On Mon, 2005-02-14 at 10:16 -0800, Andrew Morton wrote:
> Fruhwirth Clemens <clemens@endorphin.org> wrote:
> >
> > First, one has to make kmap fallible.
> 
> I think it would be relatively simple and sane to modify the existing
> kmap() implementations to add a new try_kmap() which is atomic and returns
> failure if it would have needed to sleep.

Is anyone going to implement that? I would be willing to rework my
scatterwalk code one more time, but I'm not going to touch the kernel
vm.

-- 
Fruhwirth Clemens <clemens@endorphin.org>  http://clemens.endorphin.org