Skip to content

[code-documentation] Correct benchmark gate thresholds - #400

Merged
Pedro Henrique Penna (ppenna) merged 1 commit into
devfrom
code-documentation/benchmark-gate-thresholds-37382439709-5993f30a435046a1
Oct 7, 2026
Merged

Pedro Henrique Penna (ppenna) merged 1 commit into
devfrom
code-documentation/benchmark-gate-thresholds-37382439709-5993f30a435046a1

Conversation

@ppenna

Copy link
Copy Markdown
Contributor

Summary

Correct the stale performance-gate thresholds in doc/benchmarks.md:

A metric regresses only when it is more than 50% worse. Lower-is-better millisecond metrics must also be more than 10 ms slower.

The current source of truth in scripts/nvx_tools/performance.py:2002-2011 sets the defaults to a 40% threshold and a 5 ms absolute tolerance. python3 scripts/nvx.py performance gate --help confirms the command surface. The existing doc/usage.md section also documents the current 40% and 5 ms values.

This is a single documentation mismatch. I reviewed recent documentation workflow outcomes and active pull-request file lists; no active or prior proposal covers these benchmark-gate threshold statements. The checked paths are doc/benchmarks.md, scripts/nvx_tools/performance.py, doc/usage.md, and the performance gate --help command.

Scope and validation

  • Changed files: doc/benchmarks.md
  • Total changed lines: 4 (2 additions, 2 deletions)
  • python3 scripts/nvx.py performance gate --help — passed
  • python3 -m unittest scripts/test_performance.py -q — passed (53 tests)
  • git diff --check — passed
  • git diff --numstat / git diff --raw — one allowed Markdown file; no gitlink change

No dependency, public API/CLI/ABI, gitlink, or OpenVMM change was made; no executable code was modified.

Generated by code-documentation · copilot · gpt56 · 4.93 AIC · ⌖ 0.71 AIC · ⊞ 17.6K · ◷

  • expires on Oct 19, 2026, 10:50 PM UTC

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Copilot AI balanced review requested due to automatic review settings October 5, 2026 22:50

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot review overview

🟢 Approval recommended

The documentation now accurately reflects the implemented defaults and regression logic.

Review effort: Balanced
Findings: None

What changed in this PR

Updates benchmark regression-gate documentation to match current CLI defaults and implementation.

Changes:

  • Corrects the relative regression threshold from 50% to 40%.
  • Corrects latency tolerance from 10 ms to 5 ms.
File Description
doc/​benchmarks.md Aligns documented gate thresholds with implementation and usage documentation.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

@ppenna
Pedro Henrique Penna (ppenna) marked this pull request as ready for review October 6, 2026 06:10
@ppenna
Pedro Henrique Penna (ppenna) merged commit 10ba817 into dev Oct 7, 2026
49 checks passed
@ppenna
Pedro Henrique Penna (ppenna) deleted the code-documentation/benchmark-gate-thresholds-37382439709-5993f30a435046a1 branch October 7, 2026 13:36
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants