>> Is my understanding correct. No, not really. I however imagine some such scenario: your TCB code starts (directly or indirectly) an SRB. It then WAITs for completion of this processing. Something like VTAM fast path for example. The SRB runs and POSTs the waiting TCB which is then dispatchable. The SRB has control of a CPU and wants to pass that to the TCB code, which it does using TCTL. This causes the SRB to terminate and the TCB code (after WAIT) is dispatched directly, without further waiting for the dispatcher to get around to it (after the SRB has otherwise terminated). Of course, the TCB might possibly already have been dispatched, in which case the SRB just terminates. I guess that's the kind of situation the function is provided for. Or there's RESUME RETURN=N for the TCB RB which will make the RB dispatchable first (in fact the RESUME RETURN=N function jumps to the TCTL exit in the MVS code, it's pretty much the same function). That's if one's up to playing with things like that.
-----Ursprüngliche Nachricht----- Von: IBM Mainframe Assembler List [mailto:[email protected]] Im Auftrag von [email protected] Gesendet: Sonntag, 24. Februar 2013 15:57 An: [email protected] Betreff: Re: AW: Curosity Question About TCTL Let me see If I understand this. Lets Say I Attach a Task (TCB) - the program (RB) associated with that TCB runs and ends. The TCB is not detached. At some undertimed time latter an SRB is scheduled into this address space. The SRB Issues TCTL with/to the TCB. So Do I understand that the program (RB) will be re-executed Again ? Is my understanding correct. Paul D'Angelo ---------- Original Message ---------- From: David Stokes <[email protected]> To: [email protected] Subject: AW: Curosity Question About TCTL Date: Sat, 23 Feb 2013 17:48:23 +0000 The TCTL or RESUME kicks off the TCB code without it having to wait for the usual scheduling, as long as it is dispatchable. What the TCB code does then is up to it. >>When the resumed routine finishes executing under the TCB should it re-enter >>the WAIT state again ? No, but it presumably will at some point or terminate. -----Ursprüngliche Nachricht----- Von: IBM Mainframe Assembler List [mailto:[email protected]] Im Auftrag von [email protected] Gesendet: Freitag, 22. Februar 2013 17:31 An: [email protected] Betreff: Curosity Question About TCTL I was reading about Scheduling SRBs using IEAMSCHD and discovered the TCTL macro. The TCTL (transfer control) macro allows an SRB routine to exit from its processing and to pass control to a task with minimal system overhead. When an SRB specifies RESUME RETURN=N, control transfers to the resumed TCB. Control then passes to the top RB on the TCB/RB chain, but only if the system determines that the RB is dispatchable. Again this is a curosity question and I have nop real reason to use TCTL. Am I to understand that the TCB needs to be in a WAIT state ? When the resumed routine finishes executing under the TCB should it re-enter the WAIT state again ? Its not clear to me how this facility should work. Can Some one explain the mechanics behind using TCTL and the target Task Control Block. Paul D'Angelo ---------------
