[tor-bugs] #3049 [Tor Client]: Allow a Tor process to be ‘owned’ by a controller process
Tor Bug Tracker & Wiki
torproject-admin at torproject.org
Fri May 13 07:34:27 UTC 2011
#3049: Allow a Tor process to be ‘owned’ 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 rransom):
Replying to [comment:6 arma]:
> The "TAKEOWNERSHIP" part should be quite simple and, I agree, useful. It
should get us most of what we want.
>
> I wonder if we might want to leave the OwningControllerPID part to Tor
0.2.3 though, since it may introduce bugs (e.g. portability issues) that
will take a while to settle.
We can postpone the OS-specific extensions to `__OwningControllerProcess`
until 0.2.3.x, or test them on 0.2.3.x and backport later. But we
currently only support Unixoid systems and Windows NT systems, and the
generic Unixoid implementation is both highly portable across Unixoids and
Good Enough for a first implementation.
(I've renamed the option so that we can leave room for additional OS-
specific arguments after the PID, because that will allow us to make that
piece of process ownership less race-prone.)
> On the Vidalia implementation side, I believe vidalia already starts out
by running 'tor --version' before trying to run tor for reals. See
TorProcess::version() in src/torcontrol/TorProcess.cpp. So it should be
able to predict whether Tor will recognize the OwningControllerPID option.
Or it could just refuse to start a too-old version of Tor.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/3049#comment:9>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
More information about the tor-bugs
mailing list