> On Thu, 23 Nov 1995, Andrew Warren wrote:
>
> > It was bad enough when only instructions that performed an EXPLICIT
> > read or read-modify-write kept the interrupt from happening... This
> > new information makes the change-on-portB interrupt almost
> > COMPLETELY worthless.  Does Microchip see this as a bug and plan to
> > correct it in future revs of the chip?

[bboles replied]
>
>      Actually, we understand this as a bug and it is being fixed in all new
>      devices including upcoming 'A' versions of '74, '73, '65, '64 , etc.
>
>      Rgds, Brian.

What precisely is Microchip fixing in the 'A' versions?  Most of the quirks
I can see with the PortB behavior (and read-modify-write in general) are
things that a "clever programmer" might exploit in some way.

Also, what is the specific behavior of the register read logic?  Does it
do the read for any opcode whose top six bits are not "00 0000", or does
it read regardless?  If the latter, does this cause any potential problems
with serial I/O on the '74 [since reading the input latch clears it, and
the output latch shares the input latch's address]?

PS--Personally, my preferred implementation for an "interrupt-on-change"
pin feature would be to have the interrupt triggered if/when the signal on
the port pin does not match the output latch.  This could then be easily
cleared either by "movf port,f" or by most read-modify-write operations on
the port.