tools/gasbench: per-opcode EVM gas-vs-execution-time differential microbenchmark (F2 MVP) - #3761

Open

bdchatham wants to merge 20 commits into

mainfrom

brandon2/gasbench-mvp

Commits on Jul 15, 2026

refactor(tools/gasbench): close xreview R1 findings, move rationale to docs
bdchatham
committed
fix(tools/gasbench): correct the doc-migration's own factual error
bdchatham
committed
fix(tools/gasbench): never report Significant from an unestimated Uncertainty
bdchatham
committed
test(tools/gasbench): close xreview R4 advisory items
bdchatham
committed
docs(tools/gasbench): add operator Quickstart runbook to README
bdchatham
committed
docs(tools/gasbench): correct -count file semantics in runbook, validated E2E
bdchatham
committed
fix(tools/gasbench): distinct operands per case; emit nominal const_gas

bdchatham
and
claude
committed
feat(tools/gasbench): cross-run statistics on x/perf/benchmath; effect-size floor

bdchatham
and
claude
committed
test+docs(tools/gasbench): peer-sweep follow-ups

bdchatham
and
claude
committed
Merge branch 'main' into brandon2/gasbench-mvp
bdchatham
authored
fix(tools/gasbench): address PR review findings (NaN bounds, lint gate)

bdchatham
and
claude
committed
docs(tools/gasbench): EC2 recipe for the pinned-host run

bdchatham
and
claude
committed
fix(tools/gasbench): nolint the third guarded G115 site CI's gosec flags

bdchatham
and
claude
committed
fix(tools/gasbench): failed subtest emits a status=error row, not NaN

bdchatham
and
claude
committed
fix(tools/gasbench): taint self-check failures; run.sh core-range fallback; README enum

bdchatham
and
claude
committed
docs(tools/gasbench): -benchtime=1x does not neutralize -count

bdchatham
and
claude
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

tools/gasbench: per-opcode EVM gas-vs-execution-time differential microbenchmark (F2 MVP) - #3761

tools/gasbench: per-opcode EVM gas-vs-execution-time differential microbenchmark (F2 MVP)#3761
bdchatham wants to merge 20 commits into
mainfrom
brandon2/gasbench-mvp

Commits on Jul 14, 2026

feat(tools/gasbench): compute-opcode gas-vs-execution-time microbench MVP

fix(tools/gasbench): resolve xreview R1 dissent — significance gate, persisted diff, shift-op operands

docs(tools/gasbench): fix run.sh comment accuracy per xreview R2 (idiomatic-reviewer)

refactor(tools/gasbench): re-prioritize acceptance gate per noise-isolation research

Commits on Jul 15, 2026

refactor(tools/gasbench): close xreview R1 findings, move rationale to docs

fix(tools/gasbench): correct the doc-migration's own factual error

fix(tools/gasbench): never report Significant from an unestimated Uncertainty

test(tools/gasbench): close xreview R4 advisory items

docs(tools/gasbench): add operator Quickstart runbook to README

docs(tools/gasbench): correct -count file semantics in runbook, validated E2E

fix(tools/gasbench): distinct operands per case; emit nominal const_gas

feat(tools/gasbench): cross-run statistics on x/perf/benchmath; effect-size floor

test+docs(tools/gasbench): peer-sweep follow-ups

Merge branch 'main' into brandon2/gasbench-mvp

fix(tools/gasbench): address PR review findings (NaN bounds, lint gate)

docs(tools/gasbench): EC2 recipe for the pinned-host run

fix(tools/gasbench): nolint the third guarded G115 site CI's gosec flags

fix(tools/gasbench): failed subtest emits a status=error row, not NaN

fix(tools/gasbench): taint self-check failures; run.sh core-range fallback; README enum

docs(tools/gasbench): -benchtime=1x does not neutralize -count