On Wed, Aug 25, 1999 at 08:35:43AM -0400, Perry E. Metzger wrote:
> > Did it use PIO ?
> 
> No, DMA. One note: DMA probed but the machine initially refused to use
> DMA because of a lack of driver support. (Its an Intel 810 "everything
> but the kitchen sink" chipset). I set the flag on the pciide entry in
> the kernel config to make it try DMA anyway. I haven't touched the cd
> driver's flags, though.

It works fine under PIO ?
Well, maybe DMA was not properly set for the CD (lots of BIOS don't set up
DMA for ATAPI devices, it's one of the motivations I had to add mode settings
in the driver).
I don't know what the Intel 810 chip is, is it old or new ?
I guess I should try to find the docs :)


> 
> > If so it may be stuck in an interrupt loop (because the
> > drive and driver don't agree on how much data was transfered, there was more
> > than just a 'lost interrupt' problem on the IDE bus).
> > It should be fixed in -current, and I asked for a pullup yesterday.
> > 
> > Any chances to try this again once the pullup is done (or try a -current
> > kernel) ?
> 
> Sure, I could try anything you like. Just let me know what you want me
> to do.

Well, if the hang occured in DMA mode then this is another problem than the one
I fixed. But as you forced DMA maybe it's just that the BIOS didn't set up
DMA properly for the CD. Did the BIOS announce somthing about DMA modes used ?

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--