[RC5] Linux ....

gindrup at okway.okstate.edu gindrup at okway.okstate.edu
Mon Mar 30 14:41:40 EST 1998


     Theoretically, the SIGHUP signal is sent to every proc whose parent 
     is the shell instance.
     
     nohup disassociates the proc from the shell by making it the lead 
     process in its own job/process group.  Screen is supposed to do 
     something similar, but I think it overrides its own signal handlers 
     to pass signals it receives to all of the non-detached screens it 
     owns.  (The internals of screen are unknown to me, but this is 
     equivalent to observed experimental behaviour.)
     
     HP-UX doesn't appear to implement nohup this way since *some* 
     nohup'ped procs are sent SIGNUP on logout.  The discriminator seems 
     to be that processes that were nohup'ped directly from the shell do 
     not have their parent's id set to 0 and thus receive signals 
     propagated from the parent, but procs that were nohup'ped from a 
     script (and thus another (sub-)shell) do correctly set their parent 
     id (to 0, indicating that they are lead procs) and do not receive 
     the SIGHUP.
     
     This was discussed a few months ago on the list.  I think a short 
     list of *ix variants were identified as misimplementing nohup unless 
     "odd" invocations were used.  I'd hit the archives for detailed 
     discussions.
            -- Eric Gindrup ! gindrup at okway.okstate.edu


______________________________ Reply Separator _________________________________
Subject: Re: [RC5] Linux .... 
Author:  <rc5 at llamas.net> at SMTP
Date:    3/27/98 5:27 PM


     [snip]
Hmm... you are just running it like "rc5des" from the bash command-line? 
If so, then how do you logout?  You probably want to be running "rc5des 
&", then when you logout it shouldn't be killed -- but it might be anyway. 
Run "nohup rc5des &", just to be on the safe side.  (Explanation, for 
those inclined: on exit, the shell normaly sends a "hup" (for hangup) 
signal to some of the processes that it spawned (I'm unclear on the 
alogrithim that it uses to decide which).)
     
        -=- James Mastros
--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net 
rc5-digest subscribers replace rc5 with rc5-digest
     
     

--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list