Trace 1793100270_2515 (in 0.8.0 batch 2): refinement gas exactly equals the budget

[emielsebastiaan]

Regarding trace 1793100270_2515 (in 0.8.0 batch 2): refinement gas exactly equals the budget

• Disputed trace: PolkaJam accepts; PyJAMaz returns refine_load_too_high.
• Cause: In this trace the summed refinement gas is exactly 1,000,000,000, Tiny’s G_R. PyJAMaz rejects gas_used >= G_R.
• GP: 14.9 prints the strict inequality Σ refinement budgets < G_R; However, the immediately preceding prose in GP says "at most".
• Proposed fix: Ambiguity in Gray Paper between prose and equation: solution either rephrase "at most" or replace both "<" with "<=".
• PyJAMaz follows the equation and PolkaJAM seems to follow the prose.

(bonus find: the same seems to apply to the accumulate gas limit (constant G_A)

https://graypaper.fluffylabs.dev/#/07f041d/1b9f011b9f01?v=0.8.0

[davxy]

Changed by gavofyork/graypaper#563

Part of the listed deviations in #212

[jaymansfield]

GP#563 was listed as a deviation in the PR for batch 1 and changes the GA and GR gas checks to be inclusive.

[emielsebastiaan]

Check... we missed this; closed this discussion.