Re: IDE-DMA strangeness (another one)

Rik van Riel (
Wed, 2 Dec 1998 10:42:58 +0100 (CET)

On Tue, 1 Dec 1998, Andre M. Hedrick wrote:
> On 1 Dec 1998, Mark Gray wrote:
> > Rik van Riel <> writes:
> >
> > > I've just moved my IDE disks around, taking out an old
> > > disk and moving my new 6.4GB disk from hdd to hda.
> This sounds your chipset........did you happen to
> run 'W-word'? Older PIIX's lie about capabilty because of crappy BIOS
> writers/hackers............

Yes, Wide transfers were enabled.

> > > With this move I saw in improvement (hdparm -t) from 3
> > > to 8.4 (!) MB/second... Now I wonder what kind of bug
> Your PIO3/4 disk that had a recomended of 180ns for DMA was locked out.
> The previous location was a non-correctly allocated PDT.
> Skip the assigned PIO settings in the bios and use AUTO/AUTO

I normally use hdparm to set things -- I really was under
the impression that Linux took care of things (I never got
any messages about failed DMA mode or stuff like that).

> > > in the Linux I/O subsystem could hold performance back
> > > _that_ badly.
> Not a bug...........8.4 MB/second is a UDMA drive at DMA mode 2
> on an 430HX/VX/TX, and I bet the usable buffer is more.

So that is the maximum speed I can expect from my IDE subsystem?
Seems fair enough to me :)

> > Some motherboards use a crippled controller for ide1 in the belief
> > that it will only be used for cdroms.
> That is usually because of lazy BIOS writers/hackers (ass)(u)(me)
> "Nobody will use more than two hardrives"

:) I had 5, 3 IDE and 2 SCSI. And an ATAPI CDROM and a SCSI
optical. Now one of the IDE drives is removed (slow, 250MB)
and the others are repartitioned. With 4 disks your system
can run pretty smooth....


Rik -- now completely used to dvorak kbd layout...
| Linux memory management tour guide. |
| Scouting Vries cubscout leader. |

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at