   17.661428] ath5k phy0: gain calibration timeout (2412MHz)
Apr  9 02:14:55 localhost kernel: [   17.662512] ADDRCONF(NETDEV_UP): wlan0: link is not ready
Apr  9 02:14:55 localhost kernel: [   18.358020] ath5k phy0: gain calibration timeout (2417MHz)
Apr  9 02:14:56 localhost kernel: [   18.870449] ath5k phy0: gain calibration timeout (2422MHz)
Apr  9 02:14:56 localhost kernel: [   19.384749] ath5k phy0: gain calibration timeout (2427MHz)
Apr  9 02:14:57 localhost kernel: [   19.897030] ath5k phy0: gain calibration timeout (2432MHz)
Apr  9 02:14:57 localhost kernel: [   20.420107] ath5k phy0: gain calibration timeout (2437MHz)
Apr  9 02:14:58 localhost kernel: [   20.937227] ath5k phy0: gain calibration timeout (2442MHz)
Apr  9 02:14:58 localhost kernel: [   21.452738] ath5k phy0: gain calibration timeout (2447MHz)
Apr  9 02:14:59 localhost kernel: [   21.963524] ath5k phy0: gain calibration timeout (2452MHz)
Apr  9 02:14:59 localhost kernel: [   22.477717] ath5k phy0: gain calibration timeout (2457MHz)
Apr  9 02:15:00 localhost kernel: [   22.990711] ath5k phy0: gain calibration timeout (2462MHz)
Apr  9 02:15:00 localhost kernel: [   23.114736] r8169 0000:02:00.0: eth0: link down
Apr  9 02:15:00 localhost kernel: [   23.123943] ADDRCONF(NETDEV_UP): eth0: link is not ready
Apr  9 02:15:01 localhost kernel: [   23.517741] ath5k phy0: gain calibration timeout (2467MHz)
Apr  9 02:15:01 localhost kernel: [   23.994298] ath5k phy0: gain calibration timeout (2472MHz)

It would then sweep those frequencies again repeatedly, occasionally saying:

Apr  9 02:15:11 localhost kernel: [   34.119917] net_ratelimit: 1 callbacks suppressed

I'm totally stumped on this one, folks, and would really appreciate any help you could give to me. Avahi seems to be the key but it isn't even in my "/etc/rc.conf" DAEMONS array. Is anybody else having this issue?

Last edited by Miblo (2012-06-16 03:47:07)

ewaller · 2012-04-09 03:26:07

If you would, can you post the output of:
ip link
and of
lspci -nn (That is two n's, not m)

?? Thanks

Miblo · 2012-04-09 03:30:04

Urgh! And after powering down my netbook for half an hour, without having done any changes to the configuration, my wireless works again. This is on linux-3.2.14-1 with avahi-0.6.31-1. I'll upgrade these packages again tomorrow and see if the problem reoccurs.

Miblo · 2012-04-09 03:40:24

ewaller wrote:

If you would, can you post the output of:
ip link
and of
lspci -nn (That is two n's, not m)
?? Thanks

Oh, I am sorry, ewaller, but the problem has disappeared for now. However, if it's still relevant while my wireless is up, those commands give the following:

$ ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN qlen 1000
    link/ether 00:1e:68:c0:85:95 brd ff:ff:ff:ff:ff:ff
3: wlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether 00:22:69:3d:04:e1 brd ff:ff:ff:ff:ff:ff

$ lspci -nn
00:00.0 Host bridge [0600]: Intel Corporation Mobile 945GME Express Memory Controller Hub [8086:27ac] (rev 03)
00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 945GME Express Integrated Graphics Controller [8086:27ae] (rev 03)
00:02.1 Display controller [0380]: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller [8086:27a6] (rev 03)
00:1b.0 Audio device [0403]: Intel Corporation N10/ICH 7 Family High Definition Audio Controller [8086:27d8] (rev 02)
00:1c.0 PCI bridge [0604]: Intel Corporation N10/ICH 7 Family PCI Express Port 1 [8086:27d0] (rev 02)
00:1c.1 PCI bridge [0604]: Intel Corporation N10/ICH 7 Family PCI Express Port 2 [8086:27d2] (rev 02)
00:1c.2 PCI bridge [0604]: Intel Corporation N10/ICH 7 Family PCI Express Port 3 [8086:27d4] (rev 02)
00:1c.3 PCI bridge [0604]: Intel Corporation N10/ICH 7 Family PCI Express Port 4 [8086:27d6] (rev 02)
00:1d.0 USB controller [0c03]: Intel Corporation N10/ICH 7 Family USB UHCI Controller #1 [8086:27c8] (rev 02)
00:1d.1 USB controller [0c03]: Intel Corporation N10/ICH 7 Family USB UHCI Controller #2 [8086:27c9] (rev 02)
00:1d.2 USB controller [0c03]: Intel Corporation N10/ICH 7 Family USB UHCI Controller #3 [8086:27ca] (rev 02)
00:1d.3 USB controller [0c03]: Intel Corporation N10/ICH 7 Family USB UHCI Controller #4 [8086:27cb] (rev 02)
00:1d.7 USB controller [0c03]: Intel Corporation N10/ICH 7 Family USB2 EHCI Controller [8086:27cc] (rev 02)
00:1e.0 PCI bridge [0604]: Intel Corporation 82801 Mobile PCI Bridge [8086:2448] (rev e2)
00:1f.0 ISA bridge [0601]: Intel Corporation 82801GBM (ICH7-M) LPC Interface Bridge [8086:27b9] (rev 02)
00:1f.2 IDE interface [0101]: Intel Corporation 82801GBM/GHM (ICH7-M Family) SATA Controller [IDE mode] [8086:27c4] (rev 02)
00:1f.3 SMBus [0c05]: Intel Corporation N10/ICH 7 Family SMBus Controller [8086:27da] (rev 02)
02:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8101E/RTL8102E PCI Express Fast Ethernet controller [10ec:8136] (rev 02)
03:00.0 Ethernet controller [0200]: Atheros Communications Inc. AR242x / AR542x Wireless Network Adapter (PCI-Express) [168c:001c] (rev 01)

I'm now feeling very sheepish about all of this and half-hoping that I do actually encounter the problem again after reupgrading linux and avahi. Thank you very much for starting to help me.

bsdard · 2012-04-09 23:09:16

kernel 3.3.1 (686) | xorg + dwm (suckless.org)

I've just hit the same bug after less then 24 hours after the kernel upgrade to 3.3.1.

23:45:31 ~$ dmesg | grep -i ath
[ 7.766608] ath5k 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
[ 7.766627] ath5k 0000:03:00.0: setting latency timer to 64
[ 7.766714] ath5k 0000:03:00.0: registered as 'phy0'
[ 8.271586] ath: EEPROM regdomain: 0x65
[ 8.271589] ath: EEPROM indicates we should expect a direct regpair map
[ 8.271594] ath: Country alpha2 being used: 00
[ 8.271597] ath: Regpair used: 0x65
[ 8.417908] ath5k phy0: Atheros AR2425 chip found (MAC: 0xe2, PHY: 0x70)

Apr 9 01:36:50 ubox kernel: [ 7.984148] ath5k phy0: Atheros AR2425 chip found (MAC: 0xe2, PHY: 0x70)
Apr 9 21:38:21 ubox kernel: [23103.648598] ath5k phy0: gain calibration timeout (2412MHz) <== looks like the issue happened about 20 hours exactly after the device was initialised (it was a reboot, so uptime ~20hours too)
Apr 9 21:38:21 ubox kernel: [23103.992778] ath5k phy0: calibration of channel 1 failed
Apr 9 21:38:22 ubox kernel: [23105.160970] ath5k phy0: gain calibration timeout (2412MHz)
Apr 9 21:38:23 ubox kernel: [23105.680862] ath5k phy0: gain calibration timeout (2417MHz)
Apr 9 21:38:23 ubox kernel: [23106.190942] ath5k phy0: gain calibration timeout (2422MHz)
Apr 9 21:38:24 ubox kernel: [23106.710795] ath5k phy0: gain calibration timeout (2427MHz)

daemin.log
Apr 9 21:38:21 ubox dhcpcd[692]: wlan0: carrier lost
Apr 9 23:04:13 ubox dhcpcd[4756]: sending signal 1 to pid 692
Apr 9 23:04:13 ubox dhcpcd[692]: received SIGHUP, releasing <== I've been trying to reconnect wireless with netcfg2 -r <profile>. It did not help.

There were nothing in the logs before that so the issue has just started out of nowhere.
I am not running avahi-daemon, so I suppose it's the kernel bug alone.

I rolled back kernel to 3.2.13 and rebooted the box. Important to say that reboot did not solve the issue as apparently the device was left in a bad state. So I had to power down my laptop to actually get my wifi working again.

I did not collect any output before downgrade and reboot. Sorry - my bad.
The issue sounds like a counter wrap-around issue as it seems to depend on the uptime somehow.

I do not share optimism regarding the quality of the linux kernel and would not upgrade it that often in any Linux distribution.
I was expecting that sooner or later I shall hit the buggy kernel and here I am. I feel the time has come to to roll-back and to freeze linux changes.

Last edited by bsdard (2012-04-10 08:46:02)

Hyugga · 2012-04-10 02:39:02

I had the same issue today with an atheros chip. I downgraded the kernel to 3.2.14 and i had to power off the laptop completely to make the wifi work again.
No problems so far.
Good luck

hamelg · 2012-04-10 19:39:29

same issue happens after upgrading kernel to 3.3.1

$ ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc prio state UP qlen 1000
    link/ether 00:1c:c0:97:86:67 brd ff:ff:ff:ff:ff:ff
3: wlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UNKNOWN qlen 1000
    link/ether 00:22:b0:68:31:32 brd ff:ff:ff:ff:ff:ff
4: mon.wlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UNKNOWN qlen 1000
    link/ieee802.11/radiotap 00:22:b0:68:31:32 brd ff:ff:ff:ff:ff:ff
$ lspci -nn
00:00.0 Host bridge [0600]: Intel Corporation 4 Series Chipset DRAM Controller [8086:2e20] (rev 02)
00:01.0 PCI bridge [0604]: Intel Corporation 4 Series Chipset PCI Express Root Port [8086:2e21] (rev 02)
00:19.0 Ethernet controller [0200]: Intel Corporation 82567LF-2 Gigabit Network Connection [8086:10cd]
00:1a.0 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #4 [8086:3a37]
00:1a.1 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #5 [8086:3a38]
00:1a.2 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #6 [8086:3a39]
00:1a.7 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB2 EHCI Controller #2 [8086:3a3c]
00:1b.0 Audio device [0403]: Intel Corporation 82801JI (ICH10 Family) HD Audio Controller [8086:3a3e]
00:1d.0 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #1 [8086:3a34]
00:1d.1 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #2 [8086:3a35]
00:1d.2 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #3 [8086:3a36]
00:1d.7 USB controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB2 EHCI Controller #1 [8086:3a3a]
00:1e.0 PCI bridge [0604]: Intel Corporation 82801 PCI Bridge [8086:244e] (rev 90)
00:1f.0 ISA bridge [0601]: Intel Corporation 82801JIR (ICH10R) LPC Interface Controller [8086:3a16]
00:1f.2 SATA controller [0106]: Intel Corporation 82801JI (ICH10 Family) SATA AHCI Controller [8086:3a22]
00:1f.3 SMBus [0c05]: Intel Corporation 82801JI (ICH10 Family) SMBus Controller [8086:3a30]
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation G94 [GeForce 9600 GT] [10de:0622] (rev a1)
02:02.0 Multimedia controller [0480]: Philips Semiconductors SAA7131/SAA7133/SAA7135 Video Broadcast Decoder [1131:7133] (rev d0)
02:03.0 Ethernet controller [0200]: Atheros Communications Inc. AR5212/AR5213 Wireless Network Adapter [168c:0013] (rev 01)

Last edited by hamelg (2012-04-10 19:47:27)

Miblo · 2012-04-10 20:08:10

Cheers for your data, folks. I'm afraid I haven't had the courage to upgrade the kernel again, but I have just reupgraded avahi and my wireless hasn't died. No idea why the initial upgrade of avahi (while booted into linux-3.3.1-1) triggered the issue, however I'm going along with you guys in supposing that it is actually the kernel that's at fault.

That counter wrap-around idea of bsdard's is the closest we have to suggesting a fix for it. I'm not knowledgeable enough to come up with any more suggestions, sadly. Any ideas how we can help the devs, folks?

hamelg · 2012-04-10 20:14:56

here's some clues :
http://thread.gmane.org/gmane.linux.ker … eral/86055

Miblo, could you change the title in "kernel 3.3.1 : ath5k broken"

Last edited by hamelg (2012-04-10 20:20:31)

Miblo · 2012-04-10 20:26:17

hamelg wrote:

here's some clues :
http://thread.gmane.org/gmane.linux.ker … eral/86055
Miblo, could you change the title in "kernel 3.3.1 : ath5k broken"

Ah, cheers, hamelg. It appears to be fixed, according to Xlator, so hopefully the patch will be merged into the next commit. I'll change the thread title.

hamelg · 2012-04-10 21:26:34

Miblo wrote:

It appears to be fixed, according to Xlator, so hopefully the patch will be merged into the next commit. I'll change the thread title.

It is not the same issue. it is related to ath9k driver, not ath5k.

Miblo · 2012-04-10 21:31:56

hamelg wrote:

Miblo wrote:
It appears to be fixed, according to Xlator, so hopefully the patch will be merged into the next commit. I'll change the thread title.
It is not the same issue. it is related to ath9k driver, not ath5k.

Ah, apologies! I'd better take the [Solved] back out of the title as a precaution until we're sure it is solved. Cheers for your help.

gabuzo · 2012-04-11 15:15:01

See https://bbs.archlinux.org/viewtopic.php?pid=1085916

Miblo · 2012-04-11 15:36:11

gabuzo wrote:

See https://bbs.archlinux.org/viewtopic.php?pid=1085916

Although, to be fair, they're talking about ath9k over there. We're on about ath5k, as hamelg reminded me,

hamelg · 2012-04-11 18:04:31

Unfortunately, these 2 patches are not present in the 3.3.1 :

http://git.kernel.org/?p=linux/kernel/g … 37359577d2
http://git.kernel.org/?p=linux/kernel/g … 7a1ddb9b0b

Last edited by hamelg (2012-04-11 18:05:01)

lgeek · 2012-04-11 19:23:52

Try https://bbs.archlinux.org/viewtopic.php?pid=1085916
I think that in kernel 3.3.1 there is a regression on all ath* modules. Hope the best.
Good luck.

Yorick · 2012-04-11 21:28:02

I encountered the same bug. It appears to have to do with the powernow-k8 module. I haven't encountered it with it blacklisted.
A temporary workaround to still have cpu frequency switching would be to use phc-k8 instead.

Edit: it doesn't appear to fix the problem entirely, but it seems to happen less now.

Last edited by Yorick (2012-04-12 19:52:14)

rps63ifid · 2012-04-14 21:36:25

I just saw that the 3.3.2 kernel just hit testing today; can anyone on testing provide a read as to whether it does in fact solve this problem?

--
/ron

jamesbroadhead · 2012-04-14 22:52:01

rps63ifid wrote:

I just saw that the 3.3.2 kernel just hit testing today; can anyone on testing provide a read as to whether it does in fact solve this problem?
--
/ron

Looks to me as though 5c17ddc4a047c59638c7eb8537aa887a1ddb9b0b [ath5k: do not re-run AGC calibration periodically] made it in, but the other one didn't. Neither patch is in 3.3.2.

I have the same problem here; Acer Revo (media-center) with an AR242x / AR542x. Persistent across a reboot to a previously-good kernel. Disconnected from the socket on power-off, and it worked with 3.3.1 after that (so far). Will test with 3.3.2.
Edit: it stopped working inside 2 hours with 3.3.1

Last edited by jamesbroadhead (2012-04-15 19:22:53)

zwa · 2012-04-15 04:44:37

The 3.3.2 kernel changelog does not mention ath5k, only ath9k:

http://www.kernel.org/pub/linux/kernel/ … eLog-3.3.2

jamesbroadhead · 2012-04-15 11:25:46

zwa wrote:

The 3.3.2 kernel changelog does not mention ath5k, only ath9k:
http://www.kernel.org/pub/linux/kernel/ … eLog-3.3.2

I read the source.

