Ubuntu
linux-source-2.6.15 package

Kernel problem 2.6.15-2?

Bug #62746 reported by Óscar Rodríguez Ríos
2
Affects Status Importance Assigned to Milestone
linux-source-2.6.15 (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

Binary package hint: linux-image-2.6-k7

After upgrade to 2.6.15-27 I'ving a problem with Raid 5. This is continuosly, and the error is:
"[17593006.932000] scheduling while atomic: md2_raid5/0xffffffff/2423
[17593006.932000] [<c030b79d>] schedule+0x99d/0xd20
[17593006.932000] [<c011f490>] __wake_up+0x40/0x60
[17593006.932000] [<f89570c0>] release_stripe+0xc0/0x190 [raid5]
[17593006.932000] [<c030c425>] schedule_timeout+0x75/0xc0
[17593006.932000] [<f895b132>] raid5d+0x242/0x2f0 [raid5]
[17593006.932000] [<c013c560>] prepare_to_wait+0x20/0x70
[17593006.932000] [<f896af89>] md_thread+0xf9/0x150 [md_mod]
[17593006.932000] [<c013c690>] autoremove_wake_function+0x0/0x60
[17593006.932000] [<f896ae90>] md_thread+0x0/0x150 [md_mod]
[17593006.932000] [<c013c157>] kthread+0xc7/0xd0
[17593006.932000] [<c013c090>] kthread+0x0/0xd0
[17593006.932000] [<c0101509>] kernel_thread_helper+0x5/0xc"

Every I/O access to the raid outputs and error like this (in 12 hours the /var/log/kern.log is 2GB in size) and I've to stop syslod and klogd. Not data corruption (at the moment) but I'm scared about.

After reboot the machine the raid is ok, but now a java process outputs and similar error, related with scheduling and futexes:

[17203784.544000] scheduling while atomic: java/0xffffffff/6795
[17203784.544000] [<c030b7ad>] schedule+0x99d/0xd20
[17203784.544000] [<c02b8da0>] ip_local_deliver_finish+0x0/0x280
[17203784.544000] [<c012ecb4>] lock_timer_base+0x24/0x50
[17203784.544000] [<c012ed77>] __mod_timer+0x97/0xc0
[17203784.544000] [<c030c40c>] schedule_timeout+0x4c/0xc0
[17203784.544000] [<c012f850>] process_timeout+0x0/0x10
[17203784.544000] [<c014008a>] futex_wait+0x1ea/0x220
[17203784.544000] [<c011f3c0>] default_wake_function+0x0/0x20
[17203784.544000] [<c01403e4>] do_futex+0x74/0xc0
[17203784.544000] [<c0140496>] sys_futex+0x66/0x130
[17203784.544000] [<c01033f7>] sysenter_past_esp+0x54/0x75

Not related with one process in particular, otherwise the sheduling algorithm that schedules to other process when executin a function or code that needs to be atomic (executed at once).
The problem is also with kernel version 2.6.15-26.

Could be hardware related? (mem, i/o controllers...)

Best regards,
neuromancer

See original description
Óscar Rodríguez Ríos (ingorr01)
description: updated
Revision history for this message
Chuck Short (zulcss) wrote :

Re-assigning against proper package

Revision history for this message
Óscar Rodríguez Ríos (ingorr01) wrote :

It's solved. It was a memory error. We have one DIMM not function properly. Sorry for the inconvenience.

Best regards,
neuromancer

Revision history for this message
Gareth Fitzworthington (mapping-gp-deactivatedaccount) wrote :

This bug has had no activity for a considerable period. This is a check to see if there is still interest in investigating this bug report.
Appears to have been caused by a hardware fault.
Marking "Invalid"

Changed in linux-source-2.6.15:
status: New → Incomplete
status: Incomplete → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.