>From the 1.13 Utilities RM:
v IEBCOPY uses the EXCP access method and special I/O appendages.
Therefore:
Some common DCB parameters such as BUFNO are ignored.
v Starting in z/OS V1R13, IEBCOPY is no longer APF-authorized.
A pre-V1R13, APF-authorized copy of IEBCOPY is available as
IEBCOPYO in SYS1.LINKLIB.
Hmmm. I had believed that I/O appendages required APF-authorization.
Does this imply that Starting in z/OS V1R13, IEBCOPY no longer uses
special I/O appendages.
I had understood (from gossip; I haven't RTFM) that:
o If a program executing not APF-authorized LINKs or ATTACHes
any other program, the called program executes not authorized,
regardless of its AC attribute or the status of the containing
library.
o If a program executing APF-authorized LINKs or ATTACHes any
other program from an authorized library, the called program
runs authorized, regardless of its AC attribute.
o If a program executing APF-authorized LINKs or ATTACHes any
other program from a non-authorized library, the calling
program ABENDs, regardless of the called program's AC
attribute.
So what customer requirement would not be satisfied by
simply linking the new IEBCOPY into an authorized library
with AC=0?
o It could be invoked with EXEC PGM=IEBCOPY and simply
work, as it used to, but in the unauthorized state.
o It could be invoked from an authorized program with LINK
or ATTACH and it would simply work, as it used to, still in
the authorized state.
o It could be invoked from a non-authorized program with
LINK or ATTACH and it would simply work, as it usually
didn't previously.
o The renamed IEBCOPYO wouldn't be accessible from
legacy programs that do "ATTACH EP=IEBCOPY" unless
the customer either edited and recompiled those programs
or swapped the member names with a USERMOD.
Oops! Reading further, I see:
v IEBCOPY must not be loaded in supervisor state or in
protection key zero. It is an application program that does not
use the special system interfaces assumed for the system kernel
running in supervisor state or in protection key zero.
(That paragraph has no revision bars.) But, still, if it "does not use
...", it shouldn't matter. It's a considerable logical leap to "must
not be loaded ..."
Still curious,
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
