On 04-Jan-2012 15:28 , CRPence wrote:
<<SNIP>> If for example the job had a break message handler event
active which responded with ENDRQS upon receipt of a message, then
the DLYJOB would be interrupted; but so too would the active program
[as controlled by the defined\registered invocation exit]. If the
DLYJOB were implemented in its own stack entry, the processing would
probably be easiest; the call to the program doing the DLYJOB could
signal an exception that is monitored by the caller. <<SNIP>>
Hmmm... After a re-read, the above text seems somewhat deceptive for
the capability generally available using End Request. A new call stack
entry is probably moot, because all of the stack entries would terminate
down to the request processor [that is monitoring for the CPF1907].
Reviewing the code in which I had incorrectly recalled that I had
actually implemented an EndRqs via a break message handling program, I
verified that instead, the code was implemented to send an
exception\escape message from the Break Handling exit program to the CLP
doing the DLYJOB. That messaging was the means by which the receiving
program was informed, and thus decided how to react, via global and\or
command-specific MONMSG; i.e. the CLP had not been shut down by End
Request processing as alluded might be done in the quoted text. The
program QWCDLYJB as the CPP for DLYJOB would terminate early as the
effect of the exception being signaled to the CLP earlier in the stack.
So... the code actually terminates the DLYJOB upon request, via a
message to a queue setup in break mode to a break msg program,
irrespective the time remaining for the wait-time interval.
<code>
monmsg cpf9898 exec(*N) /* Ignored [except on DLYJOB] */
mainline: /* Application program: AppPgmName */
chgmsgq AppMsgq dlvry(*break) pgm(AppBrkMsg) sev(0)
chgjob brkmsg(*normal) /* "*hold" prevents brkmsgpgm too */
...
dlyjob 600
monmsg cpf9898 exec(do) /* specific reaction to msg or: */
... /* No-Op or (global no-op + no MONMSG), ends DLYJOB */
enddo
...
</code>
<code>
pgm (&mqnm &mqlb &msgk) /* Break Handling pgm: AppBrkMsg */
dcl &mqnm *char 10
dcl &mqlb *char 10
dcl &msgk *char 4
rcvmsg msgq(nada) msgkey(&msgk) rmv(*yes) /* msgdta(&msgdta) */
sndpgmmsg msgid(cpf9898) msgf(qcpfmsg) topgmq(*same (AppPgmName)) +
tomsgq(*topgmq) msgtype(*escape) msgdta('Interruption' /* or
&msgdta */)
</code>
In the above code snippets... A message sent to the external message
queue [*MSGQ] object named AppMsgq would cause the AppPgmName [highest
in the stack] to receive an escape message CPF9898 from the AppBrkMsg
break handling program. The message text [e.g. first level text for
impromptu messages] or the message data [¿best defined as Message data
fields formats per FMT() on an ADDMSGD?] could be retrieved by the Break
Handler program and then passed to the application program, if something
specific needed to be communicated.
Regards, Chuck
midrange.com
