On Nov 30, 2013 3:10 PM, "Peter Relson" <[email protected]> wrote: > > I believe that there is no deferral. Unless RTM is terminating a higher > task, lower tasks can terminate in parallel and if it's time for an ETXR > to run, the IRB will be scheduled. If that happens to interrupt a previous > ETXR, so be it. > > If you want to see if one interrupts another, you could also have the > first wait on an ECB that never gets posted and see if any of the others > run. > (Assuming you don't mind a hung mother task for a test) > > I have not tried this. > > Peter Relson > z/OS Core Technology Design >
Thanks to all. I will remember to write my exits as thread safe code, reentrant (not real sure how they differ), and non-self modifying with appropriate use of serialization for storage which is not thread local (STORAGE OBTAIN or equivalent). May be over kill (hi NSA!), but is likely to be less bother. I prefer to code consistently, which is why my HLASM code is now LE enabled whenever possible. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
