Is there a way to make it easier to debug failing, non-terminating, tests?

Question

If a #[test] function accidentally loops forever, the test suite doesn't finish. If you kill it (eg ctrl-c), cargo test seems to exit silently, so you get neither a stack trace (if enabled) or a report of which tests passed or failed.

Is there a way to make it easier to debug failing, non-terminating, tests?

Answer 1

I've come across timebomb which looks like close to what I need, but does mean manually wrapping every test; i.e. instead of:

#[test]
fn test() {
    assert!(true);
}

I need to do:

extern crate timebomb;
use timebomb::timeout_ms;

#[test]
fn test() {
    timeout_ms(|| {
        assert!(true);
    }, 1000);
}

which is a pain (but admittedly one-time) for dozens of tests. But wait; Rust has macros! This actually seems like a reasonable solution:

extern crate timebomb;
use timebomb::timeout_ms;

macro_rules! timeout_test {
    ( $name:ident() $code:block ) => {
        #[test]
        fn $name() {
            timeout_ms(|| $code, 1000);
        }
    }
}

// the hard way
#[test]
fn foo() {
    timeout_ms(|| {
        loop {}
    }, 1000);
}

// the now easy way
timeout_test!(bar() {
   loop {}
});