On 3/29/2017 7:57 AM, Binyamin Dissen wrote:
On Wed, 29 Mar 2017 07:19:05 -0700 Ed Jaffe <[email protected]>
wrote:
:>The majority of z/OS system services are SVC and PC-entered. The linkage
:>to such services does not care about the AMODE/RMODE of the caller. I
:>would expect most of them to just work right out of the gate with
:>failures being the exception rather than the rule.
Which really p*ssed me off when so many services were documented as supporting
AMODE64 callers when the smaller print indicated that all addresses must be
below the bar.
What they did was helpful, not hurtful. Those services are nearly all
31-bit code. What should they have done differently? They could have:
a) kept them "unsupported" in the books with "use at your own risk"
caveats, allowing Level 2 to throw up their hands and not take bugs
seriously for AMODE(64) callers, or ...
b) provide formal, official support for those services and document that
the parameters must be below the bar.
They chose (b) which is much better from a service standpoint for ISVs
and customers alike.
It sounds like you're angry that they didn't (c) mass convert all z/OS
services from AMODE(31) to AMODE(64) code. That would be a LOT of work
and testing! Perhaps we'll get there one day, little by little ...
--
Edward E Jaffe
Phoenix Software International, Inc
831 Parkview Drive North
El Segundo, CA 90245
http://www.phoenixsoftware.com/
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
