🔥🚒🧯 Put out the fire

[ChrisCummins]

All CI jobs are failing and building from source, caused by a change in the checksum of the csmith dependency.

As best as I can tell, this checksum change is innocuous. By simply updating the checksum all tests continue to pass. As a precaution I've added a secondary URL for csmith tarballs.

This PR also includes a handful of tiny fixes I found while testing the #590 mitigation using make test.

Fixes #590.

[codecov-commenter]

Codecov Report

Merging #593 (207dbb9) into development (f2ca6c2) will increase coverage by 0.17%.
The diff coverage is n/a.

@@               Coverage Diff               @@
##           development     #593      +/-   ##
===============================================
+ Coverage        88.16%   88.34%   +0.17%     
===============================================
  Files              114      114              
  Lines             6708     6708              
===============================================
+ Hits              5914     5926      +12     
+ Misses             794      782      -12     

Impacted Files	Coverage Δ
compiler_gym/envs/compiler_env.py	92.32% <0.00%> (+0.49%)	⬆️
compiler_gym/envs/llvm/datasets/cbench.py	80.74% <0.00%> (+0.74%)	⬆️
compiler_gym/service/connection.py	79.01% <0.00%> (+0.98%)	⬆️
...loop_tool/service/loop_tool_compilation_session.py	91.15% <0.00%> (+3.40%)	⬆️

---

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 f2ca6c2...207dbb9. Read the comment docs.

[mostafaelhoushi]

LGTM

[ChrisCummins]

CI/test-linux-cmake is failing because of a flaky timeout in unrelated test_validate_benchmark_semantics tests. Normally I would re-run but merging anyway to unblock other development 🤞

Cheers,
Chris