> those parameters aren't going to produce a performance benefit.  Why are you
> so certain that this must be some bios configuration problem as opposed to a
> buggy driver or some sort of starvation issue caused by other traffic?

I'm not _certain_ of anything.  I don't think it's starvation caused by
other traffic.  The machine in question is pretty idle.  It serves an X
terminal and as a secondary DNS for a handful of low-traffic domains.
The SCSI bus is quiescent except during nightly backups.  It was
inactive when I was doing basic tests.  The machine has 128MB of RAM
and never uses all of it.

AFAICT, the only parts of the system getting much use during my basic
tests are:
	* General kernel stuff (page mapping, scheduler, etc.)  Mostly
	  idle, really.
	* Network (xterms and ftp test)
	* Disk (in one of the tests)

Normal disk I/O is zippy.  Network I/O without disk I/O is zippy.
Network I/O interleaved with disk I/O is not.

It could be a problem with the pciide driver or with the de driver (or
the combination).  Manuel has a machine that he says is basically the
same (pciide, de, VIA chipset) that does not exhibit these problems,
though.  For me, that tends to point to a BIOS problem, but as I've
said, I don't know PC hardware that well--maybe there are other
possibilities.

I'd be much happier if it is a buggy driver or a solvable starvation
problem that's wholly unrelated to BIOS configuration, actually.  But
then I'd be happy just to know what's going on.  ;-)

-allen