Latest udev causes kernel panics (something about libata)

xsoul · 2011-01-08 06:15:17

@andrews
oh no.. then i have no idea what is causing the kernel panic on my machine. problem still unresolved.

sm4tik · 2011-01-08 08:40:34

Daniel_F wrote:

@sm4tik
Please do not generalize...
Arch is by nature a bleeding-edge distro. As such it is the user's (aka YOU) responsability to be diligent with updates, avoiding those that might cause trouble. Pacman even makes it easy to downgrade if necessary !
But...
From the FAQ (https://wiki.archlinux.org/index.php/FAQ):
Q) Why would I not want to use Arch?
A) You may not want to use Arch if
...
* you do not want a bleeding edge, rolling release GNU/Linux distribution.
...
If you want something similar to Arch but more stable then perhaps you should try Slackware or, for something more "user friendly" try SalixOS
But make no mistake: GNU/Linux is not Windows. The open-source nature of it makes the code a lot better than the mess that is the M$ code. Yes, Windows is pretty on the outside (debatable, I know...). But ask any programmer that has done serious work on top of Windows and he will tell you how messy it is under the hood.
/rant_mode off
Sorry if that was harsh...

I wasn't supposed to generalize so sorry for that. I know my way around arch well enough and am not scared of these things, but if it was my first try with arch, I think I would've given up already. But after five years with arch, I know these things come and go

amen · 2011-01-08 10:40:25

I today turn on my computer and saw same error(libata). I reset computer and no error this time.

TigTex · 2011-01-08 16:07:08

Kernel 2.6.37 is now on the testing repo... Do any of you still have this kernel panic with that version? (if you use testing)

acoples · 2011-01-08 17:00:58

Same problem here, I've tried to solve it downgrading udev but it didn't work, the only solution that i've found was downgrading the kernel version

Daniel_F · 2011-01-08 17:42:22

To everyone: do you see a line similar to this one:

EIP is at ioread32_rep+0x38/0x50

or this one:

EIP: [<c14c4bd8>] ioread32_rep+0x38/0x50 SS:ESP 0068:f6badb68

in your Oops report ? Because if you don't then you have a different problem !!

When doing a "Me too !" post please confirm that you see this.

@Tig Tex: According to reports on the thread referred to by sm4tik, 2.6.37 still has the problem (not sure if they were talking about final or RC though).

Last edited by Daniel_F (2011-01-08 23:45:34)

madek · 2011-01-08 18:46:56

also have the same problem. with udev 165-1, downgrade to 164-3 solved the problem for me

Daniel_F · 2011-01-08 21:49:55

I've compiled a patched version of udev 165 according to post #36 of the thread referred to by sm4tik.

If anyone wants to try it:

$ sudo pacman -U http://www.electron2k.com/udev165arch/udev-165-1-i686.pkg.tar.xz

not sure if it's necessary, but I did this afterwards:

$ sudo mkinitcpio -p kernel26

As always: No guarantees ! Use it at your own risk !

EDIT [1]: According to Deadguy1 in post #38, running mkinitcpio is not necessary ! (doesn't hurt though)

EDIT [2]: This is just for 32bit (i686) systems. NOT for 64bit (x86_64) ! (I shouldn't have to say this but....)

EDIT [3]: rabid_works in post #54 says mkinitcpio is necessary ! So I'd say do it, just to be safe.

Last edited by Daniel_F (2011-01-10 18:47:38)

Painless · 2011-01-08 23:24:50

Unfortunately I can report the same problem. Doesn't happen every boot, and once up the system seems stable.

Dell Vostro 1700
Intel T7700
GeForce 8600M GT
Fujitsu HDD

Kilzool · 2011-01-09 06:07:44

I had the same problem, and suspected UDEV also. What I did to try to resolve it was just doing > sudo mkinitcpio -p kernel26

We'll see if that resolves it.

I had the problem intermittently upon booting, and once with inserting dvd / cd - and umounting it.
I never got any text or messages.
It froze the computer and both the cap lock key light and scroll lock key light flash.

I couldn't find any messages in any logs.

When the system booted sometimes, it would give out that screen that another user posted in this forum.
Again, there was no log I could find with any data in it and the cap lock key light and scroll lock key light would flash and freeze the system.

