Hi Petter! On 21.03.2010 16:14, Petter Reinholdtsen wrote: > [Petter Reinholdtsen] >> I see the point of only signaling init when the socket changed, but >> am unsure if we want to look for upstart specific binaries in the >> generic scripts. Is there still a bug with upstart if this change >> is implemented instead? > > I commited and uploaded the change to move the kill command in version > 2.87dsf-9. Please close this bug if it is enough to not conflict with > upstart, or provide more information if the other part of the patch is > also needed.
The current code -9 is not sufficient unfortunately. upstart does not use (or for that matter create) /dev/initctl, so the check if [ ! -p $INITCTL ] will be true when upstart is installed and the SIGUSR1 signal sent to pid 1. So I'd appreciate if the additional check "&& [ ! -x /sbin/initctl ]" would be added. It ensures that this code is not run with upstart (and it is not needed anyway with upstart). > >> I am not sure if this code is still needed, but believe we should >> keep it until someone can verify that it is obsolete. > > If someone can check if this code creating the socket in /dev/ is > obsolete, that would be valuable too. :) If it could be removed altogether, even better. But I'm not qualified to answer that. Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
signature.asc
Description: OpenPGP digital signature