12 Mar
2003
12 Mar
'03
5:23 a.m.
> > I might be able to use the 8088 docs. The
V20 is *supposed* to be a drop-in
> > replacement, but IIRC it needed some (minor) BIOS alterations to get it to
> > work properly in PCs when the 8088 was swapped with a V20.
> Not that I know off. The V20 did behave exactly like an 8088.
> At some point I did use them on every PC and XT I could. The
> additional speed was quite cheap. The only time I couldn't
> use a V30 was in the Siemens PC-MX, a small 8086 Unix box.
Besides, it wasn't a completely Bug-For-Bug
replacement. NEC failed to
implement some of the quirks of the 8088. For example, if an interrupt
occurred during the execution of an instruction with a double prefix, the
NEC would continue, but the intel would drop one of the prefixes when
resuming.
REP MOVSB DS:[SI] DS:[DI]
will resume with a V20, but with an 8088 will only do one more rep after
an interrupt.
Well, that's eactly the 186 behaviour. Now considering that the
timeing was also exactly like 186 (or real mode 286), and the
instruction set is the same, I still think NEC just used the
186 (188) core as template for the V20/V30 - Considering that
V40/50 where like the 186 with integrated perhipherals, you
may considere it as a two step knock off...
Gruss
H.
--
VCF Europa 4.0 am 03./04. Mai 2003 in Muenchen
http://www.vcfe.org/