Skip to content

fix: reap the Vale child when writing its stdin fails - #34

Open
chad-loder wants to merge 1 commit into
vale-cli:mainfrom
chad-loder:fix/reap-vale-child
Open

chad-loder wants to merge 1 commit into
vale-cli:mainfrom
chad-loder:fix/reap-vale-child

Conversation

@chad-loder

Copy link
Copy Markdown

If Vale exits before reading all of stdin (a config error, say), write_all fails with a broken pipe and the early return drops the Child without waiting on it, leaving a zombie per failed lint for the life of the server. The child is now always waited on, and its own output, which explains the failure, is what gets parsed.

The test uses a stand-in vale that exits without reading a 1 MiB buffer.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant