Subject: Re: Faulty STM paging
To: Kim G. S. OEyhus <kim@iq.pvv.ntnu.no>
From: Neil A. Carson <neil@causality.com>
List: port-arm32
Date: 04/29/1998 17:14:54
Greetings,

Just some more replies to stuff...

Clustered page-ins may help some of the problem (UVM will page three
pages at a time anyway). But you could still branch to the end which
will throw it some of the time---a compiler building a branch to the end
of a function is I think quite common.

> A big problem here, is that those affected, can't fix the problem,
> since they cant compile the system.

I think Mark's view is that he's not going to cripple the port just
because Acorn shipped bad cards, but of course it'd be nice if he said
that here ;-)

	Neil