c286f4f107
Add tests/ansible/requirements.txt to Tox.
2018-11-05 13:00:47 +00:00
cd6486b0e9
tests: fix more DisconnectTest raciness.
2018-11-05 13:00:47 +00:00
1c24a13560
tests: add Ansible back to requirements
...
Needed for Tox
2018-11-05 13:00:47 +00:00
8d709fdfb4
Merge remote-tracking branch 'origin/dmw'
...
- #418
- 2.6/3.x compat
- earlier bugfix fallout
2018-11-04 20:45:31 +00:00
6d5facec4c
su/sudo: fallout from previous commits
...
issue #418 and FD cleanup work.
2018-11-04 20:43:29 +00:00
e180d310b5
tests: fix fork_test compat on 3.x.
2018-11-04 20:27:52 +00:00
a7eca5b55e
docs: update Changelog.
2018-11-04 20:20:41 +00:00
045db6f689
Fix iter_read() FD leaks on 3.x; closes #418 .
2018-11-04 20:18:23 +00:00
3f46c9569c
tests: 3.x syntax compat for tests/data/stubs/
2018-11-04 20:17:09 +00:00
27a4001f4f
tests: handle NameError when faulthandler is not installed.
2018-11-04 19:49:17 +00:00
5482b4d528
tests: poller_test 3.x fix.
2018-11-04 19:48:42 +00:00
6e1f9e2596
core: 2.6 str.decode() compat fix.
2018-11-04 19:47:25 +00:00
3836c6a220
tests/bench: run roundtrip.py a ton more to reduce variance
2018-11-04 19:45:02 +00:00
d1b7c232bf
tests: image_prep needs sudo
2018-11-04 18:57:11 +00:00
1064778dfd
Merge remote-tracking branch 'origin/dmw'
...
- #413
2018-11-04 17:52:30 +00:00
7a1dfa388a
docs: update Changelog; closes #413 .
2018-11-04 17:51:51 +00:00
76ec4f201c
issue #413 : paper over harmless duplicate del_route()
...
Ideally it would only be called once, and in future maybe it can, but
right now we need to cope with these cases:
* Downstream parent notifies us of disconnection (DEL_ROUTE)
* We notify ourself of disconnection
* We notify ourself and so does downstream parent
It's case 3 that causes the error.
2018-11-04 17:49:45 +00:00
fea0fb41fc
docs: update Changelog; closes #288
2018-11-04 16:06:52 +00:00
33011af9a5
issue #387 : update Changelog.
2018-11-04 15:29:06 +00:00
cf97932fad
core: dead messages have optional body, use it everywhere; closes #387 .
2018-11-04 15:26:25 +00:00
fdcd6810e1
Merge remote-tracking branch 'origin/dmw'
...
- issue #416 .
2018-11-04 14:38:52 +00:00
176fe55bbd
issue #416 : update Changelog
2018-11-04 14:38:39 +00:00
c09780aeb0
core: fix add_handler(respondent=..) memory leak
...
Closes #416 .
2018-11-04 14:38:39 +00:00
7141e9c11f
Merge remote-tracking branch 'origin/dmw'
...
- move more docs into source code
- make tests detect leaked threads, FDs
- #405 , #406 , #417
2018-11-04 13:55:10 +00:00
4ac9cdce7c
docs: update Changelog; closes #417 .
2018-11-04 13:54:34 +00:00
01e65d7865
docs: update Changelog; closes #406 .
2018-11-04 13:53:43 +00:00
802de6a8d5
issue #406 : clean up DiagLogStream handling and connect() failure.
...
When Stream.connect() fails, have it just use on_disconnect(). Now there
is a single disconnect cleanup path.
Remove cutpasted DiagLogStream setup/destruction, and move it into the
base class (temporarily), and only manage the lifetime of its underlying
FD via Side.close(). This cures another EBADF failure.
2018-11-04 13:48:34 +00:00
e01c8f2891
issue #406 : 3.x syntax fixes.
2018-11-04 11:57:50 +00:00
586c6aca9a
issue #406 : unix: fix ordering of stop_receive/close.
2018-11-04 11:52:02 +00:00
661e274556
issue #406 : ensure is_path_dead() socket is finalized.
2018-11-04 11:49:24 +00:00
411af6c167
issue #406 : unix: don't leak already-closed socket object
...
if Side.close() closes the socket (which it does), and it gets reused,
GC will cause socketobject.__del__ to later delete some random FD.
2018-11-04 11:25:22 +00:00
dc3db49c5a
issue #406 : more leaked FDs when create_child() fails.
2018-11-04 02:35:56 +00:00
003526ef7b
issue #406 : fix thread leaks in unix_test too.
2018-11-04 01:54:50 +00:00
17631b0573
issue #406 : parent: close extra_fd on failure too.
2018-11-04 01:50:19 +00:00
b3841317dd
issue #406 : clean up FDs on failure explicitly
...
The previous approach was crap since it left e.g. socketpair instances
lying around for GC with their underlying FD already closed, coupled
with FD number reuse, led to random madness when GC finally runs.
2018-11-04 01:49:10 +00:00
eae1bdba4e
tests: make minify_test print something useful on failure
2018-11-04 01:48:04 +00:00
802efa6ea6
issue #406 : ensure broker_test waits for broker exit.
2018-11-03 20:55:30 +00:00
10af266678
issue #406 : attempt Broker cleanup in case of a crash.
2018-11-03 20:54:05 +00:00
175fc377d2
tests: remove hard-wired SSL paths from fork_test.
2018-11-03 20:38:39 +00:00
3da4b1a420
tests: verify only main/watcher threads exist at teardown
2018-11-03 20:13:11 +00:00
b0dd628f07
issue #406 : parent_test fixes, NameError in log_fd_calls().
2018-11-03 19:49:13 +00:00
375182b71b
issue #406 : don't leak side FDs on bootstrap failure.
2018-11-03 19:48:36 +00:00
14b389cb46
issue #406 : don't leak FDs on failed child start.
2018-11-03 15:49:52 +00:00
6ff1e001da
issue #406 : log socketpair calls too.
2018-11-03 15:48:31 +00:00
70c550f50c
issue #406 : close stdout pipes in parent_test
2018-11-03 15:02:54 +00:00
9b3cb55a8b
issue #4096 : import log_fd_calls() helper.
2018-11-03 15:01:45 +00:00
8a0b343760
issue #406 : test for FD leak after every TestCase
2018-11-03 13:28:37 +00:00
e9a6e4c3d2
issue #406 : add test.
2018-11-03 12:39:29 +00:00
4230a93557
issue #406 : update Changelog.
2018-11-02 17:05:22 +00:00
d1c2e7a834
issue #406 : call Poller.close() during broker shutdown.
2018-11-02 17:03:27 +00:00
David Wilson