On Jun 28, 2007, at 11:55 AM, Murali Vilayannur wrote:
Sam,
I misread your earlier email.. sorry :)
I did not realize that you were also facing the same issue.
Did you see any errors on module unload?
So module load/unload works the first time, 2nd time it oopses right?
Right. Its not reproducible every time, but it seems to happen after
the fs has been mounted for a while, the module gets unloaded, and on
reload it oopses. No errors in dmesg on the unload -- I think the
pvfs2-client may be dying, which is why the module is getting
unloaded in the first place. I guess I'm wondering if the cleanup
path on the unload is somehow skipping the kmem_cache_destroy calls...
-sam
Murali
On 6/28/07, Sam Lang <[EMAIL PROTECTED]> wrote:
The one I sent earlier? The bits from PVFS are:
0xc00000005ab57a40 0xc00000000009a0a8 .kmem_cache_create +0x524
0xc00000005ab57b30 0xd0000000003f75a0 .op_cache_initialize +0x3c
0xc00000005ab57bc0 0xd00000000040085c .pvfs2_init +0x58
0xc00000005ab57c70 0xc000000000084cd4 .sys_init_module +0x23c
0xc00000005ab57d10 0xc00000000001045c .ret_from_syscall_1 +0x0
-sam
On Jun 28, 2007, at 11:30 AM, Murali Vilayannur wrote:
> Sam,
> Can you send the oops? This is something new, isn't it?
> I havent seen any such reports go by. Did I miss something?
> Murali
>
>> Yeah looks good. Since we're on the topic, any ideas why we would
>> see an oops on a reload of the kernel module that I mentioned?
>> -sam
>>
>> > -- Pete
>> > _______________________________________________
>> > Pvfs2-users mailing list
>> > [email protected]
>> > http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
>> >
>>
>>
>
_______________________________________________
Pvfs2-users mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
