fix data race in dynamic plugin registry tests #12554
Conversation
| for { | ||
| select { | ||
| case <-ctx.Done(): | ||
| t.Fail() |
There was a problem hiding this comment.
Perhaps use t.Errorf here? It's still safe for concurrent use (unlike Fatalf and FailNow) but would allow you to output a hint as to what went wrong.
At a glance I'm not sure what a failure here means other than "an unexpected thing happened," but I'm also not familiar with these tests.
There was a problem hiding this comment.
Good call. The original tests were just "we didn't finish" and that's not very helpful to repeat when fixing the test 😁
| rcv1 = true | ||
| if rcv1 { | ||
| return | ||
| } |
There was a problem hiding this comment.
rcv1 seems like a noop. Can this be replaced with return?
There was a problem hiding this comment.
Oops, yup. Copypasta from the previous test. Will fix.
These tests have a data race where the test assertion is reading a value that's being set in the `listenFunc` goroutines that are subscribing to registry update events. Move the assertion into the subscribing goroutine to remove the race. This bug was discovered in #12098 but does not impact production Nomad code.
tgross
force-pushed
the
b-registry-data-race
branch
from
1bb0958
to
ae1b029
Compare
|
I'm going to lock this pull request because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active contributions. |
These tests have a data race where the test assertion is reading a
value that's being set in the
listenFuncgoroutines that aresubscribing to registry update events. Move the assertion into the
subscribing goroutine to remove the race. This bug was discovered
in #12098 but does not impact production Nomad code.