dj at linuxfromscratch.org
Mon Jul 26 22:55:40 PDT 2004
Nathan Coulson wrote:
> so, think we should keep with the PIDFILE=/var/run/[progname] killproc
> [progname] then?
> I like the idea of making sure the actual file is terminated, by the
> program name in the end [we only got the problem trying to kill it].
Yep. If somebody configured the package wrong and PIDFILE is not set in
the init script, not a problem...it should've been set correctly. The
existing testcase was left alone for now.
Now, a finished idea. See attached (and locally tested) patch.
n/a - July 27, 2004
* Moved PIDFILE tests into getpids
* Removed signal arguments from killproc and reloadproc
funtions. Now default
signals are set at the top of functions and can be overridden
in the individual
init scripts by setting the variables STOPSIG, FALLBACK, and
-- DJ Lucas
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
More information about the lfs-dev