René's URL Explorer Experiment
Title: fix(bench): re-measure a trace report that differs only in the measured sets by Unisay · Pull Request #355 · purescript-lua/purescript-lua · GitHub
Open Graph Title: fix(bench): re-measure a trace report that differs only in the measured sets by Unisay · Pull Request #355 · purescript-lua/purescript-lua
X Title: fix(bench): re-measure a trace report that differs only in the measured sets by Unisay · Pull Request #355 · purescript-lua/purescript-lua
Description: Closes #346.
./bench/ci is the repository's benchmark oracle: it links the Bench.* PureScript modules to standalone Lua, meters what LuaJIT does with them, and diffs the result against reports committed under bench/goldens/, so a codegen change that adds a closure allocation or stops a loop from being compiled shows up as a reviewable diff. Two of the three meters are static walks over the artifact and byte-stable by construction. The third, the per-spec trace report, records which source locations ended up carrying a compiled trace (LuaJIT rewrites the opcode to a J* form) or a blacklist (an I* form) — and whether a given location does is a per-process dice roll, because LuaJIT's hot counters live in a small hashed table keyed by bytecode address, so removing one hoisted table from an artifact reshuffles which counters collide. bench/tools/trace_report.lua already absorbs that with a majority vote over nine fresh LuaJIT processes, which pins any spot whose per-process formation probability p sits near 0 or 1 and leaves a spot near one half a coin flip.
The failure is a class, not the one spot
Bench.RefLoop.lua:32 JFUNCF — the entry of the ST thunk wrapping a hot loop — is the entry the issue names, having become marginal when an unrelated codegen change shifted a line. Measuring every spec's entries over 30 raw single-process trials (trace_report.lua --trial prints one unvoted measurement) shows it is one of fourteen: fourteen entries across 8 of the 12 specs land strictly between "never" and "always".
spec raw/30 golden entry
ref_loop 21/30 present Bench.RefLoop.lua:32 JFUNCF
ref_loop 8/30 absent Bench.RefLoop.lua:35 -- inner loop in root trace
record_set 4/30 absent Bench.RecordSet.lua:6 -- inner loop in root trace
ctor_build 4/30 absent Bench.CtorBuild.lua:6 -- inner loop in root trace
ctor_build 26/30 present Bench.CtorBuild.lua:3 JFUNCF
record_set 26/30 present Bench.RecordSet.lua:3 JFUNCF
record_set 26/30 present Bench.RecordSet.lua:29 -- NYI: bytecode FNEW
array_foldl 27/30 present Bench.ArrayFoldl.lua:20 -- NYI: bytecode FNEW
Two facts in that table decide the design. Both measured signals are affected, not just opcode end states: -- inner loop in root trace and -- NYI: bytecode FNEW are abort-site entries, which live in the report's other section. And the two ref_loop entries are complementary (21 + 8 ≈ 30) — the abort is what gets recorded when the entry trace does not form — so this is one phenomenon surfacing in two places, not two independent problems.
What changes
The comparison moves out of bench/ci into bench/tools/diff_counters, which classifies each mismatch before deciding what it costs. A trace report differing only in its two measured sets (the indented entries and the counts: line tallying them) is re-measured — a fresh nine-process vote — and fails only if every attempt still differs; two re-measurements is the default, BENCH_TRACE_RETRIES overrides it, and 0 restores a plain diff. Everything else is deterministic by construction and fails at once, buying no fresh processes: a census difference, a difference in a trace report's header lines (spec:, runtime:, workload: — derived from the spec plus a deterministic checksum), or a report existing on one side only.
That second half is not just tidiness. A LuaJIT bump changes the runtime: line of all twelve reports, and re-measuring twelve reports twice only to rediscover a version string would spend about a minute of fresh processes reaching the same verdict:
trace_ref_loop.txt: differs from the golden outside the measured sets
--- bench/_build/demo/goldens/trace_ref_loop.txt 2026-07-29 16:09:00.122835133 +0200
+++ bench/_build/demo/out/trace_ref_loop.txt 2026-07-29 16:09:00.119835126 +0200
@@ -1,5 +1,5 @@
spec: ref_loop
-runtime: LuaJIT 2.1.9999999999
+runtime: LuaJIT 2.1.1741730670
workload: n=100000 reps=4 result=4999950000
aborts (distinct site -- reason):
Bench.RefLoop.lua:40 -- NYI: bytecode FNEW
Evidence
Replaying the comparison for ref_loop against its committed golden 300 times, each iteration generating a fresh canonical report and then asking both configurations about it:
retries=0 over 300 runs: red=9, absorbed-by-re-measurement=0
retries=2 over 300 runs: red=0, absorbed-by-re-measurement=10
Those nine reds are the one entry the issue names, measured in isolation. Repeating the exercise over the whole twelve-spec pass, so every marginal entry is in play at once, puts the step-level rate in the same place and shows which entry dominates it:
iteration 23 would redden today: trace_ref_loop.txt
40 iterations of the full 12-spec trace pass:
reddens with a plain diff (today): 1
reddens with 2 re-measurements: 0
absorbed by a re-measurement: 1
The other thirteen marginal entries sit far enough from one half that a nine-trial vote already pins them (each below 0.4% on its own), which is why ref_loop is the only spec that shows up; they matter because the next line-shifting codegen change can move any of them, not because they redden today.
A reproducible change to which bytecodes trace still fails, having spent the whole budget. Standing in for a regression that blacklists a hot loop instead of compiling it, JFORL → IFORL in the golden:
trace_ref_loop.txt: differs in the measured sets; re-measuring (1 of 2)
trace_ref_loop.txt: differs in the measured sets; re-measuring (2 of 2)
trace_ref_loop.txt: still differs after 2 re-measurement(s)
--- bench/_build/demo/goldens/trace_ref_loop.txt 2026-07-29 16:09:00.191835295 +0200
+++ bench/_build/demo/out/trace_ref_loop.txt 2026-07-29 16:09:00.310835574 +0200
@@ -7,5 +7,5 @@
Bench.RefLoop.lua:10 JFUNCF
Bench.RefLoop.lua:32 JFUNCF
Bench.RefLoop.lua:34 JFORI
- Bench.RefLoop.lua:34 IFORL
+ Bench.RefLoop.lua:34 JFORL
counts: aborts=1 compiled=4 blacklisted=0
An absorbed flip is green but never silent — the entries that flickered are printed, since a spot resolving in some processes and not others is the one thing a reviewer cannot see in the committed goldens. This is not a staged mismatch: generating canonical reports against the real committed golden until one genuinely differed took 36 attempts, matching the measured rate, and the 36th is what the oracle then saw.
trace_ref_loop.txt: differs in the measured sets; re-measuring (1 of 2)
trace_ref_loop.txt: matched on re-measurement 1 -- the entries below are trace-formation noise, not a codegen change
--- bench/_build/demo/goldens/trace_ref_loop.txt 2026-07-29 16:09:15.689871641 +0200
+++ bench/_build/demo/out/trace_ref_loop.txt 2026-07-29 16:09:17.405875668 +0200
@@ -5,7 +5,6 @@
Bench.RefLoop.lua:40 -- NYI: bytecode FNEW
bytecode end state (J*=compiled, I*=blacklisted):
Bench.RefLoop.lua:10 JFUNCF
- Bench.RefLoop.lua:32 JFUNCF
Bench.RefLoop.lua:34 JFORI
Bench.RefLoop.lua:34 JFORL
-counts: aborts=1 compiled=4 blacklisted=0
+counts: aborts=1 compiled=3 blacklisted=0
bench counters match goldens
Why re-measuring, and not the alternatives
Re-measurements are conjunctive — every attempt must differ for the oracle to fail — which is what makes them a better use of fresh processes than a longer vote. For one entry at probability p, a nine-trial vote lands on the wrong side with P(Binom(9,p) ≤ 4), and each re-measurement multiplies that in:
p one vote +1 re-measurement +2 (the default) 25 trials, one vote
0.60 26.7% 7.1% 1.9% 15.4%
0.70 9.9% 1.0% 0.1% 1.8%
0.80 2.0% 0.04% 0.001% 0.04%
The independence is genuine within a run: p is a property of the artifact plus the process layout, each trial is a fresh process, so the budget resamples the layout while the artifact bytes and the LuaJIT binary stay fixed. And re-measurements are free on a green run, where a longer vote costs its extra processes every time.
