Catch BlockingIOError in job_server.

[mikepurvis]

I was getting some instances of build jobs randomly bailing out when running the master branch under Python 3.8; I'm assuming as a consequence of #574:

[build] An internal error occurred!
Traceback (most recent call last):
  File "/usr/local/lib/python3.8/dist-packages/catkin_tools/verbs/catkin_build/build.py", line 555, in build_isolated_workspace
    all_succeeded = run_until_complete(execute_jobs(
  File "/usr/local/lib/python3.8/dist-packages/catkin_tools/execution/executor.py", line 374, in run_until_complete
    return loop.run_until_complete(coroutine)
  File "/usr/lib/python3.8/asyncio/base_events.py", line 616, in run_until_complete
    return future.result()
  File "/usr/local/lib/python3.8/dist-packages/catkin_tools/execution/executor.py", line 247, in execute_jobs
    (job_server.try_acquire() is not None)):
  File "/usr/local/lib/python3.8/dist-packages/catkin_tools/execution/job_server.py", line 426, in try_acquire
    token = JobServer._acquire()
  File "/usr/local/lib/python3.8/dist-packages/catkin_tools/execution/job_server.py", line 262, in _acquire
    token = os.read(cls._job_pipe[0], 1)
BlockingIOError: [Errno 11] Resource temporarily unavailable

I believe this is a proper fix to the issue.

FYI @timonegk, @wjwwood

[mikepurvis]

Both of these derive from OSError. Any other OSError exceptions will raise through, as the behaviour was before.