SIGHUP configuration reload does not update retry_join

Hi,

currently when I send a `kill -HUP` to the serf process, I can see

```
==> Caught signal: hangup
==> Reloading configuration...
```

in the logs. However, changes to `retry_join` in the configuration file are not reflected, and serf continues to try to join incorrect nodes (those that were in the config file before the change).

From the documentation it was not clear this part of the configuration is not reloaded on SIGHUP.

I'm using serf 0.6.4 but I didn't see obvious changes in the source code on 0.7.0.

Would it be possible to allow SIGHUP to also reload `retry_join` ?

Thanks.