The measured table above is what rules out the issue's other three options. Quarantining named spots would exempt fourteen entries across 8 of 12 specs — most of what the oracle pins — and the list needs re-deriving after every codegen change that shifts lines, which is the very event that creates marginal spots. Dropping the function-entry signal class (J*FUNCF) leaves the abort-site entries untouched, and those are half the table, so it would give up real signal without removing the failure mode. Classifying a middle band as a third state needs a band edge, and the measured probabilities are spread across the whole range (4, 8, 21, 26, 27, 28, 29 out of 30), so some spot sits on whatever edge is chosen.
What re-measuring gives up is a regression that itself lands near p = 0.5. Such a regression has no pinnable golden on either side even with a single vote, so nothing previously catchable becomes uncatchable.
Guard
bench/tools/counters_selftest pins the verdict logic against synthetic reports, so it runs in milliseconds instead of spawning the hundred-odd LuaJIT processes a real set of reports needs — and, more to the point, it can stage a marginal spot deterministically, which the real tool cannot. Each case pins both the verdict and its cost, because a retry budget spent on something deterministic is as much a defect as a wrong verdict. Written first, it was red on exactly the cases about re-measurement when run against a stand-in for today's plain diff -ru:
ok matching reports pass without re-measuring (exit 0, 0 regenerations)
FAIL marginal trace entry absorbed by a re-measurement
exit 1 (want 0), regenerations 0 (want 1)
ok marginal trace entry reddens with the retry budget at zero (exit 1, 0 regenerations)
FAIL reproducible trace change fails after spending the budget
exit 1 (want 1), regenerations 0 (want 2)
ok census difference fails without re-measuring (exit 1, 0 regenerations)
ok trace header change fails without re-measuring (exit 1, 0 regenerations)
ok report missing from the output fails without re-measuring (exit 1, 0 regenerations)
ok report absent from the goldens fails without re-measuring (exit 1, 0 regenerations)
FAIL a failed re-measurement aborts instead of spending the budget
exit 1 (want 1), regenerations 0 (want 1)
3 check(s) failed
and green after:
ok matching reports pass without re-measuring (exit 0, 0 regenerations)
ok marginal trace entry absorbed by a re-measurement (exit 0, 1 regenerations)
ok marginal trace entry reddens with the retry budget at zero (exit 1, 0 regenerations)
ok reproducible trace change fails after spending the budget (exit 1, 2 regenerations)
ok census difference fails without re-measuring (exit 1, 0 regenerations)
ok trace header change fails without re-measuring (exit 1, 0 regenerations)
ok report missing from the output fails without re-measuring (exit 1, 0 regenerations)
ok report absent from the goldens fails without re-measuring (exit 1, 0 regenerations)
ok a failed re-measurement aborts instead of spending the budget (exit 1, 1 regenerations)
bench counter verdict logic behaves as pinned
Two of those cases pin decisions rather than mechanics. marginal trace entry reddens with the retry budget at zero makes the differential permanent — BENCH_TRACE_RETRIES=0 is the pre-change behaviour, so deleting the re-measurement reddens the guard instead of only showing up in this branch's history. And a failed re-measurement aborts instead of spending the budget pins the regen contract: trace_report.lua exits non-zero when a trial process fails or when the deterministic part of a report diverges between its own trials, and both mean the measurement did not happen, so burning the remaining attempts would end in "still differs after 2 re-measurements" — blaming the counters for what the measuring tool said about itself. That was raised by the automated review as a bug (a transient regen failure not being treated as a retry attempt); the behaviour is deliberate and unchanged, now documented at the regen contract in the tool header and pinned by that case.
CI runs the self-test as its own step ahead of the counters, so a broken verdict reads as itself instead of as a mystery counter diff.
