Skip to content

Conversation

@lifubang
Copy link
Member

No description provided.

@cyphar
Copy link
Member

cyphar commented Nov 18, 2025

"on" is the wrong word to use here, "through" is probably better. ("on" to me means that the bind-mount is on top of the dangling symlink, which isn't the case.)

Your tests here all use .. but that isn't actually an example nor root cause of the dangling symlink issues we had reported to us when I added 3f92552 (it would be nice to reference that commit as well). It would be nice to have a test with a symlink to /foo/bar/baz/a/b/c/d/e/f (and a relative version too), both of which would cause issues without ...

I would also check cat /proc/self/mounts (or at least look at the inode number with stat) to confirm it is actually a bind-mount in the test.

Did you want me to carry this in #4985 instead? (I was going to write tests there as well.)

@lifubang
Copy link
Member Author

Did you want me to carry this in #4985 instead? (I was going to write tests there as well.)

Yes

@lifubang lifubang force-pushed the ci-dangling-symlink branch from 3bb9deb to 82ec641 Compare November 20, 2025 07:31
@cyphar
Copy link
Member

cyphar commented Nov 22, 2025

Carried in #4985.

@cyphar cyphar closed this Nov 22, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants