You MUST await a future returned from `loop.run_in_executor()` to avoid the leak.

Yuri, what should we do with the issue? I see the second similar report in the last half of the year.
Sure, we can add weakrefs somewhere in futures._chain_future() but I pretty sure that the proper fix is just enforcing `await` here, e.g.

1. rename existing run_in_executor() into a private _run_in_executor() function.

2. write async trampoline:
async def run_in_executor(self, executor, func, *args):
    return await self._run_in_executor(executor, func, *args)