Sibling audit
The shape searched for was "a committed oracle compared by exact equality over a measurement that is not a pure function of the artifact". Inside bench/, that search is what produced the fourteen-entry table: the class is handled uniformly here rather than one spec at a time, and the two censuses are the deliberate contrast — they never run the code, and ./bench/ci still generates each twice and compares byte for byte.
Outside bench/, the eval goldens (test/ps/**/eval/golden.txt) do run generated Lua and compare its printed output exactly, so they would have the same exposure if anything reachable iterated a table in unspecified order. Two such sites exist — PSLUA_object_update in lib/Language/PureScript/Backend/Lua/Fixture.hs and unsafeSet/unsafeDelete in the prelude fork's Record/Unsafe.lua — and both use pairs only to build an identically-keyed copy, so iteration order never reaches printed output. The differential spec compares against luac -p and lua themselves, which are deterministic. Nothing to fix, and nothing filed.
Notes
Compiler code is untouched, so the committed goldens do not move and cabal test all is unaffected (1236 examples, 0 failures). nix fmt reports no changes.
Open Graph Description: Closes #346.
./bench/ci is the repository's benchmark oracle: it links the Bench.* PureScript modules to standalone Lua, meters what LuaJIT does with them, and diffs the result against reports ...
X Description: Closes #346.
./bench/ci is the repository's benchmark oracle: it links the Bench.* PureScript modules to standalone Lua, meters what LuaJIT does with them, and diffs the result against repo...
Opengraph URL: https://github.com/purescript-lua/purescript-lua/pull/355
X: @github
direct link
Domain: github.com
| react-profiling | 0 |
| react-import-map | react,react-dom,react-dom/client,react-dom/profiling,react-is,react/compiler-runtime,react/jsx-dev-runtime,react/jsx-runtime,scheduler@b8488ee2c322 |
| route-pattern | /:user_id/:repository/pull/:id/files(.:format) |
| route-controller | pull_requests |
| route-action | files |
| fetch-nonce | v2:19981968-2747-39af-2c03-e0076b8f3a27 |
| current-catalog-service-hash | ae870bc5e265a340912cde392f23dad3671a0a881730ffdadd82f2f57d81641b |
| request-id | D61E:29AEE8:745B14:9D271B:6B00DBB3 |
| html-safe-nonce | 39d7fea0a4f92ca7f662ca91654aaaec7a85cdaa8478c29a36b2b16508f38698 |
| visitor-payload | eyJyZWZlcnJlciI6IiIsInJlcXVlc3RfaWQiOiJENjFFOjI5QUVFODo3NDVCMTQ6OUQyNzFCOjZCMDBEQkIzIiwidmlzaXRvcl9pZCI6IjQ1MzE4NDc0NDg2ODI4Nzk5MjMiLCJyZWdpb25fZWRnZSI6ImlhZCIsInJlZ2lvbl9yZW5kZXIiOiJpYWQifQ== |
| visitor-hmac | d1fc3b2078e92beb0a023ae6004f3dd678a235f30c1cf13ec805c99558253462 |
| hovercard-subject-tag | pull_request:4161276278 |
| github-keyboard-shortcuts | repository,pull-request-list,pull-request-conversation,pull-request-files-changed,copilot |
| google-site-verification | Apib7-x98H0j5cPqHWwSMm6dNU4GmODRoqxLiDzdx9I |
| octolytics-url | https://collector.github.com/github/collect |
| analytics-location | ///pull_requests/show/files |
| fb:app_id | 1401488693436528 |
| apple-itunes-app | app-id=1477376905, app-argument=https://github.com/purescript-lua/purescript-lua/pull/355/files |
| twitter:image | https://avatars.githubusercontent.com/u/1009751?s=400&v=4 |
| twitter:card | summary_large_image |
| og:image | https://avatars.githubusercontent.com/u/1009751?s=400&v=4 |
| og:image:alt | Closes #346.
