ATA/66 - DMA - whatever

[Gabe Turner (officer) <dopp@acm.cs.umn.edu>]

Ok, I'm not ashamed to admit that I know very little about some of the new
fangled IDE tech out there.  All I know is that it sucks that I get these 
errors

hda: status timeout: status=0xd0 { Busy }
hda: no DRQ after issuing WRITE
ide0: reset: success

all the time!  Also, my machine slows to a bloody crawl when I'm writing from
CD to HDD or reading from CD or basically when ever I use my IDE bus.  Now,
I know IDE is fairly CPU/memory intensive.  Both of these devices are on the
same IDE controller (I'd seperate them, but my place of employment doesn't have
any spare IDE cables at the moment).  One thing I've noticed, is that if I'm
copying from my 50x IDE CDROM to my 8x SCSI CDR, it starts out nice and speedy,
maybe even for a couple of sequencial burns, but then it slows to a crawl and 
the burn times out and dies.  At this point, top reveals that my system is
swapping out to disk a lot.

Anyway, while talking to a co-worker about this, he mentioned I give hdparm a
shot. 

Info:  I'm running kernel verison 2.4.0-test1 and have all of the fancy IDE and 
stuff compiled in.  I've got an Athlon 600, ASUS mobo with the AMD Viper IDE
controller.  At least I think it's the Viper.  I get this in my logs:

ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
VP_IDE: IDE controller on PCI bus 00 dev 21
VP_IDE: chipset revision 6
VP_IDE: not 100% native mode: will probe irqs later
    ide0: BM-DMA at 0xffa0-0xffa7, BIOS settings: hda:DMA, hdb:DMA
    ide1: BM-DMA at 0xffa8-0xffaf, BIOS settings: hdc:pio, hdd:pio

In my BIOS, I've got S.M.A.R.T enabled on my drive.
The drive in question is a 13GB Seagate Medalist Pro, 7200 RPM, ATA/66.

I've checked out www.linux-ide.org and it seems that my chipset may be 
supported in Linux.  This is on the list

AMD7409 VIPER

Now, my question is this: Has anyone on the list delved into the realm of hdparm
and have any relavent info?  I just want better performance, and I thought that
the new kernel options might do it, but apparently not.

Any advice would be greatly appreciated.

TIA!

Gabe
-- 
--------------------------------------------------------------------------------
Gabe Turner		X-President, ACM @ U of MN	     dopp@acm.cs.umn.edu

"Wanna buy a rubber walrus protector?!"
					- Mr. Horse in "Rubber Nipple Salesmen"
--------------------------------------------------------------------------------