[tor-bugs] #17946 [Stem]: launch_tor should kill process on exception while bootstrapping (for example, on KeyboardInterrupt)
Tor Bug Tracker & Wiki
blackhole at torproject.org
Wed Dec 30 16:00:16 UTC 2015
#17946: launch_tor should kill process on exception while bootstrapping (for
example, on KeyboardInterrupt)
-----------------------------------+------------------------
Reporter: germn | Owner: atagar
Type: defect | Status: closed
Priority: Medium | Milestone:
Component: Stem | Version:
Severity: Normal | Resolution: fixed
Keywords: launch_tor, bootstrap | Actual Points:
Parent ID: | Points:
Sponsor: |
-----------------------------------+------------------------
Changes (by atagar):
* status: new => closed
* resolution: => fixed
Comment:
> If KeyboardInterrupt (or asyncio.CancelledError, or other) was raised,
we should reraise exactly this error after resources cleanup, we shouldn't
raise our type of error instead. Top-level code might want to know if
exactly KeyboardInterrupt happened.
Yup, I can see an argument for this. Reason I did this was because
launch_tor() is only documented as raising OSErrors. I'm unfamiliar with
asyncio.CancelledError, but you're right that KeyboardInterrupt is an
interesting critter. If the caller explicity invokes an exception within
us then I could see an argument for re-raising that.
[https://gitweb.torproject.org/stem.git/commit/?id=ae9d81b279bb51b2f1c3d088b86c7aa310b9db11
Changed].
> One another little thing is about line 141:
I removed a couple tor_process.kill() calls since they're now covered by
our except block. No need to kill the process twice. :)
Feel free to reopen if ya spot anything else.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/17946#comment:3>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
More information about the tor-bugs
mailing list