> This is odd because I have programs that call subroutines at the very end
of
> a 2K PIC from the ISR routine at ORG 4 with no trouble.  I don't save
PCLATH
> and never adjust it in the program.  The only trouble I run into is when I
> do a computed GOTO jump, in which case I set the PCLATH manually.
>
> If your statement is true, I'd never be able to return to nearly 80% of my
> calls, since they run the gamut of program memory locations from 0x00A
> through 0x700.  Yet I have no problems.
>
> --Andrew

Ah... there's pages and then there's PAGES:

Sometimes it matters what 256 byte page you're talking about.

Sometimes it matters what 2048 byte PAGE you're talking about.


For ADDWF PCL,F or MOVWF PCL or other things that directly hack PCL, PCLATH
better contain the high half of the correct 256 byte page.

For CALL or GOTO, the low order 3 bits of PCLATH are ignored, and thus
PCLATH only has to refer to any 256 byte page with the 2048 byte PAGE.

[Note: this is how it is on some PICs, but not all!]

Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.