Re: Bitwise ops in RPG

[boldt@xxxxxxxxxx]

Scott wrote:
>Just out of curiosity, how much work would it be to add
>these BIFs into the compiler?  From my point of view
>(keeping in mind that I've never written a compiler) they
>seem like one of the more basic things that a computer
>can do...

Well, first, there's the scan phase which has to parse the
new built-in functions.  This can be the most work since
we have to allow the BIF's in D-Specs as well as in
expressions.  Allowing these BIF's as operands to keyword
INZ requires evaluation of the BIF's at compile-time.

Then there's the relational diagnostics phase which checks
the types of the operands.

Then the code generation phase.  Code gen isn't just a
matter of emitting the ANDI, ORI, COM, and XORI W-Code
instructions.  It also has to handle unequal sized operands,
as well as operands that are longer than what the W-Codes
normally support (depending on how the BIFs are defined).

Add onto that writing and running unit and component
testcases, as well as design and code inspections for all
of this.

Fortunately, these BIF's should be relatively straight-
forward to design and implement.  Some enhancements,
though, require lots of thinking and analysis and heavy
debugging.  One good example in V5R1 was allowing the
argument to keyword DIM to be forward-defined.  In V4R4,
one change that required some heavy-duty design and
analysis was the enhancements to keyword OVERLAY.

Cheers!  Hans

Hans Boldt, ILE RPG Development, IBM Toronto Lab, boldt@ca.ibm.com

+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---