[TEST] Testing Framework Issues -- Segmentation Faults

[srajadnya1]

I've noticed that when a test fails because of a segmentation fault, it causes the whole testing framework to bug out since runtime fails to cleans itself properly.

For example, after restarting my docker and running a test that works, I'll uncomment a previously uncommented block of code that is intended to fail -> raises a seg fault. After the seg fault is raised and the test fails, I'll re-comment out the buggy block of code and run the test again -> the test that previously worked now fails.

I was thinking that we force runtime to clean itself preemptively when calling start_test(). Any thoughts on this?

[benliao1]

Perhaps related to #179 ?