Great, thanks.
$ pacman -Ql sane-runit
sane-runit /etc/
sane-runit /etc/runit/
sane-runit /etc/runit/sv/
sane-runit /etc/runit/sv/saned/
sane-runit /etc/runit/sv/saned/run
$ ls -l /run/runit/service/sane-runit
lrwxrwxrwx 1 root root 24 Apr 28 15:57 /run/runit/service/sane-runit -> /etc/runit/sv/sane-runit
so
# rm /run/runit/service/sane-runit
and again
# ln -s /etc/runit/sv/saned /run/runit/service/
# sv status saned
down: saned: 0s, normally up, want up
# sv start saned
timeout: down: saned: 0s, normally up, want up
# sv status saned
down: saned: 1s, normally up, want up
# sv up saned
# sv status saned
run: saned: (pid 6099) 0s
# sv status saned
down: saned: 0s, normally up, want up
and the last status is the same after reboot. Does it mean the service running is correctly now?