On 19 Jan 2001, Nathan J. Williams wrote:

> Anders Dinsen <anders@dinsen.net> writes:
> Well, for other classes of devices (most notably SCSI devices), there
> are "quirk tables" that have bits describing the various and sundry
> failures of a device. I don't see the overhead in having one of those
> here to be worthwhile, though, until and unless we discover more
> brokenness like this.

I agree, but in this case, all the information I have been able to dig up
indicates that the problems only occur in certain situations (with certain
cards). So disabling memory access completely seems to me to be a bit on
the safe side.

But, I may not have digged for information in the right places. I hope
Jason Thorpe (or someone else) remembers it or can dig up something.

- Anders


-- 
Anders Dinsen
anders@dinsen.net
http://dinsen.net/anders/