At 3:02 AM 8/17/95, Markus Wild wrote:
>> I think that at this point we can fairly safely conclude that it's
>> the Retina Z3 (I have the same problem, and have tried to get others
>> to respond, but the only one other person who posted about this with
>> their configuration has a Z3 also).
>> 
>> So far, it doesn't look like anybody's showing much of an interest in
>> fixing it.
>
>If it's the fault I already reported quite some time ago: someone in his
>unbelievable wisdom decided to make something a default that he/she obviously
>didn't even bother to test. Check out ite_rh.c, there are sections 
>ifdef'd out with something like RETINA_SPEED_HACK or so. The code which
>gets executed if this define is *not* defined (thus, the default!) causes
>these vm_faults. Define this symbol -> no more vm_faults. I was slightly
>pissed when I found out about this...

>From the log for ite_rh.c

date: 1995/04/06 19:19:45;  author: chopps;  state: Exp;  lines: +20 -8
made current retina vertical scroll code conditional on the new
option RETINA_SPEED_HACK.  Old commented code has been resurected.
This reportedly fixes the smear problems some retinas have.
John Vrolijk <dsnjvro@etmsun.etm.ericsson.se> noticed this.

I believe this "smearing" was being seen on 040 systems.  How can I test
a board I don't have?

Chris.