Writing error routines

Use the ERROR statement to create an error routine. The error routine defines an action to be taken when a CLIST receives a non-zero return code from something other than a CLIST subprocedure. (Table 1 lists the CLIST error codes.) The action can be any executable statement and is often a DO-group that performs operations tailored to the indicated error. You can structure an ERROR action as follows:

ERROR + 
 DO
⋮
  (action)
⋮
 END

The ERROR statement and its action must precede the code to which it applies. An action may contain TSO/E commands and subcommands, subject to the mode in which the CLIST is executing when the error occurs. Unlike attention routines, error routine actions can issue multiple TSO/E commands or subcommands.

If an error routine action does not end the CLIST, it must include a RETURN statement. The RETURN statement returns control to the CLIST statement, TSO/E command, or TSO/E subcommand following the one that was executing when the error occurred. Repeated errors which activate the same error routine may cause the CLIST to terminate.

You may also code error routines that override previous ones. Each error routine overrides all previous ones. You may initialize new error routines as many times as you want.