3ware rmmod - Unable to handle kernel NULL pointer dereference

[Tru Huynh]

Hello,

We are installing the same hardware here.
Rh7.1 standard install and SGI-xfs.
I have not tried your harddrive swapping but 
I would get the latest firmware and drivers upgrade...

There are some serious issues with raid5 with
version prior to firmware release 6.6 and driver 6.7 

firmware and driver download from:
http://www.3ware.com/support/3warednload_6000_driver.asp
and 3dm v7.11 from:
http://www.3ware.com/support/3warednload_3dm_driver.asp

1)quote from http://www.3ware.com/support/:
" 3ware has determined that there is an IRQ priority conflict with
Intel's 815EEA motherboard as it relates RAID applications.
This is shown when there is a failure to successfully boot
an Escalade array using this motherboard.
NOTE: Intel is aware of the problem and is working on a solution.

3ware recommends the use of the older Intel CA810E board, or any Intel
motherboard using Intel's 100mhz 440BX chipset until this problem is
resolved by Intel."

2) quote from http://www.3ware.com/support/raid5TechBulletin.shtml

"3ware has found a bug in our RAID 5 code that can cause
file system errors resulting in possible loss of data....

3ware products affected: Escalade 6400, 6410 and 6800 
3ware software versions affected: 6.5 and 6.6 
RAID level affected: RAID 5 
Operating systems affected:
        Known: Linux with default ext2 file system"

3) quote from 3w-xxxx.c (the one with RH7.1 is version 1.02.00.004)
   1.02.00.005 - Allocate bounce buffers and custom queue depth for
raid5 for
                 6000 and 5000 series controllers.
                 Reduce polling mdelays causing problems on some
systems.
                 Fix use_sg = 1 calculation bug.
                 Check for scsi_register returning NULL.
                 Add aen count to /proc/scsi/3w-xxxx.
                 Remove aen code unit masking in tw_aen_complete().
   1.02.00.006 - Remove unit from printk in tw_scsi_eh_abort(), causing
                 possible oops.
                 Fix possible null pointer dereference in
tw_scsi_queue()
                 if done function pointer was invalid.

Best regards,

Tru 

Stephen Johnston wrote:
> 
> Hi
> 
> We have a 3ware Escalade 6800 8 port ide card which the o/s sees as
> scis, it works fine, except when you put a new disk onto the card.
> 
> With 8 disks if we
> 
> rmmod 3w-xxxx
> remove 2 disks
> swap them
> modprobe 3w-xxx
> 
> no problems, the card see the disks have been swapped and its fine.
> 
> However if we use a new disk the card didnt know about to begin with
> 
> rmmod 3w-xxxx
> remove a disk
> replace with a new disk
> modprobe 3w-xxx
> 
> there is a segmentation fault with the following error;
> 
> 3w-xxxx: tw_poll_status(): Flag 0x40000 not found.
> Unable to handle kernel NULL pointer dereference at virtual address
> 00000036
>  printing eip:
> e08dc162
> pgd entry ddfb7000: 0000000000000000
> pmd entry ddfb7000: 0000000000000000
> ... pmd not present!
> Oops: 0000
<...>
-- 
Dr Tru Huynh          | Bioinformatique Structurale
mailto:tru at pasteur.fr | tel/fax +33 1 45 68 87 37/19
Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris CEDEX 15 France