Hmm,
I tried this - and the processes do exit.
... so that's OK ...
I've commented out my smtp_receive_timeout and I'll see how it goes with the default.
Thanks for your time.
>
>
> > Yes,
> > they're blocked in
> > read(1,0x... ,8192);
>
> Looks like the same old problem. I'm afraid I really don't know how to
> go about debugging this one any further, because, as I said, every test
> I set up for a timing out connection works fine. It's as though its lost
> the alarm() signal, or possibly it has lost the signal handling setting.
> Does a process loosen up (and complete) if you send it a SIGALRM signal?
>
> --
> Philip Hazel University of Cambridge Computing Service,
>
--
All the Best to you and yours.
Jonathan Kyme
jrk@???
