On my laptop (Ubuntu 17.10, Python 2.7.14 in a virtualenv),
`test_regular_mod` fails with
```
AssertionError: "\nimport sys\n\n\ndef say_hi():\n print 'hi'\n" !=
'\x03\xf3\r\n\xbbW\xd5Yc\x00\x00\x00\x00\x00\x00\x00\x00\x02\x00\x00\x00@\x00\x00\x00s\x19\x00\x00\x00d\x00\x00d\x01\x00l\x00\x00Z\x00\x00d\x02\x00\x84\x00\x00Z\x01\x00d\x01\x00S(\x03\x00\x00\x00i\xff\xff\xff\xffNc\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00C\x00\x00\x00s\t\x00\x00\x00d\x01\x00GHd\x00\x00S(\x02\x00\x00\x00Nt\x02\x00\x00\x00hi(\x00\x00\x00\x00(\x00\x00\x00\x00(\x00\x00\x00\x00(\x00\x00\x00\x00sF\x00\x00\x00/home/alex/src/mitogen/tests/data/module_finder_testmod/regular_mod.pyt\x06\x00\x00\x00say_hi\x05\x00\x00\x00s\x02\x00\x00\x00\x00\x01(\x02\x00\x00\x00t\x03\x00\x00\x00sysR\x01\x00\x00\x00(\x00\x00\x00\x00(\x00\x00\x00\x00(\x00\x00\x00\x00sF\x00\x00\x00/home/alex/src/mitogen/tests/data/module_finder_testmod/regular_mod.pyt\x08\x00\x00\x00<module>\x02\x00\x00\x00s\x02\x00\x00\x00\x0c\x03'
```
`__file__` contains the path of the compiled `.pyc`, not the `.py`
source file.
Since the above if block ends in a call to os.execv() this block will
only ever run when the if condition was false. Hence putting it in an
else clause is unnecessary.
Ubuntu 17.04 provides Docker 1.12.6, which has API version 1.24.
`dev_requirements.txt` specifies the docker-py 2.5.1, which by default
requests API version 1.30.
Hence when the SSH unit tests try to run the container specified in
`DockerizedSshDaemon` an error occurs
```
APIError: 400 Client Error: Bad Request ("client is newer than server
(client API version: 1.30, server API version: 1.24)")
```
On Ubuntu 17.10 something (probably Docker) appears to be accepting
connections, before sshd is fully ready. This results in a race
condition, and hence connection errors for the first few tests (2-3 on
my laptop).
testlib.wait_for_port() checks not only that the port can be connected
to, but also something resembling the sshd banner is sent.
Fixes#51
py_packages is not a field know by distutils or setuptools. The closest
is `py_modules`, which perhaps what the erroneus line here started as.
Fixes#43
Without this, it's possible for Waker to be start_received() after the
shutdown signal has already been sent, resulting in 5 second delay
during shutdown.
Additionally mask EBADF during os.write() to waker's write side.
Necessary since nothing synchronizes writer threads from the broker
thread during shutdown. Could be done with a lock instead, but this is
cheaper.
Can't figure out what it's supposed to do any more, and can't find a
version of Ansible before August 2016 (when I wrote that code) that
seems to need it.
Add some more mitigations to avoid sending dylibs.
Alex Willmer