[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #3049 [Tor Client]: Allow a Tor process to be âownedâ by a controller process
#3049: Allow a Tor process to be âownedâ by a controller process
-------------------------+--------------------------------------------------
Reporter: rransom | Owner: rransom
Type: enhancement | Status: accepted
Priority: major | Milestone: Tor: 0.2.2.x-final
Component: Tor Client | Version:
Keywords: | Parent:
Points: | Actualpoints:
-------------------------+--------------------------------------------------
Comment(by nickm):
BTW, you'll want to have the option named __OwningControllerPID so that
SAVECONF never writes it to a file, and so that it's treated as for-use-
by-controllers.
I'm not sure how you go about implementing that feature, fwiw. Is there a
standard way to watch for a process with a particular pid to change its
status? waitpid() and friends seem to only be for child processes. You
can kill(0) the pid, but that doesn't tell you if *that* process is still
there: it only tells you if some process with that pid is still there.
On BSDish systems, you can use kevent to wait for a process, but I don't
believe libevent exposes that atm. On linuxy systems, you could pull the
process's start time out of /proc/[pid]/stat and see whether it's changed
or somethign. On Windows, you can grab a handle to the process somehow
and poll it from time to time to see whether it's been signalled.... but
all of these approaches seem pretty heavyweight atm.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/3049#comment:4>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs