frank@phoenix.owl.de (Frank Wille) writes:

>Even if it does, it might just have been luck, because you didn't run into a
>page-fault while CAS was executing (also had this myself once).

AFAIK kernel mutexes aren't paged out nor is the kernel map.

>Yes, 68030 definitely works.

The 68851 is mostly a superset of the 68030MMU. If the page tables
only use 68030 features then it should work the same.

The only gotcha would be the smaller 68030 ATC (22 vs 64 entries).
If we forgot to flush the ATC correctly this might have an effect.

-- 
-- 
                                Michael van Elst
Internet: mlelstv@serpens.de
                                "A potential Snark may lurk in every tree."