daemon: restart child process if it died before signaling its readiness

daemon: restart child process if it died before signaling its readiness

The child process (the one being monitored) could die before it was able
to call fork_notify_startup() function.  If such situation arises, then
parent process (the one monitoring child process) would also terminate
with a fatal log message:

...|EMER|fork child died before signaling startup (killed (...))

This patch changes that behavior by always restarting child process
if it was able to start up at least once in the past.  However, if
child was not able to start up even once, then the monitor process
would still terminate, because that would most likely indicate a
persistent programming or system error.

To reproduce use following script:

while : ; do kill -SIGSEGV `cat /var/run/openvswitch/ovs-vswitchd.pid`; done

Signed-Off-By: Ansis Atteka <aatteka@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
VMware-BZ: 1273550