Latest udev causes kernel panics (something about libata)

Gullible Jones · 2011-01-04 16:50:55

So I just updated:
- pciutils
- udev
- device-mapper
- cryptsetup
- groff
- lvm2

And rebooted. Shortly after boot I was met with a stack trace (but no kernel panic). Several tries revealed that my default initcpio causes a kernel crash every boot, but the fallback one works just fine (thank you, Arch devs!).

My mkinitcpio.conf has these modules

MODULES="ata_piix sd_mod ext3"

And these hooks

HOOKS="base resume"

And nothing else. Anyone know what might have caused this problem? I suspect it's the new udev (that was where the kernel crashed IIRC, I wasn't able to get a log) but I'm not sure.

Edit: yup it's the new udev, the crash occurs on processing uevents.

Last edited by Gullible Jones (2011-01-05 03:36:59)

eldragon · 2011-01-04 18:48:15

hmm, have you tried rebuilding your boot image?

filed a bug report?

tavianator · 2011-01-04 18:52:41

It's very odd that the fallback image would behave differently than the default one in this case, since all the fallback one does differently is skip the autodetect hook, which you don't even use.

If you can copy/write down/photograph the stack trace that would help.

skunktrader · 2011-01-04 19:04:04

The fallback image probably includes udev in the HOOKS list

Gullible Jones · 2011-01-04 19:21:13

Well that was most strange. The problem has vanished now. I think rebuilding my initrd did the trick.

(FWIW, the stack trace mostly showed stuff about libata.)

Edit: or not. I just got a kernel panic, completely at random. More stuff about libata and something about the swap partition being tainted(?). Again, I wasn't able to get a log.

Edit again: it was "swapper tainted", not the swap partition.

Last edited by Gullible Jones (2011-01-05 03:47:42)

andrews · 2011-01-06 00:05:20

I have the same kernel panic. At this moment I dont know nothing about this, is random , the system works fine if the boot is ok.

Edit : On my desktop after a normal boot I put a dvd in the drive and kernel panic again.
Downgrading now to udev 164.3

Last edited by andrews (2011-01-19 06:14:04)

clu · 2011-01-06 02:03:49

I'm having the same problem -- and even worse -- it's intermittent. Has anyone filed a bug report or found a solid solution? Running mkinitcpio again does *not* solve it for me.

TigTex · 2011-01-06 02:40:04

Maybe all of you have the same hardware (chipset)... Can you post your system specifications? (motherboard / cpu / hdd). Because I'm using my arch instal in several pc's and I don't have any kernel panic to report...

clu · 2011-01-06 04:18:05

I'm on a macbook pro that's not the latest.

cpu: Intel core 2

disk:
ATA device, with non-removable media
Model Number: Hitachi HTS541616J9SA00
Serial Number: SB2421SJH8YGNB
Firmware Revision: SB4AC7CP

Based on the libata stuff I'm guessing it has something to do with the disk...

soundman · 2011-01-06 10:50:54

I have same problem both laptop and desktop computer. It's random, after several resets booting normally.

KKata · 2011-01-06 13:16:21

Hi!

I had the same issue today on my dekstop computer (Acer eMachine). Upgraded udev on 4th Jan., that is 2 days ago. In the meantime there was nothing wrong, but today a random panic, with the libata stuff described above. I tried to boot in 3 times, then I used the fallback image which worked well, and then I could boot in normally with the default Arch image again.

I also have a Hitachi sata hdd, maybe that's the issue.

pacman.log

[2011-01-04 17:40] Running 'pacman -Syu'
[2011-01-04 17:40] synchronizing package lists
[2011-01-04 17:40] starting full system upgrade
[2011-01-04 17:45] upgraded pciutils (3.1.7-1 -> 3.1.7-2)
[2011-01-04 17:45] upgraded udev (164-3 -> 165-1)
[2011-01-04 17:45] upgraded device-mapper (2.02.78-1 -> 2.02.79-1)
[2011-01-04 17:45] upgraded cryptsetup (1.1.3-1 -> 1.2.0-1)
[2011-01-04 17:45] upgraded gnome-media (2.32.0-2 -> 2.32.0-3)
[2011-01-04 17:45] upgraded groff (1.20.1-4 -> 1.21-1)
[2011-01-04 17:45] upgraded gtk-doc (1.15-2 -> 1.15-3)
[2011-01-04 17:45] upgraded libmysqlclient (5.1.51-1 -> 5.1.54-1)
[2011-01-04 17:45] upgraded lvm2 (2.02.78-1 -> 2.02.79-1)
[2011-01-04 17:45] upgraded mysql-clients (5.1.51-1 -> 5.1.54-1)
[2011-01-04 17:45] upgraded mysql (5.1.51-1 -> 5.1.54-1)
[2011-01-04 17:45] upgraded pam (1.1.1-2 -> 1.1.3-1)
[2011-01-05 13:55] Running 'pacman -Syu'
[2011-01-05 13:55] synchronizing package lists
[2011-01-05 13:55] starting full system upgrade
[2011-01-05 14:22] upgraded gcc-libs (4.5.2-3 -> 4.5.2-4)
[2011-01-05 14:22] upgraded pacman-contrib (3.4.0-1 -> 3.4.2-1)
[2011-01-05 14:22] upgraded rrdtool (1.4.4-2 -> 1.4.5-1)

lshw output:

