On Tue, 21 Apr 2015 18:01:10 -0500, Walt Farrell wrote: > >Yes, but only for abnormal termination of the Initiator TCB, or for failure of >the entire address space. In normal or abnormal termination of the jobstep >task those resource managers won't help the OP, and the ENQ would remain >outstanding because it's not owned by the task that's terminating. > Hmmm. Can a program fork() a child that can outlive the job step?
Or the child could be another job submitted via INTRDR. Such a child could issue the ENQ and read() from a FIFO. The final job step of the parent job could write a token to the FIFO; the child could accept that and terminate. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
