A more complete log about my disk access errors

[Shaun Oliver]

you've gotta remember though kerry, that ultra dma isn't enabled by
default on dma66 or dma33 hard drives.
No wait a minute, it is but the running speed isn't set properly. it's
usually set to dma33 and you have to use a utility to enable dma66 for
some stupid reason.
You do under windows any way. not sure about linux.


Shaun..
"We realise we have a problem with communication. However, we're not going
to discuss it with our staff."
EMAIL: shauno at goanna.net.au ICQ: 76958435
YAHOO ID: blindman01_2000 IRC NICK/SERVER: |3|1ndm4n on #aussiefriends on
www.jong.com:6667

On Sun, 3 Jun 2001, Kerry Hoath wrote:

> You may also find that certain drives want to go into udma66
> which makes sense since that is their fastest powerup dma mode. Now if you don't happen
> to have a UDMA66 cable, then you can experience dma timeouts or crc errors,
> because Linux thinks the drive is in superfast mode, the drive does too but the signals on the cable
> are getting corrupted due to the lack of conductors in the cable.
> UDMA 66 has 80 conducters and normal ide has 40, although both only have 40 pins.
> 
> I found that with my Quantum 30gig drive, I couldn't enable dma reliably
> unless I had an UDMA66 cable on the Via 82c586 controller. Some new
> drives suck in pio mode providing a lack lustre 5 megabytes per second
> but run at 15 megabytes when udma is enabled.
> If anyone wants any more information on Linux and DMA/UDMA let me know under Linux,
> sadly under Windows you'll have to ask Kirk Wood about the dark side :-)
> 
> Regards, Kerry.
> On Sat, Jun 02, 2001 at 04:53:14PM -0500, Kirk Wood wrote:
> > First, understand why Windows 2000 doesn't complain about the DMA. If the
> > drive times out twice (ever) then it no longer uses DMA ever. So it won't
> > complain. It won't tell you. In fact, win winclows 98 they burried any
> > indication that this had happened. It happens more often them many people
> > realize.
> > 
> > Having said that, if you didn't have DMA support compiled into the kernel
> > you wouldn't get the errors. Yes, it must be compiled in if it is to be
> > used. (DMA falls back to PIO in case of failure and PIO always remains
> > available.)
> > 
> > As for fixing this, it is either the motherboard or the drive. I know that
> > doesn't help. You might check to see that DMA is turned on in the
> > BIOS. (Linux will try anyway, winblows just acts like it is using DMA and
> > doesn't really.) Otherwise try another DMA drive and see if the problem
> > resolves.
> > 
> > =======
> > Kirk Wood
> > Cpt.Kirk at 1tree.net
> > 
> > Nothing is hard if you know the answer or are used to doing it.
> > 
> > 
> > _______________________________________________
> > Speakup mailing list
> > Speakup at braille.uwo.ca
> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> 
> -- 
> --
> Kerry Hoath: kerry at gotss.net
> alternatives: kerry at gotss.eu.org or kerry at gotss.spice.net.au
> ICQ UIN: 8226547
> 
> _______________________________________________
> Speakup mailing list
> Speakup at braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup
>