./bench/ci is the repository's benchmark oracle: it links the Bench.* PureScript modules to standalone Lua, meters what LuaJIT does with them, and diffs the result against reports ... |
| og:site_name | GitHub |
| og:type | object |
| hostname | github.com |
| expected-hostname | github.com |
| None | 043c3f30fbcbe2c6c8e729e10cbbbc90c406546f9fbbb67c648d38f6b7d16daf |
| turbo-cache-control | no-preview |
| diff-view | unified |
| go-import | github.com/purescript-lua/purescript-lua git https://github.com/purescript-lua/purescript-lua.git |
| octolytics-dimension-user_id | 131581790 |
| octolytics-dimension-user_login | purescript-lua |
| octolytics-dimension-repository_id | 587200487 |
| octolytics-dimension-repository_nwo | purescript-lua/purescript-lua |
| octolytics-dimension-repository_public | true |
| octolytics-dimension-repository_is_fork | false |
| octolytics-dimension-repository_network_root_id | 587200487 |
| octolytics-dimension-repository_network_root_nwo | purescript-lua/purescript-lua |
| turbo-body-classes | logged-out env-production page-responsive full-width |
| disable-turbo | true |
| browser-stats-url | https://api.github.com/_private/browser/stats |
| browser-errors-url | https://api.github.com/_private/browser/errors |
| release | 5425dc24e8bd8d6fca9641a12afc984825a43875 |
| ui-target | full |
| theme-color | #1e2327 |
| color-scheme | light dark |
Links:
| Skip to content | https://github.com/purescript-lua/purescript-lua/pull/355/files#start-of-content |
| https://github.com/ |
| Sign in | https://github.com/login?return_to=https%3A%2F%2Fgithub.com%2Fpurescript-lua%2Fpurescript-lua%2Fpull%2F355%2Ffiles |
| GitHubRepository hosting, code review, and collaboration | https://github.com/platform |
| CopilotAI coding agent for tasks, projects, and automated work | https://github.com/features/copilot |
| ActionsCI/CD and automation | https://github.com/features/actions |
| Code QualityQuality analysis, fixes, and enforcement | https://github.com/features/code-quality |
| CodespacesCloud development environments | https://github.com/features/codespaces |
| Code SecurityVulnerability detection and remediation | https://github.com/security/advanced-security/code-security |
| Secret ProtectionSecret scanning and leak prevention | https://github.com/security/advanced-security/secret-protection |
| All Features | https://github.com/features |
| Documentation | https://docs.github.com/en |
| Training | https://learn.github.com/skills |
| Certifications | https://learn.github.com/credentials |
| Build with Copilot | https://github.com/features/ai |
| Customize Copilot | https://docs.github.com/en/copilot/how-tos/copilot-on-github/customize-copilot/customize-copilot-overview |
| Integrations | https://github.com/marketplace |
| Trending repos | https://github.com/trending |
| GitHub Sponsors | https://github.com/open-source/sponsors |
| Developer articles | https://github.com/resources/articles |
| ChangelogYour source for the latest features, improvements, and fixes across GitHub products, all in one place. | https://github.blog/changelog/ |
| Enterprises | https://github.com/enterprise |
| Team | https://github.com/team |
| Startups | https://github.com/enterprise/startups |
| Nonprofits | https://github.com/solutions/industry/nonprofits |
| Healthcare | https://github.com/solutions/industry/healthcare |
| Financial services | https://github.com/solutions/industry/financial-services |
| Manufacturing | https://github.com/solutions/industry/manufacturing |
| Government | https://github.com/solutions/industry/government |
| See all industries | https://github.com/solutions/industry |
| Why GitHub | https://github.com/why-github |
| Customer Stories | https://github.com/customer-stories |
| Business insights | https://github.com/solutions/executive-insights |
| Trust Center | https://github.com/trust-center |
| GitHub Advanced SecurityEnterprise-grade code security | https://github.com/security/advanced-security |
| Copilot for businessEnterprise-grade AI features | https://github.com/features/copilot/copilot-business |
