Skip to content

Conversation

fishcakez
Copy link
Member

Previously a backend could be removed from logger and the watcher would not have exited before the Logger.remove_backend/1 call returned. This meant that a new watcher could not always be started because the child (watcher) might still exist but be about to die.

I need to investigate whether switching to a simple_one_for_one with non-temporary children that can return :ignore will cause a problem - it might.

Previously a backend could be removed from logger and the watcher would
not have exited before the remove_backend/1 call returned. This meant
that a new watcher could not always be started because the child
(watcher) might still exist but be about to die.
josevalim pushed a commit that referenced this pull request Aug 23, 2014
Fix logger watcher race condition
@josevalim josevalim merged commit 6827431 into elixir-lang:master Aug 23, 2014
@fishcakez fishcakez deleted the logger_rc branch April 22, 2017 14:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

Successfully merging this pull request may close these issues.

2 participants