Kerneloops and disconnecting wlan

[PrototypeX29A]

Binary package hint: kerneloops

Distro: Ubuntu 9.10

After a while of using WLAN (especially when causing huge traffic) the kernel will oops with one of the following error messages in kernel.log

[ 2253.163051] BUG: unable to handle kernel NULL pointer dereference at 00000001
[ 107.994107] BUG: unable to handle kernel paging request at 4a709527

After that the wlan will disconnect not reconnect until rebooting. The error messages are taken from two different sessions.

[Matthew (matthew-redhat-bugs)]

Description of problem:

Oops in ath9k driver under light load and subsequent loss of networking. A reboot must be done to make the system usable again.

This bug seems to be fixed in at least 2.6.32-rc6-git3

Version-Release number of selected component (if applicable):

2.6.31.5-127 PAE

How reproducible:

100% of the time

Steps to Reproduce:

1) rm -rf /var/cache/yum/*

2) yum -y update

Obtaining the compressed repo data is sufficient to trigger the oops.

Actual results:

yum -y update stalls and networking stops working.

Expected results:

I would expect the yum -y update to complete successfully.

Additional info:

Oops is attached.

[Matthew (matthew-redhat-bugs)]

Created attachment 368980
ath9k oops

[Matthew (matthew-redhat-bugs)]

I am currently building 2.6.31.6 with the F12 config and debugging enabled in the ath9k driver and will test that as soon as it is done.

[Matthew (matthew-redhat-bugs)]

This does not reproduce with 2.6.31.6 using the F12 kernel-PAE .config plus ath9k debugging enabled.

[John (john-redhat-bugs)]

I don't see any patches in 2.6.31.5..2.6.31.6 that would seem to relate -- no ath9k ones and only a couple of mac80211 that don't seem related to the oops path.

Maybe it's a Heizenbug? Could you try your 2.6.31.6 build using the stock F12-PAE config (i.e. w/o turning-on ath9k debugging)?

[Luis (luis-redhat-bugs-1)]

John, although 2.6.31.6 did not have any new ath9k or mac80211 specific patch there was the PCI memory rounding fix that also fixed an issue with another netbook. So far that we know if this only affected some Aspires (if the sky2 issue was with an Aspire as well).

I cannot be too sure that that same patch fixed the issue but it would be easy to test by just reverting it.

Matthew, if you have spare time can you try reverting this patch to see if it did indeed fix the issue:

Author: Yinghai Lu <email address hidden>

   pci: increase alignment to make more space for hidden code

So far the other issue refer to:

ath9k load issue:

http://bugzilla.kernel.org/show_bug.cgi?id=14402

iwlagn + sky combo issues:

http://bugzilla.kernel.org/show_bug.cgi?id=13940

These issues are loading related though so it would be surprising to see an issue not related to that. The patch in question would change the pci memory address start.

Only thing that I can think of between 2.6.31.5 and 2.6.31.6 but perhaps this wasn't it, only testing would tell.

[Matthew (matthew-redhat-bugs)]

I came home today after work and the laptop had oopsed while running 2.6.31.6 with debugging enabled in the ath9k driver. It looks like the same oops signature. I am attaching it.

[Matthew (matthew-redhat-bugs)]

Created attachment 369122
oops while running 2.6.31.6 with ath9k debugging enabled

[Pierre (pierre-redhat-bugs)]

I have exactly the same problem, and disabling kernel modesetting seems to correct it. To reproduce it, I do the same as you, booting linux on kernel 2.6.31.6-134.fc12.i686, launching a web application (like yum or firefox) and right after that, it stops working, and I can't shutdown properly.

I have this wifi card :
03:00.0 Network controller: Atheros Communications Inc. AR928X Wireless Network Adapter (PCI-Express) (rev 01)
        Subsystem: Foxconn International, Inc. Device e01f
        Kernel driver in use: ath9k
        Kernel modules: ath9k

on an ACER laptop, and the same message as you, at the end of your attachment (the kernel oops about null pointer dereferencing and the same last sysfs used).

I was desperate seeing no one else with this problem xD

[Pierre (pierre-redhat-bugs)]

Here is what my report says, if it can be userful :

Nov 18 12:47:55 localhost NetworkManager: <info> (wlan0): supplicant connection state: disconnected -> scanning
Nov 18 12:47:55 localhost kernel: BUG: unable to handle kernel NULL pointer dereference at 00000001
Nov 18 12:47:55 localhost kernel: IP: [<c049b517>] put_page+0xe/0x76
Nov 18 12:47:55 localhost kernel: *pde = bf730067
Nov 18 12:47:55 localhost kernel: Oops: 0000 [#1] SMP
Nov 18 12:47:55 localhost kernel: last sysfs file: /sys/devices/LNXSYSTM:00/device:00/PNP0A08:00/device:0e/PNP0C09:00/PNP0C0A:00/power_supply/BAT0/charge_full
Nov 18 12:47:55 localhost kernel: Modules linked in: vboxnetadp vboxnetflt vboxdrv sunrpc ipv6 cpufreq_ondemand acpi_cpufreq fuse dm_multipath uinput arc4 ecb snd_hda_codec_atihdmi snd_hda_codec_realtek snd_hda_intel ath9k snd_hda_codec mac80211 snd_hwdep snd_seq snd_seq_device uvcvideo ath snd_pcm snd_timer videodev cfg80211 snd acer_wmi tg3 rfkill v4l1_compat i2c_i801 soundcore serio_raw iTCO_wdt snd_page_alloc iTCO_vendor_support joydev wmi usb_storage video output radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core [last unloaded: scsi_wait_scan]
Nov 18 12:47:55 localhost kernel:
Nov 18 12:47:55 localhost kernel: Pid: 681, comm: phy0 Not tainted (2.6.31.5-127.fc12.i686 #1) Aspire 5738
Nov 18 12:47:55 localhost kernel: EIP: 0060:[<c049b517>] EFLAGS: 00010282 CPU: 0
Nov 18 12:47:55 localhost kernel: EIP is at put_page+0xe/0x76
Nov 18 12:47:55 localhost kernel: EAX: 00000001 EBX: f1887840 ECX: f6e84188 EDX: 00000000
Nov 18 12:47:55 localhost kernel: ESI: 00000001 EDI: f1887860 EBP: f41c7e44 ESP: f41c7e34
Nov 18 12:47:55 localhost kernel: DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
Nov 18 12:47:55 localhost kernel: Process phy0 (pid: 681, ti=f41c6000 task=f6332640 task.ti=f41c6000)
Nov 18 12:47:55 localhost kernel: Stack:
Nov 18 12:47:55 localhost kernel: f6c003ac f1887840 00000001 f1887860 f41c7e54 c06d81d8 f1887840 f6e8c1a0
Nov 18 12:47:55 localhost kernel: <0> f41c7e60 c06d7e20 f1887840 f41c7e68 c06d7eaa f41c7e94 f8be3c3a f6e8d148
Nov 18 12:47:55 localhost kernel: <0> f41c7ea8 00000009 f19544aa f1887860 f19544ae f6e8c9f4 f1887840 f41c7ee4
Nov 18 12:47:55 localhost kernel: Call Trace:
Nov 18 12:47:55 localhost kernel: [<c06d81d8>] ? skb_release_data+0x56/0x96
Nov 18 12:47:55 localhost kernel: [<c06d7e20>] ? __kfree_skb+0x17/0x72
Nov 18 12:47:55 localhost kernel: [<c06d7eaa>] ? consume_skb+0x2f/0x31
Nov 18 12:47:55 localhost kernel: [<f8be3c3a>] ? ieee80211_tx_status+0x367/0x36f [mac80211]
Nov 18 12:47:55 localhost kernel: [<f7d484e5>] ? ath_tx_complete_buf+0x111/0x166 [ath9k]
Nov 18 12:47:55 localhost kernel: [<f7d4934d>] ? ath_draintxq+0x129/0x1b9 [ath9k]
Nov 18 12:47:55 localhost kernel: [<f7d49eec>] ? ath_drain_all_txq+0xd8/0xe6 [ath9k]
Nov 18 12:47:55 localhost kernel: [<f7d45da0>] ? ath_set_channel+0x4d/0xe3 [ath9k]
Nov 18 12:47:55 localhost kernel: [<f7d45fbc>] ? ath9k_config+0x186/0x1bd [ath9k]
Nov 18 12:47:55 localhost kernel: [<f8be381c>] ? ieee80211_hw_config+0x91/0x99 [mac80211]
Nov 18 12:47:55 localhost kernel: [<f8be744f>] ? ieee80211_scan_work+0xeb/0x178 [mac80211]
Nov 18 12:47:55 localhost kernel: [<c0446238>] ? wor...

