On Fri, 16 Jul 1999, Thierry Vignaud wrote:
> Mauro Tortonesi wrote:
> >
> > On Mon, 12 Jul 1999, Marco Fioretti wrote:
> >
> > > hda: timeout waiting for DMA
> > > hda: irq timeout: status=0x59 { DriveReady SeekComplete DataRequest
> > > Error }
> > > hda: irq timeout: error=0x84 { DriveStatusError BadCRC }
> > > hda: DMA disabled
> > > ide0: reset: success
> >
> > What are these lines? By the time I've added "append "ide0=autotune"" to
> > my lilo.conf I keep getting this kind of errors while using my WD AC4300
> > UDMA-capable. The other disk, an old Conner 1,2 GB only PIO3-capable, does
> > not report any error on I/O.
> > My chipset is Intel 430FX (ok, a P120 with an Intel Zappa MB today is a
> > dinosaur, but is still works fine).
> If it occurs one time, it is safe: it's hust means that the linux kernel
> try to enable dma transfert with the disk but it fails so it roll back
> to old eide modes.
> If it keeps to occur, disable the ide0=autotune or just put an "hdparm
> -d0 /dev/whatever_is_your_drive" somewhere in /etc/rc.d/rc.sysinit
> (where there is already an hdparm line for example).
> ANother solution : i've seen some udma disks display this kind of errors
> when they were on the same ide cable, but not if on different cable. You
> may put your udma one on the first cable (ide0) and the other on the
> second cable (ide1). Whatsmore, the bandwith may be more important.
>
> --
> MandrakeSoft http://www.mandrakesoft.com/
> --Thierry
>
My disc/Mb combination requires -u1 for -d1, if that helps any.