| Premium SupportEnterprise-grade 24/7 support | https://github.com/enterprise/premium-support |
| Documentation | https://docs.github.com/en |
| Resource center | https://github.com/resources |
| Blog | https://github.blog/ |
| Events & webinars | https://github.com/resources/events |
| Community Forum | https://github.com/orgs/community/discussions |
| Customer Stories | https://github.com/customer-stories |
| Integrations | https://github.com/marketplace |
| Customer support | https://help.github.com/ |
| Training | https://learn.github.com/skills |
| GitHub Status | https://www.githubstatus.com/ |
| About | https://github.com/about |
| Partners | https://github.com/partners |
| Changelog | https://github.blog/changelog/ |
| GitHub UniverseThe event for developers, by developers — talks, workshops, and what’s next for GitHub. | https://githubuniverse.com/ |
| Pricing | https://github.com/pricing |
| Sign in | https://github.com/login?return_to=https%3A%2F%2Fgithub.com%2Fpurescript-lua%2Fpurescript-lua%2Fpull%2F355%2Ffiles |
| Sign up | https://github.com/signup?ref_cta=Sign+up&ref_loc=header+logged+out&ref_page=%2F%3Cuser-name%3E%2F%3Crepo-name%3E%2Fpull_requests%2Fshow%2Ffiles&source=header-repo&source_repo=purescript-lua%2Fpurescript-lua |
| Reload | https://github.com/purescript-lua/purescript-lua/pull/355/files |
| Reload | https://github.com/purescript-lua/purescript-lua/pull/355/files |
| Reload | https://github.com/purescript-lua/purescript-lua/pull/355/files |
| purescript-lua | https://github.com/purescript-lua |
| purescript-lua | https://github.com/purescript-lua/purescript-lua |
| Code | https://github.com/purescript-lua/purescript-lua |
| Issues48 (48) | https://github.com/purescript-lua/purescript-lua/issues |
| Pull requests | https://github.com/purescript-lua/purescript-lua/pulls |
| Discussions | https://github.com/purescript-lua/purescript-lua/discussions |
| Actions | https://github.com/purescript-lua/purescript-lua/actions |
| Projects | https://github.com/purescript-lua/purescript-lua/projects |
| Security and quality | https://github.com/purescript-lua/purescript-lua/security |
| Insights | https://github.com/purescript-lua/purescript-lua/insights |
| Sign up for GitHub
| https://github.com/signup?return_to=%2Fpurescript-lua%2Fpurescript-lua%2Fissues%2Fnew%2Fchoose |
| terms of service | https://docs.github.com/terms |
| privacy statement | https://docs.github.com/privacy |
| Sign in | https://github.com/login?return_to=%2Fpurescript-lua%2Fpurescript-lua%2Fissues%2Fnew%2Fchoose |
| Unisay | https://github.com/Unisay |
| main | https://github.com/purescript-lua/purescript-lua/tree/main |
| issue-346/bench-trace-oracle-retry | https://github.com/purescript-lua/purescript-lua/tree/issue-346/bench-trace-oracle-retry |
|
Conversation
0
| https://github.com/purescript-lua/purescript-lua/pull/355 |
|
Commits
2
| https://github.com/purescript-lua/purescript-lua/pull/355/commits |
|
Checks
2
| https://github.com/purescript-lua/purescript-lua/pull/355/checks |
|
Files changed
| https://github.com/purescript-lua/purescript-lua/pull/355/files |
| Please reload this page | https://github.com/purescript-lua/purescript-lua/pull/355/files |
|
fix(bench): re-measure a trace report that differs only in the measured sets
| https://github.com/purescript-lua/purescript-lua/pull/355/files#top |
|
Show all changes
2 commits
| https://github.com/purescript-lua/purescript-lua/pull/355/files |
|
a4adfff
fix(bench): re-measure a trace report that differs only in the measur…
Unisay Jul 29, 2026
| https://github.com/purescript-lua/purescript-lua/pull/355/commits/a4adfffc8a6bc30f3eae5052ca91d0db6f216372 |
|
1865e7e
test(bench): pin that a failed re-measurement aborts rather than spen…
Unisay Jul 29, 2026
| https://github.com/purescript-lua/purescript-lua/pull/355/commits/1865e7ea7a785d285affabcd9c9af9a1f0af7504 |
|
Clear filters
| https://github.com/purescript-lua/purescript-lua/pull/355/files |
| Please reload this page | https://github.com/purescript-lua/purescript-lua/pull/355/files |
