Messages in this thread | | | Date | Sat, 5 Sep 1998 18:25:12 +0100 | From | Jamie Lokier <> | Subject | Re: Odd 2.0.35 problems with APM on VIA motherboards |
| |
On Fri, Sep 04, 1998 at 06:38:41PM +0200, Hans de Goede wrote: > When you say "wether apm is enabled or not" do you mean in your bios/ or > in your kernel. And is only apm disabled or also hd-sleeping.
I mean in the BIOS. I've tried these combinations over time:
BIOS Kernel off off off on on on
I've never encountered lockups with hd-sleeping. I've only used BIOS-controlled hd-sleeping though.
> I happen to have a quantum fb st 6.4A drive too. And both I and a > friend of mine who also has this hd (and doesn't have a via mb!) are > experiencing hardlockups when udma is enalbed and the hd spinsup.
Well, I've never had a lockup while the disk is spinning up. As far as I am aware, Linux has no part in the spin-up process except that it is suspended while it happens. As said, I only used BIOS/APM-controlled spin up.
> So disabling/enabling apm doesn't help. The hdspinning after set by the > bios is > controlled by the hd itself and can only be set from hdparm afaik.
I think you are talking about the hdparm options to spin down the disk after a certain amount of inactivity, independent of APM. I've never tried these options.
> So it seems that the quantum st6.4A has a bug which causes it to hang > with udma + hd-sleeping. > I would like to have this investigated further.
I'm now using DMA with BIOS/APM-controlled hd sleeping, without any lockups. The wonders of modern technology are such that I simply don't know if it is using UDMA or non-U DMA.
The bootup message from the IDE driver saying UDMA doesn't help any. It says that even when I don't use any DMA at all.
Would someone care to tell me how I can tell the difference?
> It's also very interesting that you say it doesn't hang with -X34. > what would that do exactly ?
I don't know. It does change the output of hdparm -I, suggesting to me that maybe I am using non-U DMA now. But the hdparm -t timings are the same as ever. (9MB/s).
It is poorly documented. But it works for me!
-- Jamie
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html
| |