Ensure all listeners are cleaned up on ServiceBrowser cancelation #290
Conversation
bdraco
force-pushed
the
fix_listener_leak
branch
4 times, most recently
from
e3772ec
to
8569101
Compare
|
I added coverage for the areas that were previously not actually covered with assertions. |
|
May be one of the causes of home-assistant/core#38862 |
bdraco
force-pushed
the
fix_listener_leak
branch
4 times, most recently
from
dbc568a
to
7251dc7
Compare
When creating listeners for a ServiceBrowser with multiple types they would not all be removed on cancelation. This led to a build up of stale listeners when ServiceBrowsers were frequently added and removed.
bdraco
force-pushed
the
fix_listener_leak
branch
from
7251dc7
to
28a2d16
Compare
|
@jstasiak If you have a moment, could we get a release bump with this fix. I'm trying to close out home-assistant/core#38862 and home-assistant/core#38933 and this is the one of the last ones to be resolved. |
zeroconf/test.py
Outdated
| @@ -860,6 +861,20 @@ def test_cache_empty_does_not_leak_memory_by_leaving_empty_list(self): | |||
| cache.remove(record2) | |||
| assert 'a' not in cache.cache | |||
|
|
|||
| def test_cache_empty_multiple_calls_does_not_throw(self): | |||
There was a problem hiding this comment.
Does this patch actually fix something that's verified by this test? Let's keep this out of this pull request if that's not the case.
There was a problem hiding this comment.
I'll pull it out. It should probably go in a new PR after this merges
|
Sure, I'll make a release after merging. |
|
Just released as 0.28.1. :) |
|
TYVM |
Sorry I missed this in the original submission #249. 馃檴
When creating listeners for a ServiceBrowser with multiple types
they would not all be removed on cancelation. This led
to a build up of stale listeners when ServiceBrowsers were
frequently added and removed.