Re: LBA48 bug and the wd_quirk_table in wd.c for 2.5" drives

[Bernd Ernesti <netbsd%lists.veego.de@localhost>]

On Mon, Oct 13, 2008 at 07:14:19PM +0100, Patrick Welche wrote:
> On Sun, Oct 12, 2008 at 11:52:27PM +0200, Manuel Bouyer wrote:
> > On Sun, Oct 12, 2008 at 11:45:03PM +0200, Bernd Ernesti wrote:
> > > On Sun, Oct 12, 2008 at 11:29:08PM +0200, Manuel Bouyer wrote:
[..]

> > > > Did your kernel report a "Id not found" error, or something else ?
> > > 
> > > It reported something about a LBA48 bug, so it should work later on.
> > 
> > OK, that what is expected then.
> 
> I commented out the explicit quirk list, and all is working happily with
> disks which matched that list, i.e., the generic code works well...

That was not the point of my mail.

First I thought that my disks were bad when I got a kernel message and was
about to get the drives replaced.
Using the generic code path also adds also some retries on the disk level.

I prefer to not get hit by a kernel error message about my disks when
it hits the affected area.

I should have known why i don't like seagate drive so much, that 'bug'
and the broken s.m.a.r.t. implementation.

Bernd