I wrote:
> This patch circumvent the problem by a very ugly hack: it
> simply sets the PG_reserved bit of all those pages used for
> guest memory, thereby declaring them as 'hardware range' ;-)
> This works fine on my system (2.0.36), but I'm not at all sure
> that it will work on other versions ...
Hmmm. I just looked at the 'bttv' driver code in the latest 2.3.15
(bttv is always used by Alan Cox as an example how you *should* do
buffers accessed both from kernel and user mode, IIRC), and guess what:
it sets the PG_reserved bit ;-)
So I seems the hack wasn't all that ugly after all ...
B.t.w. Kevin, are you going to make another release including my
recent patches, or are we waiting for CVS going online? (Or is
there a problem with those patches?)
Bye,
Ulrich
