replace python benchmark with cargo bench

[arvidn]

cargo bench is more rigorous than my home-built run-benchmarks.py.

This patch does this:

• ports benchmark/run-benchmark.py to rust, specifically in the form of cargo bench benchmarks
• moves all the environments (arguments) passed into the benchmark programs to be generated programatically. Currently they are being generated in clvm-form, compiled and then loaded from disk when running, which is slower.
• pre-compiles the benchmark programs. compiling takes a surprisingly long time and there's no need to do that every CI run
• drop dependencies on clvm and clvm_tools. The remaining dependency on opc is tests/run-programs.py. To make this available in CI, I install clvm_tools_rs

[coveralls-official]

Pull Request Test Coverage Report for Build 6186220250

• 0 of 0 changed or added relevant lines in 0 files are covered.
• 45 unchanged lines in 2 files lost coverage.
• Overall coverage decreased (-0.8%) to 93.018%

---

Files with Coverage Reduction	New Missed Lines	%
src/chia_dialect.rs	5	67.52%
src/more_ops.rs	40	93.72%

Totals
Change from base Build 5953615517:	-0.8%
Covered Lines:	5156
Relevant Lines:	5543

---

💛 - Coveralls

[github-actions]

'This PR has been flagged as stale due to no activity for over 60
days. It will not be automatically closed, but it has been given
a stale-pr label and should be manually reviewed.'