The problem is that pg_receivexlog calls select(2) with timeout=0 and
goes into busy loop when --status-interval option is set to 0. This bug
was introduced by the commit,
74cbe966.

Per report from Sawada Masahiko