Manuel Bouyer wrote:
> On Tue, May 09, 2006 at 01:20:09AM +1000, Carl Brewer wrote:
>> Manuel Bouyer wrote:
>>> On Tue, May 09, 2006 at 12:49:20AM +1000, Carl Brewer wrote:
>>>> Manuel Bouyer wrote:
>>>>
>>>>> Hum, it may be another issue here: 268435392 is not the "magic" number.
>>>>> And this drive seems to be an older revision, which shouldn't have the
>>>>> problem.
>>>>>
>>>>> But this request is crossing the 128Gb barrier, this can be an issue
>>>>> (and I would understand a drive bailing on this). Carl, can you try the
>>>>> attached patch ?
>>>> I would, except I already hacked at wd.c and added the ID for my
>>>> drives for the LBA48 quirk, which seems to have worked.
>>> Yes, this work, but this is not what will be commited because it may
>>> break things for others.
>> Understood.  The box is now almost ready for production use, and I
>> don't want to break it, can I install your patch and then test
>> easily without killing what I've installed on the machine? I can
>> break the mirror by pulling out one of the drives if there's any
>> risk?
> 
> Yes, either pull the drive out or mark it failed: raidctl -f /dev/wd1a raid0
> 
> I would test it myself, but I don't have drives larger than 128GB connected
> to a plain IDE controller.

Ok, I'll test it tomorrow, it's bedtime :)

Any suggestions for a test?  The machine will boot off the
disk without the LBA48 hack, what's a good way to see if the
fix works?