[net.works] 68000 page faults

Leonard@sri-unix (07/08/82)

I believe the consensus \at that time/ was that one needed a second
processor to handle page faults, hanging the primary one while the
page was brought in.

However, I remember another message (no longer on this system but
somewhere on paper, *sigh*) which talked about the availability of
various newer 68000 parts.  One was a 68000 memory mangler, and one
was a version of the 68000 which preserved enough information
internally to be able to handle a true page fault (i.e., one in which
context switching can take place while the page fault is being
handled).

I can find the information on these parts sometime today.  As I
recall, the part numbers were of the form 68010, 68020, etc, and were
going to be available from six months ago to six months from
now---therefore what you're looking for may well be available.

If anyone wants to look throuh the archives, s/he may hit paydirt
looking for the strings "68010" or "FQ 82" or "FQ '82" or something
similar.  The message was kinda short.

Ciao.

						<LNF>

phw (07/08/82)

I have in my grubby little hands a product information sheet
published by Motorola Semiconductors on the MC68010.  This processor has
the capability to run in both virtual memory mode and virtual machine mode.
It is upward compatible with the MC68000 and MC68008 processors.
I have no idea of the development schedule of this processor, but at the time
of the writing of the sheet (4/82) it was still under development.