Re: [PATCH] sata_mv: Fix broken Marvell 7042 support.

[Mark Lord <liml@rtr.ca>]

Mark Lord wrote:
> Hein-Pieter van Braam wrote:
>> I zero'd all the disks, and now they show up as 'new' in the highpoint
>> BIOS again... so, I guess there IS some reserved part somewhere that's
>> accessible with the sata_mv driver.
> ..
> 
> Yeah, that's my suspicion too.. they must have just simply moved
> the metadata to near the end instead of sector-8.
> 
> Mmm...  I'll try a smaller drive here,
> and zero the entire drive, then reboot with it
> on the Highpoint board, and then see what sector it modified.
> 
> This'll take a while.
..

Actually, it doesn't take long at all,
with some creative use of fdisk.

And I haven't yet figured out their logic,
but here's what I've found.

They write their RAID metadata near-ish to the end of the drive.
On my 320GB drives, it ended up at about -199853 sectors from
the end of the drive.  I have no idea what logic leads to them
choosing such a peculiar location for it.

That's NOT an even cylinder or anything else I can figure out.

But if you leave the last 100KB of the drive unallocated 
for any partitions, then your data might survive.

This needs more research, and probably another patch of some kind
in the driver for #upstream-fixes.  I'm busy with a course all week,
this won't happen (from me) until the weekend.

What a screwy BIOS.

-ml