> >Keep in mind that when executing an EI the interrupts will be enabled
after
> >the next instruction. So this is the way to keep the interrupts disabled
as
> >short as possible, which is always preferred (although not required).
>
> I just wondered... Does DI have this behaviour as well? If so, the example
> above can execute incorrectly.\
No. DI hasn't got the same behaviour.
> And remember that instructions like LDIR and OTIR are treated like "ultra
> short loops" in terms of interrupts, meaning an interrupt can occur
> "inside" an OTIR.
Oh, really??? Well, that's great, for then I don't have to avoid OTIR-alike
instructions in project "i" to keep the interrupts ready!!! Disn't know
about that.
And in fact, it is also quite logical, because if the instruction didn't end
yet it just executes itself again. So an interrupt can occur inbetween.
> >> Is it legal to call 256 times #FD9F with interrupts disabled at VDP
level?
> >
> >Absolutely. But it might re-enable the interrupt. You'll have to check
that.
>
> It might execute "EI". But unless there is some really weird routine
> hanging on #FD9F, it won't enable interrupts at VDP level.
That's about exactly what I said. :)
> >But why do you need to keep the interrupts disabled???
>
> Sony F700.
But if you replace the DiskROM then ALL diskaccess will work fast.
~Grauw
****
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
****
