Assembly-level optimizations

[iamdefinitelyahuman]

What I did

• find and remove unreachable instructions within the assembly
• merge sequential jumpdests

How I did it

Added some logic in assembly_to_evm within compile_lll.py

Both of these situations occur as a result of LLL generation happening in small chunks that don't see the "bigger picture" around them. There is likely a more elegant solution here, but this is effective and simple to implement. I've left comments about what's happening, we can try to address this properly as we refactor out parser in the future and give ourselves more effective data structures to catch and fix this as it happens rather than after the fact.

How to verify it

Run tests, confirm that nothing broke. If this created any issues we'd see tests failing everywhere, especially within the e2e tests of the examples.

Cute Animal Picture

Notice me senpai!

[codecov-io]

Codecov Report

Merging #2304 (ec4b289) into master (fbd6880) will decrease coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #2304      +/-   ##
==========================================
- Coverage   85.82%   85.81%   -0.02%     
==========================================
  Files          85       85              
  Lines        8546     8560      +14     
  Branches     2036     2042       +6     
==========================================
+ Hits         7335     7346      +11     
- Misses        728      730       +2     
- Partials      483      484       +1     

Impacted Files	Coverage Δ
vyper/compile_lll.py	92.85% <100.00%> (+0.29%)	⬆️
vyper/functions/functions.py	89.25% <0.00%> (-0.40%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fbd6880...a6b4fab. Read the comment docs.

[fubuloubu]

Big change, but if it passes the test suite, hopefully should work just fine.