| Please reload this page | https://github.com/purescript-lua/purescript-lua/pull/355/files |
|
ci.yaml
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-944291df2c9c06359d37cc8833d182d705c9e8c3108e7cfe132d61a06e9133dd |
|
README.md
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
|
ci
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8419eee1239e6d308c2b77a01f4b002fffd650ae1e245b0bf8678caacde67cbf |
|
counters_selftest
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-f8317b8036304c8ff3a7d4f77081d4583934c59413e0fcd86c31c4bfb5da06f5 |
|
diff_counters
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8d1c00294e40e6cddc59ca91aab5a5c53f9444ffb1456754e4652fbc06291c1e |
|
trace_report.lua
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-2a08844aaaaaef7bcc1a841e4023aee323d9293832908f951aa2e9d7bac3c0e3 |
|
20260729_190000_unisay_bench_trace_oracle_remeasure.md
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-84cfb0d5c4a97429d3b86fc8f9e099564617c48a9a5589f1f19a109a590d3297 |
| .github/workflows/ci.yaml | https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-944291df2c9c06359d37cc8833d182d705c9e8c3108e7cfe132d61a06e9133dd |
|
View file
| https://github.com/purescript-lua/purescript-lua/blob/1865e7ea7a785d285affabcd9c9af9a1f0af7504/.github/workflows/ci.yaml |
|
Open in desktop
| https://desktop.github.com |
| https://github.co/hiddenchars |
| https://github.com/purescript-lua/purescript-lua/pull/355/{{ revealButtonHref }} |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-944291df2c9c06359d37cc8833d182d705c9e8c3108e7cfe132d61a06e9133dd |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-944291df2c9c06359d37cc8833d182d705c9e8c3108e7cfe132d61a06e9133dd |
| bench/README.md | https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
|
View file
| https://github.com/purescript-lua/purescript-lua/blob/1865e7ea7a785d285affabcd9c9af9a1f0af7504/bench/README.md |
|
Open in desktop
| https://desktop.github.com |
| https://github.co/hiddenchars |
| https://github.com/purescript-lua/purescript-lua/pull/355/{{ revealButtonHref }} |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-46ff119234b9456477b06dc1699710eebe72f9d9bfa2f7edf5f63707f292c623 |
| bench/ci | https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8419eee1239e6d308c2b77a01f4b002fffd650ae1e245b0bf8678caacde67cbf |
|
View file
| https://github.com/purescript-lua/purescript-lua/blob/1865e7ea7a785d285affabcd9c9af9a1f0af7504/bench/ci |
|
Open in desktop
| https://desktop.github.com |
| https://github.co/hiddenchars |
| https://github.com/purescript-lua/purescript-lua/pull/355/{{ revealButtonHref }} |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8419eee1239e6d308c2b77a01f4b002fffd650ae1e245b0bf8678caacde67cbf |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8419eee1239e6d308c2b77a01f4b002fffd650ae1e245b0bf8678caacde67cbf |
|
| https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-8419eee1239e6d308c2b77a01f4b002fffd650ae1e245b0bf8678caacde67cbf |
| bench/tools/counters_selftest | https://github.com/purescript-lua/purescript-lua/pull/355/files#diff-f8317b8036304c8ff3a7d4f77081d4583934c59413e0fcd86c31c4bfb5da06f5 |
|
View file
| https://github.com/purescript-lua/purescript-lua/blob/1865e7ea7a785d285affabcd9c9af9a1f0af7504/bench/tools/counters_selftest |
|
Open in desktop
| https://desktop.github.com |
| https://github.co/hiddenchars |
| https://github.com/purescript-lua/purescript-lua/pull/355/{{ revealButtonHref }} |
| Please reload this page | https://github.com/purescript-lua/purescript-lua/pull/355/files |
|
| https://github.com |
| Terms | https://docs.github.com/site-policy/github-terms/github-terms-of-service |
| Privacy | https://docs.github.com/site-policy/privacy-policies/github-privacy-statement |
| Security | https://github.com/security |
| Status | https://www.githubstatus.com/ |
| Community | https://github.community/ |
| Docs | https://docs.github.com/ |
| Contact | https://support.github.com?tags=dotcom-footer |
Viewport: width=device-width
URLs of crawlers that visited me.