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?

-----Original Message-----
From: Allen Briggs [mailto:briggs@ninthwonder.com]
Sent: Tuesday, October 05, 1999 5:51 AM
To: Adam Glass (Exchange)
Cc: Manuel Bouyer; Andrew Gillham; port-i386@netbsd.org
Subject: Re: de0/pciide0 performance conflict?


> Generally this is not a parameter that an OS should screw with.
> Theoretically these parameters are balanced by the bios to reflect a
latency
> appropriate to the number/type devices in the system.  No MS OS modifies
> this parameter.  You'll get some performance this way but you might starve
> something else.

I don't doubt it.  I was fooling with it mainly to see if it did make
any difference.  I don't really know what to try since this behavior
seems so odd.  It seems that _something_ isn't configured right and I
don't know how to go about figuring out what it could be.  Do you have
any ideas?

The only PnP configuration I have is "Resources controlled by" with
values "Auto/Manual".  I have it set to "Auto", which is supposed to
mean that the "system" (BIOS, I presume) will set it up.  If it's
switched to "Manual", then I can set the IRQ and DMA to either "Legacy
ISA" or "PCI/ISA PnP".  I have no ISA cards in the system.  I have one
PCI Trident video card, one PCI Netgear 10/100 card, and one PCI Asus
SCSI card.  It's a pretty basic system.

It seems like it must be something glaring to kill performance this
badly, but I don't know enough about the PC/PCI/DMA/bus architecture
to make even a reasonable hypothesis.  All I know is that if I take
the pciide drive out and use a SCSI disk, it'll probably run just fine,
but that doesn't solve the problem.  It also doesn't explain why M.
Bouyer's Very Similar System runs perfectly well.

-allen