Wouter van Ooijen <PICLIST@mitvma.mit.edu> wrote:

> As far as I can make out from the datasheet they realy have less
> FLASH, the remaining is ROM.

    That seems unlikely, especially since they have the same device
    ID as an all-flash PIC.

> > it's trivially easy, in most cases, to overcome the restrictions
> > on CALL destination addresses in the 12-bit parts.
>
> It is just the place of that one fixed bit I was complaining about.
> IMHO it should be either the LSB or the MSB, not somewhere inbetween.

    On the 16C54, that bit IS the MSB... And if you don't count the
    page-select bits as part of the address, it's the MSB on the
    other 12-bit PICs, too.

> And how do you overcome that limitation easily, without using call
> vesctors or re-arraging your code? Note that I am looking at it
> from the compiler writers perspective.

    Although the destination of every CALL must be in the lower half
    of a 512-word page (i.e., in the 0xnn00-0xnnFF area), there's no
    similar restriction on the destination of GOTOs... So you just
    put your subroutine entry-points at the low end of a page and
    have them GOTO the body of the subroutine which can be anywhere
    in memory.

    For an example "from a compiler writer's perspective", see the
    output of Bytecraft's C compiler; Walter Banks uses what I
    believe he calls "CALL islands" to solve this problem and the
    related problem of cross-page GOTOs.  Presumably, the other C
    compilers deal with the problem in a similar way.

    I'm not familiar with JAL; is there something inherent in its
    design that makes this approach unworkable?

    -Andy

=== Andrew Warren -- aiw@cypress.com
=== Principal Design Engineer
=== Cypress Semiconductor Corporation
===
=== Opinions expressed above do not
=== necessarily represent those of
=== Cypress Semiconductor Corporation

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads