[Nottingham] Various kernel oops

[Martin]

OK, so these are my first kernel oops...

Anyone seen anything like this and any good (or quick!) fixes?


The system uses a Gigabyte GA-MA790FX-DS5 motherboard with 8 GBytes RAM
and sata HDDs.

It fully passes memtest86+ and HDD diagnostic tests.

(Of course) I've installed Mandriva:
2.6.24.4-server-3mnb #1 SMP Mon May 5 20:28:22 EDT 2008 x86_64 AMD
Athlon(tm) 64 X2 Dual Core Processor 6000+ GNU/Linux

And I get various oops listed in /var/log/messages, for example:

May 25 04:02:22 server kernel: Unable to handle kernel paging request at
ffff8100bca02000 RIP:
May 25 04:02:22 server kernel:  [<ffffffff80356f97>] clear_page_c+0x7/0x10
May 25 04:02:23 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 04:02:23 server kernel: Oops: 000b [4] SMP
May 25 04:02:23 server kernel: CPU 1

May 25 04:02:23 server kernel: Unable to handle kernel paging request at
ffff8100bca03000 RIP:
May 25 04:02:23 server kernel:  [<ffffffff80356fe5>] copy_page_c+0x5/0x10
May 25 04:02:23 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 04:02:23 server kernel: Oops: 000b [5] SMP
May 25 04:02:23 server kernel: CPU 0

May 25 04:02:23 server kernel: Unable to handle kernel paging request at
ffff8100bca0c000 RIP:
May 25 04:02:23 server kernel:  [<ffffffff80356fe5>] copy_page_c+0x5/0x10
May 25 04:02:23 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 04:02:23 server kernel: Oops: 000b [6] SMP
May 25 04:02:23 server kernel: CPU 0

And a few more, all for logrotate and msec.

May 25 22:19:47 server kernel: Unable to handle kernel paging request at
ffff8100bca10000 RIP:
May 25 22:19:47 server kernel:  [<ffffffff80357720>] memset_c+0x20/0x30
May 25 22:19:47 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 22:19:47 server kernel: Oops: 000b [1] SMP
May 25 22:19:47 server kernel: CPU 1

May 25 22:19:51 server kernel: Unable to handle kernel paging request at
ffff8100bca11000 RIP:
May 25 22:19:51 server kernel:  [<ffffffff80356f97>] clear_page_c+0x7/0x10
May 25 22:19:51 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 22:19:51 server kernel: Oops: 000b [2] SMP
May 25 22:19:51 server kernel: CPU 1

Those were for whilst using tar for a hdd copy over to a USB external hdd.

May 25 22:20:01 server kernel: Unable to handle kernel paging request at
ffff8100bca2f000 RIP:
May 25 22:20:01 server kernel:  [<ffffffff80356f97>] clear_page_c+0x7/0x10
May 25 22:20:01 server kernel: PGD 8063 PUD b063 PMD c3d50000bca001e3
May 25 22:20:01 server kernel: Oops: 000b [3] SMP
May 25 22:20:01 server kernel: CPU 0

Those were whilst in some python tool.


And I've various other examples. Most seem to involve "pdflush" during
hdd activity. Or is it all a problem of memory paging or filesystem paging?

Turning off all swap space seems to reduce the frequency of the oops.

Is this known?
Any fix?
What next?

Thanks,
Martin

-- 
----------------
Martin Lomas
martin at ml1.co.uk
----------------