[PrototypeX29A (preineke)]

AplayDevices:
 **** List of PLAYBACK Hardware Devices ****
 card 0: Intel [HDA Intel], device 0: ALC272 Analog [ALC272 Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
Architecture: i386
ArecordDevices:
 **** List of CAPTURE Hardware Devices ****
 card 0: Intel [HDA Intel], device 0: ALC272 Analog [ALC272 Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: peter 1836 F.... pulseaudio
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
 Card hw:0 'Intel'/'HDA Intel at 0xd6700000 irq 22'
   Mixer name : 'Realtek ALC272'
   Components : 'HDA:10ec0272,10250212,00100001'
   Controls : 13
   Simple ctrls : 8
DistroRelease: Ubuntu 9.10
HibernationDevice: RESUME=UUID=4c689dbc-a530-4d6e-aab2-d6fe6c038a3d
InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
MachineType: Acer Aspire 5732Z
Package: linux (not installed)
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-16-generic root=UUID=5c746146-a177-4c6e-a8bc-ead45d7f0d6d ro quiet splash
ProcEnviron:
 SHELL=/bin/bash
 LANG=en_US.UTF-8
ProcVersionSignature: Ubuntu 2.6.31-16.53-generic
RelatedPackageVersions:
 linux-backports-modules-2.6.31-16-generic N/A
 linux-firmware 1.25
RfKill:
 0: phy0: Wireless LAN
  Soft blocked: no
  Hard blocked: no
Uname: Linux 2.6.31-16-generic i686
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare
XsessionErrors:
 (gnome-settings-daemon:1848): GLib-CRITICAL **: g_propagate_error: assertion `src != NULL' failed
 (gnome-settings-daemon:1848): GLib-CRITICAL **: g_propagate_error: assertion `src != NULL' failed
 (nautilus:1871): Eel-CRITICAL **: eel_preferences_get_boolean: assertion `preferences_is_initialized ()' failed
 (polkit-gnome-authentication-agent-1:1894): GLib-CRITICAL **: g_once_init_leave: assertion `initialization_value != 0' failed
dmi.bios.date: 08/03/2009
dmi.bios.vendor: Acer
dmi.bios.version: V2.05
dmi.board.asset.tag: Base Board Asset Tag
dmi.board.name: Aspire 5732Z
dmi.board.vendor: Acer
dmi.board.version: V2.05
dmi.chassis.type: 10
dmi.chassis.vendor: Acer
dmi.chassis.version: V2.05
dmi.modalias: dmi:bvnAcer:bvrV2.05:bd08/03/2009:svnAcer:pnAspire5732Z:pvrV2.05:rvnAcer:rnAspire5732Z:rvrV2.05:cvnAcer:ct10:cvrV2.05:
dmi.product.name: Aspire 5732Z
dmi.product.version: V2.05
dmi.sys.vendor: Acer

[jtniehof (jtniehof)]

I think the Fedora bug is close enough to consider this one confirmed. A related, but not identical, Ubuntu bug is #426130 .
There is also a mainline kernel bug on the general issues with this driver: http://bugzilla.kernel.org/show_bug.cgi?id=13807

[jtniehof (jtniehof)]

Ooops, LP didn't auto-link... bug #426130 or https://bugs.launchpad.net/ubuntu/+source/linux/+bug/426130

[PrototypeX29A (preineke)]

Just wanted to add, that i still have the problem with the latest version from karmic-repo:

peter@yoggoth:~$ uname -r
2.6.31.6-zarkov

[PrototypeX29A (preineke)]

It seems, that I can NOT reproduce it with 2.6.33-999-generic.

[DennyHalim.com (dennyhalim.com-redhat-bugs)]

same problem on acer aspire 4540.

i 'upgrade' to rawhide kernel and seems everything's ok now.

[PrototypeX29A (preineke)]

It IS reproducible with karmic's: 2.6.31.17.30

[Thomas Glaser (thomas-glsr)]

I'm also experiencing this bug with a freshly installed Ubuntu 9.10 on an Acer Aspire 5732Z.
Kernel 2.6.31.17. Any hints on workarounds appreciated.

[PrototypeX29A (preineke)]

To me it does not happen anymore with the official Kernel

[PrototypeX29A (preineke)]

To be more specific: the bug isn't occurring anymore with 2.6.31-21-generic while I am too using 'Acer Aspire 5732Z'
With Network controller: Atheros Communications Inc. AR928X Wireless Network Adapter (PCI-Express) (rev 01)
Ethernet controller: Attansic Technology Corp. Atheros AR8132 / L1c Gigabit Ethernet Adapter (rev c0)

[Jeremy Foshee (jeremyfoshee)]

per the original reporter, i am marking this bug for Ubuntu as Fix released. Those still affected by a similar issue, please open new bugs so that we may address them.

Thanks!

~JFo

[Stanislaw (stanislaw-redhat-bugs)]

F-12 use now 2.6.32 based kernel, which according to comment 0 have this bug fixed.