Hopefully we can track this down.. I wish I could supply some logs and further details.

Gullible Jones · 2011-01-09 16:28:05

Hmm, there's a new version of the LTS kernel out... Has anyone affected upgraded to that yet?

scarecrow · 2011-01-09 16:48:58

Same sort of problems here. Resolved (...well, no issues up to now) by installing 2.6.37 from testing.

Last edited by scarecrow (2011-01-09 16:49:18)

Deadguy1 · 2011-01-09 19:54:28

Daniel_F wrote:

I've compiled a patched version of udev 165 according to post #36 of the thread referred to by sm4tik.
If anyone wants to try it:
$ sudo pacman -U http://www.electron2k.com/udev165arch/udev-165-1-i686.pkg.tar.xz
not sure if it's necessary, but I did this afterwards:
$ sudo mkinitcpio -p kernel26
As always: No guarantees ! Use it at your own risk !

I too had the random kernel panics as described in this thread and in this bug report
I just wanted to confirm that Daniel_F's patched version of udev 165 (above) fixed both the random panics and dvd insertion panics for me.
also, it worked without doing "sudo mkinitcpio -p kernel26"

thanks Daniel_F

Cheers,

Deadguy1

akng · 2011-01-09 23:10:47

Same thing happening to me, nothing in logs, screen covered in the same text as rabid_works' screenshot.

No hardware similarities, install is only a couple of days old. Before finding this thread, I thought it had something to do with system encryption with lvm over luks.

beerad · 2011-01-10 00:46:27

I'm seeing this panic at boot intermittently as well. I see two different panic signatures: one has a long stack dump where any useful info has scrolled off the screen, and one has a shorter stack dump where the eip is ioread32_rep+0x38 and the Kernel panic is "not syncing : Fatal exception in interrupt". I'm pretty sure both signatures are caused by the same bug; they both happen at about the same point in the bootstrap, just after the video driver has switched modes.

I downgraded udev to 164-3 from the ARM and booted a few times with no problems yet.

Daniel_F · 2011-01-10 01:45:17

Some relevant quotes from the linuxquestions.org thread already mentioned:

rworkman wrote:

Kay Sievers (udev maintainer) says that this is definitely a kernel bug, even though the change to ata_id (in udev) is what triggered it.

resonance wrote:

The problem is probably in the kernel scsi/sg code, specifically supporting "ATA pass-through" functionality. What's new in udev-165 is a function "disk_identify_packet_device_command" which tries a scsi SPC-4 ATA 16-bit pass-through command to identify a cd/dvd drive, and if that fails, tries an SPC-3 version of the command. I believe it is the version 3 attempt which causes the oops; commenting out line 270 of extras/ata_id/ata_id.c appears to eliminate the panic.

That's what I did.

BTW: My patched UDEV had about a dozen downloads so far and only one feedback (thanks Deadguy1 !). Come on people: if it works for you (or if it doesn't !) then at least let us know !

bones · 2011-01-10 03:51:16

Downloaded and testing at the moment. Will get back with the results in a few hours but so far so good.

Gullible Jones · 2011-01-10 05:11:03

Well I have mixed news... The bad news is that the bug has hit Ubuntu. The good news is that Ubuntu has come out with a kernel and a new version of udev, which I'm hoping will fix this. I'll see where this goes.

beerad · 2011-01-10 05:58:35

I patched ata_id.c as resonance suggested, and rebooted 8 times with no panics. It was taking me 3-4 boots to get the system up, so it looks like it eliminated the panics.

bones · 2011-01-10 07:07:39

Well I'll go out on a limb here and say that for me the problem is fixed.
Thanks Daniel_F

zebulon · 2011-01-10 07:14:30

I also observe random crashes (the machine completely freezes) and am wondering if it is the same bug. Do you observe it during boot time only, or any time?

Daniel_F · 2011-01-10 07:15:24

@Gullible Jones

Gullible Jones wrote:

...Ubuntu has come out with a kernel and a new version of udev...

Would you mind posting some references to that ?

@bones

Glad to hear it worked for you. Thanks for reporting !

@zebulon

Yes, it can happen at any time. People have seen it more frequently at boot or when accessing a CD/DVD drive.

