> I've seen claims on this list (see thread "Hi-Tech C
> Compiler - Demo Version Limitations", 4/8/03) and
> elsewhere that the (free) PICC Lite C compiler can be
> "coaxed" into supporting the 16F628 with an addition
> to the piclite.ini file.  Has anyone actually looked
> at the code generated as a result?
>
> I can illustrate a problem I found using this tiny
> example:
>
> #include <pic.h>
> void main()
> {
>  CMIE = 0;
>  CMCON = 0x07;
> }
>
> For the 16F627 it generates this (correct) code:
>
> 03FA  1683main BSF STATUS, 0x5
> 03FB  130C BCF PIR1, 0x6
> 03FC  3007 MOVLW 0x7
> 03FD  1283 BCF STATUS, 0x5
> 03FE  009F MOVWF CMCON
> 03FF  2804 GOTO exit
>
> For the 16F628 it generates this (INCORRECT) code:
>
> 03FB  1683main BSF STATUS, 0x5
> 03FC  130C BCF PIR1, 0x6
> 03FD  3007 MOVLW 0x7
> 03FE  009F MOVWF CMCON
> 03FF  2804 GOTO exit
>
> The problem is that the generated 16F628 code does
> not return to bank 0 from bank 1 before setting CMCON,
> thus setting VRCON instead.
>
> Any comments (besides "you get what you don't pay
> for :))?

        Yup, noticed that too, it seems some registers don't have their bank bits
set properly when you "hack" it into 628 mode, very annoying, took me quite
a while to figure out what was going one. Now that I know what to look for
it isn't as bad, but still... I guess they purposely "sabotage" things this
way. TTYL

----------------------------------
Herbert's PIC Stuff:
http://repatch.dyndns.org:8383/pic_stuff/

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