Doesn't make a practical difference - but it's more correct this way.

Fixes race condition when epoll_ctl() gets called after the socket has
been closed, which may affect a different socket created by another
thread meanwhile.

When rpc_reconnect_requeue() gets called from inside nfs_service(),
the NfsInputStream can stall completely because the old socket has
been unregistered from epoll automatically, but the new one has never
been registered.  Therefore, nfs_service() will never be called again.

This kludge attempts to detect this condition by checking
nfs_which_events()==POLLOUT.

https://bugs.musicpd.org/view.php?id=4081

Same as in commit e02d8ad8, but this time for the sndio plugin
which can be emulated by libroar.

GCC 4.9 has incomplete C++14 support.  Specifically, it doesn't allow
switch/case in "constexpr" functions.

With "ok==false", newly created threads may quit instantly.

libupnp provides seconds, not whatever time unit is used by
std::chrono::steady_clock.

If the base class is not accessible, the "catching" the base class
won't work.  This caused the fatal error:

 terminate called after throwing an instance of 'LibmpdclientError'

Each close/open cycle resets the Filter's state, because a new Filter
instance is being created.  That results in the serials
(replay_gain_serial and other_replay_gain_serial) being out of sync
with the internal ReplayGainFilter state.

So instead of initializing those serials once, we need to initialize
them each time we create new ReplayGainFilter instances, i.e. in
OpenFilter().

 https://bugs.musicpd.org/view.php?id=4632