On Thu, Mar 08, 2007 at 09:20:48AM -0800, Bill Stouder-Studenmund wrote:
> On Thu, Mar 08, 2007 at 05:54:51PM +0100, Manuel Bouyer wrote:
> > On Thu, Mar 08, 2007 at 04:46:41PM +0000, David Laight wrote:
> > > 
> > > Doesn't that have the effect of always trying to do LBA48 transfers
> > > on disks that appear to support LBA48 ?
> > 
> > only if they are strictly larger than 128GB.
> > 
> > > 
> > > There are two problems:
> 
> [snip]
> 
> > > 2) There is a performance penalty for using LBA48 transfers.
> > > IIRC that is why 28 bit transfers are done whenever possible.
> > 
> > Long term I think we want to alway use LBA48 anyway, to have single transfers
> > larger than 128k. Right now we're limited to 64k because of MAXPHYS but this
> > is something that has to die. The performance penalty of such small
> > transfers is much highter than LBA48 ...
> 
> wd is one of the main reasons attempts to change MAXPHYS have stunbled.

wd (at last with IDE disks, maybe older drives are more limited) can
handle up to 128k (inclued). But I don't think we want a global limit,
the limit has to be per device. But that's another story.

-- 
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
     NetBSD: 26 ans d'experience feront toujours la difference
--