FS#20525 - [kernel26] 2.6.35.2 causes weird errors (possibly related to lvm) on boot

Attached to Project: Arch Linux
Opened by jeremy mills (drama) - Thursday, 19 August 2010, 20:26 GMT
Last edited by Thomas Bächler (brain0) - Friday, 27 August 2010, 22:10 GMT
Task Type Bug Report
Category Upstream Bugs
Status Closed
Assigned To Tobias Powalowski (tpowa)
Thomas Bächler (brain0)
Architecture All
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 4
Private No

Details

When i installed i setup lvm for all partitions including root. This was on kernel 2.6.34. Everything worked wonderfully. Then yesterday i upgraded to 2.6.35.2 (i always use packages from stable and never from testing). However after i rebooted i noticed some errors come across the screen

Internal error: Maps lock 14217216 < unlock 14221312
Internal error: Maps lock 14221312 < unlock 14225408
Internal error: Maps lock 14225408 < unlock 14229504
Internal error: Maps lock 14229504 < unlock 14233600
Internal error: Maps lock 14233600 < unlock 14237696
Internal error: Maps lock 14237696 < unlock 14241792

This happens right after it activates the volume groups "vgchange -ay"

My computer boots fine but when i see something that says "internal error" especially when it is related to my discs i start to think " hey this cant be good".

Im also seeing alot of "EXT4-fs (dm-x): mounted filesystem with ordered data mode. Opts: (null)" Im not sure if that means anything or even if its a problem. Ive attatched dmesg,kernel,everything,and messages.log files.

Also i can confirm that downgrading back to kernel26-2.6.34.3-1 makes the message go away.

If you need anything else dont hessitate to ask.

P.S.

Maybe this commit would fix it

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=d7824370e26325c881b665350ce64fb0a4fde24a

This task depends upon

Closed by  Thomas Bächler (brain0)
Friday, 27 August 2010, 22:10 GMT
Reason for closing:  Fixed
Additional comments about closing:  The errors are gone with newer kernels, let's consider this problem solved.
Comment by jeremy mills (drama) - Friday, 20 August 2010, 15:50 GMT
ok so late last night i decided to see if i could fix this on my test box. i recompiled kernel26-2.6.35.2 from abs. but i added the patch i referenced to originally. in the PKGBUILD i added it right after the arch patches. it appeared to compile fine. i then rebooted and observered the boot messages on my screen and the error is now gone. so it appears that would be the fix. i say appears because im not 100% certain its fixed. its possible the the patch just hides the messages and i dont know how to find out for sure. there is nothing in the logs referenceing the "internal error" but then again there never was to begin with. im not sure of the arch policy on patching though. some people say its not allowed at all other people say that it is allowed as long as the patches come directly from upstream (which is where it came from). on another not i looked in the logs and noticed messages saying that my partitions were remounted after gdm started. is this normal? ive never noticed this before. ill upload updated logs as well as a copy of the patch from upstream and the PKGBUILD i used if anyone wants give it a try. it will be a few hrs though because im not at home at the moment (posting from my phone).
Comment by jeremy mills (drama) - Friday, 20 August 2010, 21:35 GMT
Ok so i just noticed that new kernel was released upstream just a few hours ago (2.6.35.3) which includes 2 patches related to "stack guard". the patch i referenced as well as one other plus a patch that fixes a vmware build error (another problem ive been having). I am compileing the kernel from upstream as i type. Ill report back but im almost positive once the updated kernel hits our repos this bug can be closed.
Comment by Nicolas Brunet (nsb-fr) - Sunday, 22 August 2010, 18:13 GMT
Just updated to 2.6.35.3-1 and still having one Internal error at boot. Had four before the update. Still nothing about the error in logs.

Config: i686 with /, /boot, /home and swap on lvm
Comment by Feifei Jia (ffjia) - Monday, 23 August 2010, 05:15 GMT
Nicolas, I have the same config except arch x86_64. After upgrading to 2.6.35.3-1, the error messages are totally gone, here.
Comment by Nicolas Brunet (nsb-fr) - Monday, 23 August 2010, 17:19 GMT
Well, since there is no or little documentation on this error I can't really tell if it's my lvm setup that was damaged by the previous bug or if the bug is still there. Is there any tool to check the status of a lvm VG/LV ?

Edit: Sort of solved for me. The lvm setup came from my previous laptop wich had 2 HD. The new don't and using lvm over a single HD is a bit useless for me so I am reinstalling without lvm.

Loading...