*-disk
                description: ATA Disk
                product: Hitachi HDT72101
                vendor: Hitachi
                physical id: 0.0.0
                bus info: scsi@2:0.0.0
                logical name: /dev/sda
                version: ST1O
                serial: STA1L5MA182KMX
                size: 149GiB (160GB)

mkinitcpio.conf:

MODULES=""
HOOKS="base udev autodetect pata scsi sata filesystems"

rabid_works · 2011-01-06 16:11:04

[EDIT] I downgraded gcc-libs and gcc, since I suspected they were the cause of the problem. The system now boots without an issue without having to opt for the fallback image. Can anyone else confirm this as being the source of the problem ?

Same issue here following an upgrade on the 5th of January. It freezes without fail on every boot and with the same output (seen here in this photo) :

Fallback image also works for me. I haven't rebuilt the boot image as yet (although it doesn't seem to have improved matters for others in this thread from what I can tell). I also haven't downgraded any packages, since the fallback image works fine.

Other info :

pacman.log :

[2011-01-05 13:47] Running 'pacman -Syu'
[2011-01-05 13:47] synchronizing package lists
[2011-01-05 13:47] starting full system upgrade
[2011-01-05 13:48] upgraded atkmm (2.22.1-1 -> 2.22.1-2)
[2011-01-05 13:48] upgraded cloog-ppl (0.15.10-1 -> 0.15.10-2)
[2011-01-05 13:48] upgraded gcc (4.5.2-3 -> 4.5.2-4)
[2011-01-05 13:48] upgraded gcc-libs (4.5.2-3 -> 4.5.2-4)
[2011-01-05 13:48] upgraded rrdtool (1.4.4-2 -> 1.4.5-1)

lshw output:

 *-disk
                description: ATA Disk
                produit: ST3400832A
                fabriquant: Seagate
                identifiant matériel: 0
                information bus: scsi@0:0.0.0
                nom logique: /dev/sda
                version: 3.03
                numéro de série: 5NF098QW
                taille: 372GiB (400GB)
                fonctionnalités: partitioned partitioned:dos
                configuration: ansiversion=5 signature=00057f6b
*-core
       description: Carte mère
       produit: P4S533
       fabriquant: ASUSTeK Computer INC.
       identifiant matériel: 0
       version: REV 1.xx

PS : sorry about the poor photography in the above photo BTW...

Last edited by rabid_works (2011-01-06 23:50:07)

adriaticc · 2011-01-07 02:29:20

Same intermittent problem here. Seagate pata drive, downgraded udev with fingers crossed.

sliposk · 2011-01-07 16:14:08

I have the same problem on 3 Diff Pcs . #1 Dell 3000 P4 3.0 intel 865 , socket 478 Nvidia 6200 #2 Mpc P4 3.0 Intel 865 , socket 478 Nvidia 6200 #3 Nexlink {Micro Star} 6769 Sis 651 P4 2.8 , socket 478 Nvidia 6200 ... NOTE that all Pcs that I have that have this Random problem have similar hardware and the Same Exact Video Card in each PC . I have updated all 3 of the affected machines every 2 days for the past 2 years . This is what I have tryed so far ..... Reinstalled kernel26 , Nvidia Drivers , deleted vboxdrv from mod list . This happened after update on 1-4-11 . It happens on every 3 to 4 boots totaly random . Sorry nothing to log . I will continue to work on this . To all people affected by this issue please post your hardware specs. See pic that rabid_works posted . Same thing here . We might have to wait for new udev to be packaged .

myrlin · 2011-01-07 17:16:24

Same problem here. Dell Dimension 3000 P4 3.0 intel 865, Samsung SATA drive (with SATA/IDE adapter).

Random kernel panic on boot, with message "unable to handle kernel paging request at fffb2000"

I've downgraded udev, and all seems OK so far, so I'm keeping my fingers crossed!

i

defears · 2011-01-07 17:31:04

Same problem but in a virtualbox new arch install image with default settings. But the host computer running virtualbox is fine. Weird. I have to reboot the image 10 times, then it's normal.

Gullible Jones · 2011-01-07 17:56:55

Downgrading the kernel to the LTS version (2.6.32) seems to be a temporary workaround, I downgrade and haven't crashed yet (fingers crossed!).

sm4tik · 2011-01-07 19:28:39

There's an interesting link in this bug report in case someone hasn't seen it yet.
http://www.linuxquestions.org/questions … en-852843/
Looks like the latest udev isn't such a good idea afterall. I thought it was just bad luck with my new hardware, but am glad to see it isn't so. Though, as comet.berkeley in the other forum says:

If I were a new user I would give up on Linux and go back to Windows...
In the old days we used to make fun of Windows because of the "Blue Screen of Death" would occur so often.
This is exactly what I am seeing now with Linux!

