On Mon, 20 Mar 2000, Arjan J. Molenaar wrote:
>
>
> > -----Original Message-----
> > From: Valentijn Sessink [mailto:[EMAIL PROTECTED]]
> > Sent: zondag 19 maart 2000 0:05
> > To: [EMAIL PROTECTED]
> > Subject: [LINUX-ABIT] found where it hangs (at my BP6)
> >
> >
> > Hi all,
> >
> > My BP6 can be forced to hang with a DMA timeout.
> >
> > The error is:
> > hde: timeout waiting for DMA
> > ide_dmaproc: chipset supported ide_dma_timeout func only: 12
> >
> > ... after that, there's only silence...
> >
> > Now I digged into the kernel to find out that a simple single one byte
> > input instruction just never returns. This is around line
> > 1460 at ide.c
> > in linux-2.2.15pre15+ide.2.2.15pre15, where it says
> >
> > startstop = ide_error(drive, "irq timeout", GET_STAT());
> >
> > ... the GET_STAT just hangs. GET_STAT resolves to inb_p but if you
> > define REALLY_FAST_IO (which I, last resorted, did in ide.h) you get
> > inb.
> >
> > Now I'm stuck. I don't know enough of SMP to know why a simple in
> > instruction will hang the machine. Could anyone help with this?
>
> WOW!
>
> It seemed to me that appending "ide2=ata66" also does the trick. This switch isn't
>documented, I found it in the changelog in ide.c.
I put the switch in for systems that fail to detect the 80 conductor
cable.
> Have you contacted the ide-maintainer about this? It's his job/hobby to resolve this
>kinda problems...
>
> Regards,
>
> Arjan
>
> PS. "noapic" gives a hanging kernel at my place, so I don't use it...
> --
> =- To unsubscribe, email [EMAIL PROTECTED] with the -=
> =- body of "unsubscribe linux-abit". -=
>
Andre Hedrick
The Linux ATA/IDE guy
--
=- To unsubscribe, email [EMAIL PROTECTED] with the -=
=- body of "unsubscribe linux-abit". -=
