This is a known problem since PrimeFaces 3.0. It's caused by a change in how it checks if the current HTTP request is an ajax request. It's been identified by a request parameter instead of a request header. When a request parameter is retrieved for the first time before the JSF view is restored, then all request parameters will be parsed using server's default character encoding which is often ISO-8859-1 instead of JSF's own default character encoding UTF-8. For an in depth explanation see Unicode input retrieved via PrimeFaces input components become corrupted.
One of the solutions is to create a filter which does a request.setCharacterEncoding("UTF-8")
.
@WebFilter("*.xhtml")
public class CharacterEncodingFilter implements Filter {
@Override
public void doFilter(ServletRequest request, ServletResponse response, FilterChain chain) throws ServletException, IOException {
request.setCharacterEncoding("UTF-8");
chain.doFilter(request, response);
}
// ...
}