Petter Reinholdtsen <p...@hungry.com> writes: > [Robbie Harwood] >> It's pretty common for SIGHUP to reload config as well as SIGUSR1. >> >> So, my suggestion would be define a new way of doing this, file bugs >> with patches against anyone using do_reload_sigusr1, and then remove >> do_reload_sigusr1. >> >> One way that might be nice is to make a function do_reload_signal, which >> would send $SIGNAL (or $RELOAD_SIGNAL) as an argument to >> start-stop-daemon. > > I did not quite understand the problem. Can you explain in a bit more > detail? > > Today scripts wanting to use SIGHUP to reload can implement do_reload() > similar to the way do_reload_sigusr1() is implemented. Those wanting > SIGUSR1 can use do_reload_sigusr1() by providing it as an alias for > do_reload(). What more is needed?
As per this bug's original filing: do_reload_sigusr1() calls SIGHUP, not SIGUSR1.
signature.asc
Description: PGP signature
