Binyamin, Chris, I did some testing and more reading and discovered the following...
a) The Extended Addressability Guide says that each workunit (TCB or SRB) has its own stack. From that I it appears that the IRB is sharing the stack with the orginating task. So, I won't be using the stack to save the state in the exit. Hence the second question on the size of the save area provided by the system for the exit. b) I checked the contents of 4(r13) and 0(r14) upon entry to the ETXR. 4 (r13) does not contain F6SA, so I assume it's a standard 72 byte save area; and r14 does point to an SVC 3. As Binyamin noted, the system requires the ETXR to restore the regs. Since the doc is not specific, I'll have to save and restore both GPRs and ARs - just to be on the safe side. As an FYI... I'm using the ETXR for cleanup and to inform the originating task (attaching task) that a child has terminated. There are a varying number of child subtasks, so I do cleanup in the ETXR and use a single ECB to post the originating task that a child terminated. There is a queue of control blocks (one per child) that the originating task can then check to determne which child terminated and take appropriate action if necessary. Thanks again, for your responses. Regards, Vic ---------------------------------------------------------------------- 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
