On Sat, 6 Dec 1997 10:29-EST 
 "Robert.V.Baron" <rvb@gluck.coda.cs.cmu.edu> wrote:

 > I'd like to suggest that (for 1.3) a check be made in the kernel and iff
 > the sizes are such that dma can not work, I'd either have the kernel not
 > build or panic.  We could have a option NODMA that if set would not 
 > provide these warnings.  This way people would not accidentally get 
 > confused.

Well, there's another problem, here... On systems with large amounts of
memory (256M, for example), this problem happens without building with
additional buffers...

Unfortunately, I'm leaving for Wash, DC for a week in the next hour or
so.  Perhaps I'll try to look at this while on the plane.

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                            Home: +1 408 866 1912
NAS: M/S 258-6                                       Work: +1 650 604 0935
Moffett Field, CA 94035                             Pager: +1 415 428 6939