On Tue, 17 Dec 2002, Alexander Yurchenko wrote:

> > find it to be necessary: the bit in IDEDMA_CTL seems to be reliable.
>
> Yes, I agree, but hardware always has quirks, bugs, etc...
>
> > At last for the promise controllers I have.
> > Did you find controllers that required this code ?
>
> No, but may be Teemu Rinta-aho has the one ;-)

Thanks for the patch Alexander, but I still haven't
had time to test it, as the bug is on my home PC,
and lately I've not been much at home.. I hope
I can try it tonight. I will report then.

I can tell you, I have hard time reproducing this
bug, as now it seems it's not about taking the
power off for a long time either. I have found no
logic on that.. If I knew how to reproduce it, I would
of course be looking ways to debug it somehow.

BR,
Teemu