panic at "no Task is currently running"

I'm trying to write some multithreaded WebSocket code. I'm sending a Remote to another thread then calling spawn on it and sending a message inside the callback. For some reason, it works fine for a bit and then panics like this:

thread 'main' panicked at 'no Task is currently running', ~/.cargo/registry/src/github.com-1ecc6299db9ec823/futures-0.1.18/src/task_impl/mod.rs:44:8
stack backtrace:
   0: std::sys::imp::backtrace::tracing::imp::unwind_backtrace
   1: std::sys_common::backtrace::_print
   2: std::panicking::default_hook::{{closure}}
   3: std::panicking::default_hook
   4: std::panicking::rust_panic_with_hook
   5: std::panicking::begin_panic
   6: futures::task_impl::current
   7: <tokio_core::net::tcp::TcpStream as std::io::Write>::write
   8: <tungstenite::protocol::frame::FrameSocket<Stream>>::write_pending
   9: <tungstenite::protocol::WebSocket<Stream>>::send_one_frame
  10: <tungstenite::protocol::WebSocket<Stream>>::write_pending
  11: <F as tokio_core::reactor::FnBox>::call_box
  12: tokio_core::reactor::Core::notify
  13: tokio_core::reactor::Core::consume_queue
  14: tokio_core::reactor::Core::poll
...

I'm not exactly sure how to debug this, especially since it only happens in a release build. Is it the case that there might not be a current task inside a call to "notify"? Is there any way to get it to be called inside a task? Or any way of adding extra debugging checks in tokio-core that could catch the issue?

Sorry to ask here but I'm not sure where to start. Please let me know if there's a better place to ask instead. I found https://github.com/rust-lang-nursery/futures-rs/issues/234 but I wasn't sure how to apply that to my situation. I'm not calling poll() or park() or anything, just core.run().

tokio-rs / tokio-core

panic at "no Task is currently running" #299