runtime: TestMemStats fails after running tests and benchmarks for some time #22696
Note that the test succeeds fine on its own, even when it's run thousands of times.
Also note that it took my laptop over 5 minutes to reach this point. Might be reached faster on a more powerful machine.
Hmm. It's not really clear what we can do about this.
Maybe it should keep track of how many times it's run and start skipping after some N?