A posted ECB of xx400000 can be either system completion code of 400; or reg 15 value of 400000 on normal return to the system.
You can tell "which" by bit TcbEndingAbnormally (introduced in z/OS 1.12 for exactly this purpose) -- on in the first case, off in the second. I believe that the value in the low 3 bytes of the posted ECB will also match the value in STCBCMPC (not necessarily the value in TCBCMPC) Note that if you did ATTACHX with ECB, the TCB will still be around for you to look at, after the ECB is posted. I'm still curious about Steve's assertions about use of "CALLDISP" and about DETACH abending. Neither of those are likely to be correct interpretations of what's going on. I'm still curious about just when Steve "saw" the ECB contents with an RB that was waiting. Was it by display from code that executes after the POST? By a dump? The fact is what I wrote: the ECB provided on ATTACHX will be posted upon termination of the task in all cases (on a memterm, one would not consider the task to have terminated). If the code does not "wake up" then in some way the ECB provided on ATTACHX was not the ECB being waited on. But then if the code does not "wake up" no further instructions would have been executed. If the code does wake up and the ECB is not "posted" when looked at then either the ECB has been waited on by something else after the post or the display is not of the right ECB. Other posts have suggested that it would have been helpful to show the exact values that resulted (return code, ECB contents, ECB address, etc), by some print mechanism. I agree. Peter Relson z/OS Core Technology Design ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
