On Mon, Oct 20, 2003 at 12:14:28PM +0930, Brett Lymn wrote:
> As far as I know, hypersparc SMP is not working properly.  There is a
> work around available which effectively disables the cache.  There is
> a performance penalty for doing this though but it does make the
> machine run reliably.

Is said work around in-tree? I'm wondering if my success is
aberrant.

Steve