Just because of interest:
the program mask was set by the ESPIE macro, according to the
interrupts you wanted to be caught by the ESPIE routine, and it had
nothing to do with the ATTACH macro, which attached the subtask;
am I correct? So it is in fact an ESPIE topic, not related to subtasking?
And not related to LE?
Do you know, if ATTACH does anything to the program mask, or is
the program mask for the subtasks simply inherited from the master task?
Thanks for clarifying ... have a nice day.
Bernd
Am 09.11.2020 um 18:25 schrieb Thomas David Rivers:
Thomas David Rivers wrote:
I seem to have a situation where the program mask for my attached
sub-task is getting all the bits set (indicating, for example, the
receipt
of, say, FIXED-POINT OVERFLOW) when the parent task does not
have those set.
Just to reply to myself - for any possible future where this question
pops up
again.
It turned out to be the ESPIE macro. Apparently if you set up an
ESPIE requesting
a particular interrupt be caught, the OS sets the program-mask to
catch that interrupt.
We were just putting a "blanket" ESPIE to catch everything - so *poof*
- the PSW
program mask was changed to do that.
I suppose that's "helpful" - but not quite what was expected; we'll
have to do
a little program mask dance around setting ESPIEs to avoid that.
I wonder if this magic changing-of-the-program-mask is documented?
- Dave R. -
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
