tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]


On Tue, Dec 15, 2009 at 08:44:26AM +0000, Michael van Elst wrote:
> Which means, we are seeking support to access 2 more sectors

no, only the last sector :)

> for broken (or say: nonstandard) drives that are artificially
> limited to work with broken controllers.
> Do you know a way to detect such a controller?

unfortunably, no. The only way is to try.
It's probably safe to assume all SATA controllers are LBA48-capable, as well as
all ATA controllers designed after 2000.

> Seems to be a
> more reasonable approach than to deduce behaviour from
> a rather imprecise standard.
> N.B. for the specific drives you described a simple quirk would be
> to ignore the LBA48 capability if a drive reports a standard <= ATA-5
> and capacity <= 2^28. The decision logic for LBA48 transfers
> would simply not be used and everything would be addressed
> by LBA28.

Or just look at the content of words 60-61 in identify and use it
as the LBA48 threshold. I think this would cover all cases.

Manuel Bouyer <>
     NetBSD: 26 ans d'experience feront toujours la difference

Home | Main Index | Thread Index | Old Index