>>no asynchronous exit routine can interrupt another >>asynchronous exit routine
>Perhaps VTAM does something special, but in general this is not true. Nothing comes to mind that would truly allow VTAM to accomplish the statement above 100% of the time if it applied to all asynchronous exits, so it's good that the statement applies only to VTAM's asynchronous exit routine. You can of course mark a task as not interruptable by an asynchronous exit (such as by setting bit TCBNOIRB). But in order for an asynchronous exit to get control, that marking must not be set. And that means that as early as "instruction 1" of that exit, before it gets to set the task not interruptable, another asynchronous exit could interrupt it. Peter Relson z/OS Core Technology Design
