i91@nikhefh.UUCP (Fons Rademakers) (12/01/87)
I already did sent this message once, but I wonder if it ever
got over the ocean:
I am having the following problem:
I have two processes each with a gpr event loop running at the same
time. They both use (gpr/dp)_$cond_event_wait calls to get their events.
I have to use cond_event_wait calls since I also have to check
periodically another event flag set in a mapped file by again
another process.
Now, the following happens: after a short period of time the
(gpr/dp)_$cond_event_wait calls don't return any more. Using
DPAT I can see that the programs are looping inside the cond_event_wait
routines.
When I suspend one of the programs (by making it an icon) the other
program runs well again. Also, when I change in one of the programs
the gpr_$cond_event_wait call to gpr_$event_wait both run well, but that
is not what I want (having to use gpr_$event_wait, that is).
I am using SR9.2.4. Is this problem related to the global gpr event count?
If so is this problem solved by switching to SR9.6?
Thanks for any help.
Regards, Fons Rademakers
--
Org: NIKHEF-H, National Institute for Nuclear and High-Energy Physics.
Mail: Kruislaan 409, P.O. Box 41882, 1009 DB Amsterdam, the Netherlands
UUCP: {cernvax,uunet}!mcvax!nikhefh!i91 or i91@nikhefh.hep.nl
BITNET: nikhefh!i91@mcvax.bitnet Telex: 10262 (hef nl)