Re: [oops] XFS? 2.6.3 build with gcc 3.3.3 running on Opteron

[Chris Meadors]

On Tue, 2004-02-24 at 15:17, Chris Meadors wrote:
> Just got this oops, looks like it might be XFS related.  But I'm no good
> at decoding these things.
> 
> As the subject says this is a plain vanilla 2.6.3 kernel built with the
> just released gcc 3.3.3.  All file systems are XFS.

Just got another one.  I'll this time it went on for a couple tumbles. 
I'll just post the first oops.  I can get you any more info you need.

Some more details about the machine and kernel.  It is a dual Opteron,
SMP kernel, without preempt.  It was plenty stable compiling for hours
at a time.  But now with a real active load on it, it has given up twice
in 24 hours.

I'm going to try the current (2.6.3-1) X86-64 patchkit from Andi Kleen
to see if that helps any.

 

Bad page state at free_hot_cold_page
flags:0x10000000000014 mapping:0000000000000000 mapped:1 count:0
Backtrace:

Call Trace:<ffffffff8013e6fb>{bad_page+75} <ffffffff8013ee9f>{free_hot_cold_page+111}
       <ffffffff8013f530>{__pagevec_free+32} <ffffffff801449db>{release_pages+331}
       <ffffffff80154734>{free_pages_and_swap_cache+116} <ffffffff801489c7>{zap_pte_range+503}
       <ffffffff80148ae6>{zap_pmd_range+198} <ffffffff80148b50>{unmap_page_range+80}
       <ffffffff80148cb1>{unmap_vmas+305} <ffffffff8014d772>{exit_mmap+290}
       <ffffffff80123548>{mmput+88} <ffffffff801278cc>{do_exit+476}
       <ffffffff80127bdb>{do_group_exit+235} <ffffffff80130db0>{get_signal_to_deliver+1152}
       <ffffffff8010e8fd>{do_signal+125} <ffffffff80110b0f>{do_general_protection+159}
       <ffffffff8010f1e8>{retint_signal+62}
Trying to fix it up, but a reboot is needed


-- 
Chris

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/