[avalanche] Disconnect from the updated block tip handler before destructing the processor

Summary: Make sure we can't reach a use after free issue where UpdatedBlockTip() is called after the peer manager has been destructed. This can only happen in tests because the normal node shutdown take cares of this already.

Test Plan:

ninja all check-all

Reviewers: #bitcoin_abc, PiRK

Reviewed By: #bitcoin_abc, PiRK

Differential Revision: https://reviews.bitcoinabc.org/D18416