tweaks to benchmark-clvm-cost and fix sha256 per-byte measurement

[arvidn]

This is a collection of a few minor fixes to benchmark-clvm-cost.

• every ChiaDialect used to benchmark is created with flags 0. Create an abstraction to easily change the flags for all dialect objects.
• One assumption in measuring the cost-per-byte is that 0 bytes have 0 cost. When performing the linear regression, add a bias towards (0,0).
• Extend write_gnuplot_header() to allow better titles to the plots
• Fix issue in the sha256 operator test, where the parameter was hard-coded rather than substituted by the test atoms. This caused the cost-per-byte to always be 0.

per-byte benchmarks

before

time: per-byte: 2.78ns

after

time: per-byte: 2.86ns

[coveralls-official]

Pull Request Test Coverage Report for Build 20473952212

Details

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 91.549%

---

Totals
Change from base Build 20434068006:	0.0%
Covered Lines:	6294
Relevant Lines:	6875

---

💛 - Coveralls