On Tue, 17 Apr 2012 17:49:24 -0400, Tony Harminc wrote: > >>>Actually, they probrably would document that SYSLIB DD name cannot be >>>used to with z/OS 1.12 and above to allow for testing alternate >>>versions of the program. >>> >> Are you thinking, rather, perhaps, of STEPLIB? >> >> Its likely that they internally DYNALLOC DD(SYSLIB) REUSE, for whatever >> purpose. They ought to remove the REUSE key, allowing a "IKJ56246I" >> FILE in USE" message to appear and use a generated DDNAME (best). >> A poor second is to document the restriction. > >I doubt that the IEH utilities use DYNALLOC. They long predate the >very existence of DYNALLOC, and while they may, of course, have been >updated, I suspect they continue to use a combination of DADSM and >home-grown manipulation of control blocks. > Obviously, something changed between 1.11 and 1.12. >But of course there's no real way to tell in these OCO days, unless >some message or behaviour leaks the information. > In TSO:
o preallocate SYSLIB to an arbitrary DSN. o CALL *(IEHLIST) o use BPXWDYN( 'INFO ...' ) to detect any change in allocation of SYSLIB. (Or preallocate SYSLIB to a USS path. DYNALLOC REUSE doesn't deallocate PATHs. Analyze any failures.) -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
