Brad wrote:
> Is there anything that describes "timer
sockets" (PF_TIMER). I've
> never seen these before and was wondering what problem they solved.
der Mouse wrote:
Briefly, they provide a file-descriptor interface to
the timer
facilities underlying setitimer(ITIMER_REAL) and SIGALRM.
OK, but I still don't understand what problem they solve.
Well, maybe I have an idea. The one program I used setitimer() in
would core dump if the system was heavily loaded and a second timer
signal occurred before service for an earlier one completed. Does the
file descriptor interface solve this somehow?
Eric