DRL ABTJOB (octal 51) - abort batch job.

Usage:

Normal program:
    drl     abtjob
    bci     1,snumb
 
JABT subsystem:
    drl     abtjob
    bci     1,snumb
    zero    action,0

Where:

snumb

is the 5 character id of the job being aborted. The lower 6 bits are ignored. On return, from the DRL the snumb is overwritten with a status.

action

If the JABT subsystem is executing the DRL sequence is extended with a subfunction code. The following action codes are supported:

0 normal abort

This behaves the same as the normal DRL ABTJOB. The job will be aborted if it is in execution, and originated from the current terminal. The following status returns can occur:

0: job not executing.
1: job not submitted from current terminal (not your job)
2: abort initiated.

1 find J*

This is used to find the J* for an executing job, so that a userid check can be done. The following status returns can occur:

0: job not executing.
1: job is executing in activity 0. No valid J* is available.
sct,llink: the J* location from SD.SN2.

2 abort executing job

An abort request is sent to POP1 without any ownership checks. This always returns a status 2.

3 abort job in scheduler

This sends an abort request to $SCHD. The following status returns can occur:

1: scheduler queue is full.
2: request sent to scheduler.

Description:

This derail is used to abort a batch job. The job is identified by snumb. The DRL has two formats. The first is used by normal programs, and only aborts currently executing jobs submitted from the requesting TSS terminal.

The second format is used if the current subsystem is JABT. This can abort jobs without a terminal id check, and it can also abort jobs still in the system scheduler queues. It is assumed that JABT subsystem will have read the J* to check job ownership.