Hi Greg,

"Greg A. Woods" wrote:
>               -------Sequential Output-------- ---Sequential Input-- --Random--
>               -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
> Machine    MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU  /sec %CPU
> proven    100  4017 24.9  3988  9.5  2293  5.9  8197 53.5  9151 12.8  94.8  2.0

What are the benchmarks like under FreeBSD?

> (I'm very curious to find out why the block writes are slower and the
> rewrite is slower still....  And I thought these drives would be able to
> get a slightly higher seeks/s too....)

Yes, mee too.

[...]
> [[ So, after reading through a bit of dev/ic/aic7xxx*, and then the new
> version of the same from FreeBSD-3.0, I can only ask, again:  When is
> NetBSD going to switch to the FreeBSD CAM stuff? ]]

Is this really the question? Or is it first: Are we going to switch to
FreeBSD CAM stuff? If this has been decided by core/whoever is
responsible
for this decision, it's more likely that people step up to say, hey
if you guys want CAM in NetBSD, I'm glad to invest some time to port
it. Justin has already offered to support anyone trying to do so, but
to me it hasn't been clear that core wants CAM (Sounds like a good
architecture
to me, but then again, I haven't had a close look at it).

So, to make a long story short:

Do we want CAM in NetBSD???

> (A virtual high-5 to the first person to identify the make and model of
> the machine!  ;-)

Not me, but why has the machine on 64 MB RAM. This looks like it might
become a bottleneck sooner than hd-preformance :-)

  Guenther