On 2014-03-25 13:29, John McKown wrote:
Yes, I am reading the book on Metal C. But I often appreciate other
peoples' opinions and ideas. So I'm "just asking".
I am porting some C code to z/OS (nedit thread on MVS-OE). I need to be
able to do z/OS enqueues and dequeues. I don't see any C subroutine to do
this. So I need to write something. This is not a big deal to me. Normally,
I'd just whip something up in HLASM and go on with it. But I was thinking
that it might be better to try to use Metal C simply because everything
else is written in "normal" C. Yes, I know how to write LE assembler code.
The C code is xplink, so I think in either case, I will need to use
"#pragma linkage" to set the routine up as having OS linkage.
Thanks for your opinions.
Metal C follows MVS linkage conventions. The Metal C manual refers you
to the MVS Assembler Services Guide for details on those conventions,
where no mention of XPLINK is made. XPLINK is an LE convention.
If you are already using LE and XPLINK for the nedit C code, then I'm
not sure I see the benefit to writing a Metal C subroutine to do the
ENQ/DEQ. The C code in this case will likely be mostly comprised of
inline Assembler with the cryptic syntax used to map C constructs to
Assembler constructs. Your code will probably be more understandable if
you write a LE-compliant Assembler subroutine to do those functions.
--
Regards, Gord Tomlin
Action Software International
(a division of Mazda Computer Corporation)
Tel: (905) 470-7113, Fax: (905) 470-6507
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
