Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!att!mcdchg!behemoth!mph
From: mph@behemoth.phx.mcd.mot.com (Mark Huth)
Newsgroups: comp.sys.amiga
Subject: Follow up on 2620 woes
Message-ID: <10981@behemoth.phx.mcd.mot.com>
Date: 1 Jun 89 00:27:52 GMT
Reply-To: mph@behemoth.UUCP (Mark Huth)
Organization: Motorola Microcomputer Division, Tempe, Az.
Lines: 17

As i reported yesterday, I'm having troubles with my 2620 crashing.  Last
night I tried setfastrom. With the fastrom installed (it copies rom to
32bit ram and write protects it, using the MMU to map it into the same
address as the roms) I was able to complete all of the benchmarks.  I ran
overnight looping through the compilations that formed my benchmark, and
found no problems.

As soon as I removed the fastrom, every benchmark crashed.  This would point
to a problem with access to the ROM chips.  Given that most of the failures
seem to be related to instruction corruption (lockup, address trap, illegal
instruction traps) I wonder if there isn't a timing problem going to offboard
memory, and in particular, the roms.

Any thoughts?

Thanks,
Mark Huth