Last edited by Daniel_F (2011-01-10 08:27:14)

arctos · 2011-01-10 07:51:07

I've got 2 kernel panics and 1 random crash.

Daniel_F · 2011-01-10 08:18:31

@arctos

arctos wrote:

I've got 2 kernel panics and 1 random crash.

Is it related to the bug we are dealing with in this thead, or you just felt like sharing ??

Because if it is related then your post is missing a lot of information to be helpful or even relevant...

Sorry for being rude, but it would be great if people could think before posting !!

Last edited by Daniel_F (2011-01-10 08:37:10)

archman-cro · 2011-01-10 09:35:38

Here's my report:

This is completely random, although a lockup chance on my setup is fairly big. I see the ioread32_rep+... entry when I'm trying to reboot into the stock ARCH kernel 2.6.36.2.
I don't see it when I'm trying to reboot into a BFS-patched kernel (well, maybe it passed by the screen and I couldn't see it). Any way to catch the logs?

My setup:

HP nx6310 ey588es
Intel Celeron M 1.73GHz
1.2 GBRAM
Hitachi 60GB HDD:
description: ATA Disk
                product: HTS541060G9SA00
                vendor: Hitachi
                physical id: 0
                bus info: scsi@0:0.0.0
                logical name: /dev/sda
                version: MB3O
                serial: MPBCL5XGK9GSJT
                size: 55GiB (60GB)
                capabilities: partitioned partitioned:dos
                configuration: ansiversion=5 signature=41de41dd

Update: The lockup doesn't happen when I downgrade to udev 164-3 and try to boot into the stock ARCH kernel, but still happens if I try to boot into bfs-patched kernel. The pics are below...

http://i.imgur.com/L0UYY.jpg
http://i.imgur.com/OwVYs.jpg

Last edited by archman-cro (2011-01-10 10:17:39)

Arch Linux

#26 2011-01-08 06:15:17

Re: Latest udev causes kernel panics (something about libata)

#27 2011-01-08 08:40:34

Re: Latest udev causes kernel panics (something about libata)

Q) Why would I not want to use Arch?

#28 2011-01-08 10:40:25

Re: Latest udev causes kernel panics (something about libata)

#29 2011-01-08 16:07:08

Re: Latest udev causes kernel panics (something about libata)

#30 2011-01-08 17:00:58

Re: Latest udev causes kernel panics (something about libata)

#31 2011-01-08 17:42:22

Re: Latest udev causes kernel panics (something about libata)

#32 2011-01-08 18:46:56

Re: Latest udev causes kernel panics (something about libata)

#33 2011-01-08 21:49:55

Re: Latest udev causes kernel panics (something about libata)

#34 2011-01-08 23:24:50

Re: Latest udev causes kernel panics (something about libata)

#35 2011-01-09 06:07:44

Re: Latest udev causes kernel panics (something about libata)

#36 2011-01-09 16:28:05

Re: Latest udev causes kernel panics (something about libata)

#37 2011-01-09 16:48:58

Re: Latest udev causes kernel panics (something about libata)

#38 2011-01-09 19:54:28

Re: Latest udev causes kernel panics (something about libata)

#39 2011-01-09 23:10:47

Re: Latest udev causes kernel panics (something about libata)

#40 2011-01-10 00:46:27

Re: Latest udev causes kernel panics (something about libata)

#41 2011-01-10 01:45:17

Re: Latest udev causes kernel panics (something about libata)

#42 2011-01-10 03:51:16

Re: Latest udev causes kernel panics (something about libata)

#43 2011-01-10 05:11:03

Re: Latest udev causes kernel panics (something about libata)

#44 2011-01-10 05:58:35

Re: Latest udev causes kernel panics (something about libata)

#45 2011-01-10 07:07:39

Re: Latest udev causes kernel panics (something about libata)

#46 2011-01-10 07:14:30

Re: Latest udev causes kernel panics (something about libata)

#47 2011-01-10 07:15:24

Re: Latest udev causes kernel panics (something about libata)

#48 2011-01-10 07:51:07

Re: Latest udev causes kernel panics (something about libata)

#49 2011-01-10 08:18:31

Re: Latest udev causes kernel panics (something about libata)

#50 2011-01-10 09:35:38

Re: Latest udev causes kernel panics (something about libata)

Board footer