On Sun, Jul 30, 2000 at 10:07:57PM -0400, Allen Briggs wrote:

 > > If nothing else, it should be a DIAGNOSTIC panic.
 > 
 > So you are suggesting that the driver that is passing a 0 count to a
 > bus_space_read_multi (or whatever) check for that case and not do it,
 > and that the bus_space_read_multi (or whatever) functions have a
 > DIAGNOSTIC panic if called with a count of 0?

Yes, exactly.

-- 
        -- Jason R. Thorpe <thorpej@zembu.com>