[mythtv] Mythfilldatabase consuming 100% cpu

[Torbjörn Jansson]

I think there is something wierd going on with mythfilldatabase sometimes.
Mythfilldatabase sometimes decides to use 100% cpu and not exit properly,
i've seen this only once before a few weeks ago.

Before i killed mythfilldatabase the backend said "mythfilldatabase still
running, skipping checks." every 5 minutes.

When i killed mythfilldatabase this was logged in the backend log:
sh: line 1: 24519 Killed                  mythfilldatabase --refresh-today
>>/mnt/store/mythfilldatabase.log 2>&1
2006-06-20 17:53:11.176 MainServer::HandleAnnounce Monitor

And then 5 seconds later:
2006-06-20 17:53:19.984 MythSocket(81dca58:-1): writeStringList: Error,
socket went unconnected.

The last entries from mythfilldatabase was:
2006-06-20 02:05:41.408 
===============================================================
| Attempting to contact the master backend for rescheduling.  |
| If the master is not running, rescheduling will happen when |
| the master backend is restarted.                            |
===============================================================
2006-06-20 02:05:41.428 Connecting to backend server: 192.168.0.1:6543 (try
1 of 5)
2006-06-20 02:05:41.430 Using protocol version 30
2006-06-20 02:05:42.171 mythfilldatabase run complete.

So it looks as if it completed as it should, but instead of exiting nicely
it started to consume 100% cpu.

Could this problem have something to do with the recent socket changes?

Unfortunately i didn't think of attaching gdb to mythfilldatabase and
getting a backtrace before i killed it.
If it happends again, i'll try to remeber to get a bt first.