[STAGING] ath5k kernel panic

Laptop with a ATI mobility GPU is crashing a large amount. I have tried both the radeon and radeon HD xorg packages and both are equally unstable. Not yet tried the -ati driver, chipset is not supported by the vendor drivers (fglrx).

System is no more stable with desktop effects disabled.

I have also experimented with the 2.6.34-rc7 and the 2.6.33 kernels, but both appear to be equally unstable.

Any tips on a workaround would be appreciated.

Have just noticed a lockup without either the radeon or radeonhd packages installed.

I have since black-listed the ath5k driver and am experimenting with this state... clearly this stops the wireless working, which is a bit of a pain.

I have changed the abstract to reflect the fact that it appears to be the ath5k driver causing this lock up.

I would also like to confirm the suspicion of ath5k. I get a kernel panic quite frequently, always when I am connected via wifi. I recall I had this exact same problem in 9.04 with the ath5k driver. I am having trouble reproducing the crash, but will try to get some logs from the panic. Other files attached.

Thinkpad 10.04 64-bit with 2.6.32-22 kernel. wireless-backports installed, no madwifi drivers or any other modifications.

Any help would be appreciated, this is rendering my machine unusable.

Not sure if this matters, but I noticed in my log files and in the OP log files that when the ath5k driver is loaded the wireless device registered the AR5414 model, yet we both have AR5212 model on our machines (at least from lspci). Not sure if that is significant.

From dmesg:
[ 44.418566] ath5k phy0: Atheros AR5414 chip found (MAC: 0xa3, PHY: 0x61)

From lspci:
03:00.0 Ethernet controller: Atheros Communications Inc. AR5212 802.11abg NIC (rev 01)

Finally managed to capture the output of the freeze. Attached in the image. I can't make any sense from it but it clearly involves the ath5k driver.

Also, have been running the 2.6.31-21 version of the kernel for a little bit with no problems so far.

I have the same kernel panic lock ups on a ThinkPad T60p with Atheros 5 chipset.

Installing the linux-backports-modules-wireless seems to make the kernel panics even more frequent.

Kernel: 2.6.31-22

@Brian: are you suggesting to revert to the 2.6.31-22 kernel?

@Jan - Yes, I have been running the 2.6.31-21 kernel for several days now with no kernel panics (no backports installed either). The wireless is a little unreliable, but certainly usable.

Huh, running linux-backports-modules-wireless-2.6.32-23-generic on my thinkpad T60 stopped the panics for me (so far, ~8 hours of use, when I'd normally panic in less than ~15 mins).

Scratch that comment about wireless backports. It's seemingly a bit harder to trigger now (for me), but it just happened again, as Jan said.

This seems to affect me, too.
Got an IBM thinkpad X31.
If I remember right it started with ath5k and kernel 2.6.31-20.

Regarding Brian's suspicion:
My Laptop gets the model right and crashes nevertheless...

lspci:
02:02.0 Ethernet controller: Atheros Communications Inc. AR5212 802.11abg NIC (rev 01)
dmesg:
ath5k phy0: Atheros AR5212 chip found (MAC: 0x56, PHY: 0x41)

Had a similar problem with madwifi a year and a half ago.

I can not really reproduce it, although while I have been running transmission and surfing with firefox/chrome at the same time I almost everytime got the kernel panic sooner or later.

I believe this is related to Bug #593650 and #300693. For me, upgrading to the 2.6.35-rc3 from mainline completely stopped the ath5k related panics. However, it appears to break libvirt VM management in some way.

As mentioned before, I still suffer from these kernel panics, all be it very occasionally and only if I use a large amount of bandwidth over the wireless network.

I've experienced this problem with kernels 2.6.31-22 and 2.6.31-23 and with and without the backports modules (i386).

I will be trying the AMD64 version of Ubuntu soon, to see if there might be a difference.

@Jamin: Indeed bug [url=https://bugs.launchpad.net/ubuntu/+source/linux/+bug/300693]#300693[/url] looks the same.

I've still been running kernel 2.6.31-21 with no problems for over a month now.

@Jan - I've been running 64-bit, so it still gets the panics.

Bug #300693 seems to have replicated my experience exactly. I had problems with 8.10 that went away in 9+, then seem to have resurfaced again in 10.04. Was there ever a fix or resolution for #300693?

I've been using the Kernel PPA to run the mainline (2.6.35-19-generic #25~lucid1-Ubuntu SMP, as of the time of this writing), and tempting fate (using LOTS of bandwidth with ath5k, and then using Chrome's bookmark/account sync in the middle of all of it).

It seems practically bullet-proof on this kernel. I wonder why 2.6.32 is so badly broken with this wifi module, but I can't say I'll be thinking about it a lot now that I'm on Maverick's kernel.

Alex Ainscow, thank you for reporting this bug and helping make Ubuntu better. This bug was reported a while ago and there hasn't been any activity in it recently. We were wondering if this is still an issue? Can you try with the latest development release of Ubuntu? ISO CD images are available from http://cdimage.ubuntu.com/releases/ .

If it remains an issue, could you run the following command in the development release from a Terminal (Applications->Accessories->Terminal). It will automatically gather and attach updated debug information to this report.

apport-collect -p linux <replace-with-bug-number>

Also, if you could test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

[Expired for linux (Ubuntu) because there has been no activity for 60 days.]