Quite harsh words, but still.. Anyways, it seems that downgrading udev to 164 should be a temporary workaround. I'll post more experiences again once I get my networking problem solved.

Last edited by sm4tik (2011-01-07 19:30:03)

rubble · 2011-01-07 19:43:15

Started to install Arch before Christmas - left it for a couple of weeks and got back on it a few days ago, updated everything and hit the same issue as described above. Appears to crash when udev starts. Fallback image works fine though.

AMD AthlonXP 2400+, VIA KT600, WD 500 GB SATA HDD.

some of the package versions installed that were mentioned above:
gcc 4.5.2-4
kernel26 2.6.36.2-1
udev 165-1

Daniel_F · 2011-01-08 00:47:59

@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...

Last edited by Daniel_F (2011-01-08 00:58:59)

Gullible Jones · 2011-01-08 00:59:59

Not sure about code quality (I'm not knowledgeable about it, nor privy to Microsoft's proprietary code), but it looks like this has occured on only two distros, Arch and Slackware. Arch is a great distro IMO, but it's not exactly known for outstanding stability or newbie-friendliness; and in the case of Slack it was in the -current branch, which IIRC is considered potentially buggy. It seems to be limited to more bleeding-edge distros/branches, which novices should not use anyway.

Now that that's done with...

I think I should mention that one of my hard drive partitions just failed spectacularly and unrecoverably, totalling my Arch install. So it has occured to me that this bug might be a sign of impending hardware failure - maybe the newer kernel is less fault tolerant, or something. At any rate, I would strongly encourage everyone suffering from this bug to make sure your backups are up to date, should this hypothesis turn out to be correct!

Edit: False alarm... Kind of. I was able to fsck, chroot in, regenerate the initcpio image, and reboot from the partition, but was met again with the same kernel panic (using the LTS kernel this time). Still torn on whether this is purely a software issue or if there's an underlying hardware problem, I have had trouble with this laptop before...

Last edited by Gullible Jones (2011-01-08 01:17:30)

Daniel_F · 2011-01-08 01:19:10

Gullible Jones wrote:

Not sure about code quality (I'm not knowledgeable about it, nor privy to Microsoft's proprietary code), but it looks like this has occured on only two distros, Arch and Slackware. Arch is a great distro IMO, but it's not exactly known for outstanding stability or newbie-friendliness; and in the case of Slack it was in the -current branch, which IIRC is considered potentially buggy. It seems to be limited to more bleeding-edge distros/branches, which novices should not use anyway.

Glad to see you agree with me on that....

Gullible Jones wrote:

I think I should mention that one of my hard drive partitions just failed spectacularly and unrecoverably, totalling my Arch install.

Ouch !

Are you positive it's a hardware failure ? Or could it be that your data has been corrupted as a side effect of the problem ??

Maybe the devs should consider pulling udev 165 from the repos until a patch is issued....

Gullible Jones · 2011-01-08 01:30:06

No, not positive, especially since I was eventually able to recover stuff from it. Even so.

xsoul · 2011-01-08 03:24:50

too bad downgrading to udev 164-3 does not solve the problem for me, kernel panic still happens from time to time
http://imagebin.ca/view/VQ-Sz72.html

Last edited by xsoul (2011-01-08 03:25:28)

andrews · 2011-01-08 06:10:21

xsoul wrote:

too bad downgrading to udev 164-3 does not solve the problem for me, kernel panic still happens from time to time
http://imagebin.ca/view/VQ-Sz72.html

@xsoul
I don't see any libata error in the log, maybe your kernel panic isn't related to the udev 165-1 bug?
For me after downgrading to 164-3 and tons of reboots, no more panics.

Last edited by andrews (2011-01-08 06:11:39)

Arch Linux

#1 2011-01-04 16:50:55

Latest udev causes kernel panics (something about libata)

#2 2011-01-04 18:48:15

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

#3 2011-01-04 18:52:41

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

#4 2011-01-04 19:04:04

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

#5 2011-01-04 19:21:13

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

#6 2011-01-06 00:05:20

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

#7 2011-01-06 02:03:49

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

#8 2011-01-06 02:40:04

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

#9 2011-01-06 04:18:05

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

#10 2011-01-06 10:50:54

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

#11 2011-01-06 13:16:21

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

#12 2011-01-06 16:11:04

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

#13 2011-01-07 02:29:20

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

#14 2011-01-07 16:14:08

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

#15 2011-01-07 17:16:24

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

#16 2011-01-07 17:31:04

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

#17 2011-01-07 17:56:55

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

#18 2011-01-07 19:28:39

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

#19 2011-01-07 19:43:15

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

#20 2011-01-08 00:47:59

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

Q) Why would I not want to use Arch?

#21 2011-01-08 00:59:59

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

#22 2011-01-08 01:19:10

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

#23 2011-01-08 01:30:06

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

#24 2011-01-08 03:24:50

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

#25 2011-01-08 06:10:21

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

Board footer