Edit: too briefly, it seems! Neither were applied. (Note to self: no more midnight kernel source scanning).

Last edited by jamesbroadhead (2012-04-15 12:34:13)

Mr Green · 2012-04-15 11:29:08

Has any of you tried loading 3.3.2 from testing to see if that solves the issue?

jamesbroadhead · 2012-04-15 19:19:55

6h30 uptime here with a patched 3.3.2.

Last edited by jamesbroadhead (2012-04-15 19:23:10)

zwa · 2012-04-15 23:46:10

jamesbroadhead wrote:

6h30 uptime here with a patched 3.3.2.

Could you elaborate a bit more on that? what patches did you applied to the stable 3.3.2 kernel source?

thx!

c00kiemon5ter · 2012-04-16 08:26:02

I use ath5k too, and got hit by that. It happened twice, while I was sshed to my machine.
I logged out and then couldn't connect.

Back home, I tried to reconnect but was unable to.
iwlist scanning
returned no results.
modprobe -r ath5k; modprobe ath5k;
didn't help either

reboot fixes the issue here, but it's annoying, plus I can't really work when I'm not home.
..crawling the web now..

according to http://thread.gmane.org/gmane.linux.ker … eral/86055 (link from above)
the patches are merged in 3.4-rc1 , so I guess the solution for now is
either wait or manually patch and build a kernel

Last edited by c00kiemon5ter (2012-04-16 08:37:32)

Arch Linux

#1 2012-04-09 02:37:14

[Solved] linux-3.3.1-1 breaks ath5k

#2 2012-04-09 03:26:07

Re: [Solved] linux-3.3.1-1 breaks ath5k

#3 2012-04-09 03:30:04

Re: [Solved] linux-3.3.1-1 breaks ath5k

#4 2012-04-09 03:40:24

Re: [Solved] linux-3.3.1-1 breaks ath5k

#5 2012-04-09 23:09:16

Re: [Solved] linux-3.3.1-1 breaks ath5k

#6 2012-04-10 02:39:02

Re: [Solved] linux-3.3.1-1 breaks ath5k

#7 2012-04-10 19:39:29

Re: [Solved] linux-3.3.1-1 breaks ath5k

#8 2012-04-10 20:08:10

Re: [Solved] linux-3.3.1-1 breaks ath5k

#9 2012-04-10 20:14:56

Re: [Solved] linux-3.3.1-1 breaks ath5k

#10 2012-04-10 20:26:17

Re: [Solved] linux-3.3.1-1 breaks ath5k

#11 2012-04-10 21:26:34

Re: [Solved] linux-3.3.1-1 breaks ath5k

#12 2012-04-10 21:31:56

Re: [Solved] linux-3.3.1-1 breaks ath5k

#13 2012-04-11 15:15:01

Re: [Solved] linux-3.3.1-1 breaks ath5k

#14 2012-04-11 15:36:11

Re: [Solved] linux-3.3.1-1 breaks ath5k

#15 2012-04-11 18:04:31

Re: [Solved] linux-3.3.1-1 breaks ath5k

#16 2012-04-11 19:23:52

Re: [Solved] linux-3.3.1-1 breaks ath5k

#17 2012-04-11 21:28:02

Re: [Solved] linux-3.3.1-1 breaks ath5k

#18 2012-04-14 21:36:25

Re: [Solved] linux-3.3.1-1 breaks ath5k

#19 2012-04-14 22:52:01

Re: [Solved] linux-3.3.1-1 breaks ath5k

#20 2012-04-15 04:44:37

Re: [Solved] linux-3.3.1-1 breaks ath5k

#21 2012-04-15 11:25:46

Re: [Solved] linux-3.3.1-1 breaks ath5k

#22 2012-04-15 11:29:08

Re: [Solved] linux-3.3.1-1 breaks ath5k

#23 2012-04-15 19:19:55

Re: [Solved] linux-3.3.1-1 breaks ath5k

#24 2012-04-15 23:46:10

Re: [Solved] linux-3.3.1-1 breaks ath5k

#25 2012-04-16 08:26:02

Re: [Solved] linux-3.3.1-1 breaks ath5k

Board footer