Repository navigation
genhttp-ioxide: build GenHTTP from the branch of Kaliumhexacyanoferrat/GenHTTP#953 - #1783
Merged
Merged
Conversation
…t/GenHTTP#953 GenHTTP is cloned at e3974794, the head of #953, and referenced as projects rather than the 11.3.0 packages; the commit is pinned because the bench box builds on cached layers. ioxide.pg and ioxide.timer move to 0.26.563 with the branch's ioxide. The echo handler closes a rejected frame with FrameError.CloseStatus, and frames of up to 64 KiB are accepted (MaxFrameSize).
Owner
Author
|
/benchmark --save |
Contributor
|
👋 Benchmark request received. A collaborator will review and approve the run. |
Contributor
Benchmark ResultsFramework:
Composite score, main against this run, in the board's default view:
The results on main were measured in another session, so a few points up or down can be the machine and not the change. Compliance
H/1.1: 2 failed
H/2: 38 failed
H/3: 44 failed
|
This was referenced Nov 16, 2026
MDA2AV
added a commit
that referenced
this pull request
Nov 16, 2026
…s job log) [skip ci] #1783's run (job 132532753667, 2026-11-14) measured f852765, but its results commit was refused because 533ae33 reached the branch mid-run, so main kept genhttp-ioxide's results from #1749. The 32 result rows are rebuilt from that job's log with the repo's own <tool>_parse on every run's raw output, the winner picked as benchmark.sh picks it, and rows written as save_result writes them, then folded in with rebuild_site_data.merge_results. Every row matches the run's Best line and the results comment exactly. The one approximation: on the four fixed-rate profiles cpu_usec is recovered from the logged core-seconds and us/req, within +-0.003%. Compliance: H/2 (108/146) and WebSocket (244/244) are rebuilt from the results comment; H/1.1 and H/3 measured the same as main already had. The rebuilt file renders back to the posted comment byte for byte. Not recoverable from the log: the h2c listener's own failure list, WebSocket's non-strict count, and the run's container logs (site/static/logs keeps #1749's).
MDA2AV
added a commit
that referenced
this pull request
Nov 16, 2026
… be pushed (#1816) A rejected push used to print 'could not push to fork (permissions)' and leave the step green, so the results were lost without a word. #1783 lost its saved run that way: a commit reached the branch while the run was measuring, and the results push was refused as a non-fast-forward. Now the step fails, and comments on the PR with the reason: the branch moved (the numbers are for code it no longer has, so they are not rebased on top either), or the push was refused (a fork without maintainer edits).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
genhttp-ioxide now builds GenHTTP from source: the branch of Kaliumhexacyanoferrat/GenHTTP#953 (
update-ioxide-0.20.396) instead of the 11.3.0 release.What changes
28b0cd10, the head of [WTX] WebSocket evaluation #953, into/genhttp. It is pinned to that commit because the bench box builds on cached layers: a clone by branch name would keep whichever commit it fetched first. A newer commit means a newGENHTTP_COMMIT.GenHttpSource,/genhttpby default). The branch's engine pins ioxide 0.26.563, soioxide.pgandioxide.timermove from 0.15.251 to 0.26.563 with it. The entry's code builds against them unchanged.FrameError.CloseStatus, which the branch adds: 1007 for invalid UTF-8, 1009 for a frame that is too large, and 1002 for the other protocol errors.MaxFrameSize(64 * 1024), so the entry accepts frames of up to 64 KiB instead of the 16 KiB default.What #953 brings
Besides ioxide 0.26.563, the branch adds the websocket protocol checks Autobahn found missing: reserved bits, close frames, UTF-8, and the order of continuation frames.
TryReadFramealso no longer splits a fragmented message.Local results (
scripts/spec.sh)All four suites were measured on this PR's final build. Without
MaxFrameSize(64 * 1024), the six WebSocket cases with 64 KiB frames fail (238/244).validate.shpasses all 122 checks locally.A chunked body fix on the way
The first version of this PR pinned
e3974794, on whichvalidate.shfailed one check:POST /echo chunked 102400Bover TLS (:8081) came back short. GenHTTP's chunked body reader read a chunk's length after advancing past it, when the reader may already have reset the segments that length is computed from. ioxide's TLS reader does, so the last read of the body reported more bytes than its buffer held. It is fixed on #953 by28b0cd10, with a regression test, and this PR now pins that commit.🤖 Generated with Claude Code