Running stable gentoo, =net-misc/tor-0.2.1.30. /etc/init.d/tor script does server reloading wrong. # start-stop-daemon --stop --signal HUP --pidfile /var/run/tor/tor.pid start-stop-daemon: 1 process refused to stop Patch against /etc/init.d/tor included, resolves the issue. Running without --stop argument is fine and HUP signal is sent successfully. # start-stop-daemon --signal HUP --pidfile /var/run/tor/tor.pid # echo $? 0 # tail -n 2 /var/log/tor/tor.log Aug 20 11:34:52.348 [notice] Received reload signal (hup). Reloading config and resetting internal state. Aug 20 11:34:52.348 [notice] Tor 0.2.1.30 opening log file. There is a cron daily task which rotates tor logs daily (/etc/logrotate.d/tor) and reloads tor server. Cron sends email warnings then! Reproducible: Always Steps to Reproduce: 1. start tor server 2. rc-service tor reload Actual Results: * Tor configuration (/etc/tor/torrc) is valid. * Reloading Tor configuration ... * start-stop-daemon: 1 process refused to stop Expected Results: * Tor configuration (/etc/tor/torrc) is valid. * Reloading Tor configuration ...
Created attachment 283997 [details, diff] Suppress daemon stop warning
With --stop it actually does receive the HUP and reload. Luckily the --stop fails and so it "works" but this is less than ideal. I'll put it in without a rev bump since it is not a show stopper bug, and those who have the current init still have something that works, albeit in a poor way. The next release is coming out soon and will definitely have it. Thanks for catching that.
Its in the tree. Can people please test. I know its an obvious fix, but because it does effect tor-0.2.1.30.ebuild which is marked stable, I want to make sure I'm not breaking anything. Reopen if there is any badness whatsoever. Thanks again Tomas.