On Thu, 14 Oct 2004, Manuel Bouyer wrote:

> Date: Thu, 14 Oct 2004 19:47:15 +0200
> From: Manuel Bouyer <bouyer@antioche.eu.org>
> To: Karl Janmar <karl@utopiafoundation.org>
> Cc: tech-kern@NetBSD.org
> Subject: Re: ATI IXP IDE chipset support
>
> On Wed, Oct 13, 2004 at 07:19:54PM +0200, Karl Janmar wrote:
> >
> > The machine is a P4 Celeron 2.4Ghz,
> > 2 disks,
> > one SATA 120Gig Maxtor 8MB cache,
> > one ATA 120Gig Seagate 8MB cache.
> >
> > Something is not right as I get a performace increase of about 30times
> > with the ATA disk, as compare to the 10times on the SATA disk.
>
> Maybe SATA does a better job when no specific drivers are present.
> How much do you get out of each disk ?

That was what I had first? when there was no driver for the IXP?
Then I got about 2 MB/s even with the SATA disk when using dd from
/dev/zero to the filesystem.

>
> You can also try the attached program:
> ./tst /dev/rwd0d 30000
> (increase the second parameter if it completes in less than 2 seconds)
> This will benchmark the bandwith available between the drive's cache and
> the host (as opposed to the overall drive performances).
>

This is with the SATA disk:
# ./tst /dev/rwd1d 3000
9673491 us, 19.382868 MB/s

This is on the ATA disk:
# ./tst /dev/rwd0d 3000
53188011 us, 3.525231 MB/s

these results are a little bit strange(for me as I am not an expert),
because  I get this when using dd, /mnt is on wd0e (the ATA disk):
# dd if=/dev/zero of=/mnt/apslask bs=4096 count=100000
100000+0 records in
100000+0 records out
409600000 bytes transferred in 17.693 secs (23150398 bytes/sec)

This doesn't ofcourse exacly test the same thing but....

- Karl