> On Sat, Dec 05, 1998 at 11:29:47PM -0500, Rafael Reilova wrote:
> > You don't happen to have a VIA chipset on your motherboard do you? There
> > is a hardware bug with the VIA that causes a total freeze up ocassionally
> > when doing DMA (esp. with sound). the AC kernels, since 2.1.129 (IIRC)
> > have a work around for that bug (you must enable PCI quirks). I was going
> > crazy with it too.
>
> Do you know if this affects disk activity too?
>
> I have a VIA and get lockups with DMA to my IDE disk, unless
> I do hdparm -X34 as well.
I don't think so, but I can't tell for sure. The problem happened
in the sound code when the driver attempted to figure out how much the DMA
transfer had progressed, by accessing the counter in the ISA DMA
controller. Under heavy CPU load, and for some unknown reason, the VIA
would just lock up. My guess is that a DMA was initiated, even after
being disabled, but since no DMA was supposed to happen it never
completed, and the CPU was kept in the HOLD state.
OTOH, IDE doesn't play with the DMA controller like that. The driver just
runs when it gets its interrupt. Maybe Alan can aswer this one, Alan?
Cheers,
Rafael
-
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/