[nycbug-talk] Reproducible data corruption on 6.1-Stable [Long but please read]

George Georgalis george at galis.org
Thu Sep 14 09:09:04 EDT 2006


On Thu, Sep 14, 2006 at 08:20:47AM -0400, Jonathan Stewart wrote:
>George Georgalis wrote:
>> On Wed, Sep 13, 2006 at 07:00:52PM -0400, Jonathan Stewart wrote:
>>> ad4: 305245MB <Seagate ST3320620AS 3.AAC> at ata2-master UDMA133
>>> ad6: 305245MB <Seagate ST3320620AS 3.AAC> at ata3-master UDMA133
>> 
>> Sorry I didn't take a close look at your issue, but please check
>> sector 0xfffffff and 0x10000000.
>> 
>> the on their larger (?) drives seagate doc says use LBA32 for
>> sectors <= 0xfffffff and LBA48 for sectors >0xfffffff when the
>> drives actually require LBA32 < 0xfffffff and LBA48 for sectors
>> =>0xfffffff
>> 
>> I haven't tested yet, but I think these numbers are correct. In
>> netbsd many seagate drives are in a quirk table, but last I
>> checked my 400 and 500Gb ones where not.
>> 
>
>Did you mean doing something like
>sudo dd if=/dev/ad4 of=/tmp/temp skip=268435455 count=1
>sudo dd if=/dev/ad4 of=/tmp/temp skip=268435456 count=1
>sudo dd if=/dev/ad4 of=/tmp/temp2 skip=268435456 count=1
>
>to make sure I can read the sectors and the reads come back the same
>each time?  If so then yes to both.


not sure about the base change
or your device names, but
 if=/dev/rwd0d skip=0xffffffe count=3
is what I had in mind, is ad4 a raw device?

// George


-- 
George Georgalis, systems architect, administrator <IXOYE><



More information about the talk mailing list