Fix MOVE to -(An) destination cycle count bug

[habedi]

• The m68k overlapped the address predecrement with the write bus cycle for MOVE instructions targeting -(An), taking 2 fewer cycles than the generic effective address cost. Rocket 68 was charging the full EA table value (e.g., 6 for word, 10 for long), making every MOVE to a predecrement destination 2 cycles too slow. This affected all 36 MOVE opcode variants (12 source modes x 3 sizes) with -(An) destination, including common patterns like MOVE.L Dn,-(SP) for stack pushes.
• Added tests to prevent regression and set Rocket 68 version to 0.2.1.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 75.91%. Comparing base (74c60ae) to head (9b0415b).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #16      +/-   ##
==========================================
+ Coverage   75.67%   75.91%   +0.24%     
==========================================
  Files          18       18              
  Lines        4468     4496      +28     
  Branches      433      433              
==========================================
+ Hits         3381     3413      +32     
+ Misses       1077     1073       -4     
  Partials       10       10              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.