Report more information about instances of VNA00J

[reprogrammer]

When I ran Keshmesh (9a1eed4) on Cassandra with an entry point at org.apache.cassandra.gms.Gossiper.start(int), it reported 110 instances of VNA00J. However, the bug markers provided little information about why the code had been marked as an instance of VNA00J. To confirm or reject the detected instances of the bug pattern, Keshmesh should provide more information, e.g. the shared field.

[reprogrammer]

In practice, I've been able to infer the root cause of the reported problems by tracking down the call graph. However, it would be nice to provide additional information for each marker.