Automatic SASS diffs for PRs - #10780
Conversation
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
|
/ok to test |
🔬 CUB benchmark SASS comparisonThis check never blocks a merge. |
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
1 similar comment
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
1 similar comment
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
/ok to test |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
griwes
left a comment
There was a problem hiding this comment.
Reviewed throughout. Looks good. The normalization catches the usual causes of noise in the SASS diffs, from what I can tell. Tests appear appropriate.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
| "name": "cub-benchmark", | ||
| "displayName": "CUB benchmarking", | ||
| "generator": "Ninja", | ||
| "binaryDir": "${sourceDir}/build/$env{CCCL_BUILD_INFIX}/${presetName}", |
There was a problem hiding this comment.
important: there's a couple of places that you'd have to change before we can merge this. First, docs/cub/benchmarking.rst and docs/cub/tuning_infra.rst showcase building in build/, which is fine, but later illustrate build/bin/cub.bench* paths that do not include the preset directory. Besides that, benchmarks/scripts/submit_benchmark_job.sh builds in build_perf instead of build/, so can get broken after this is merged, please adjust. Most importantly, some (soon to be deleted) internal tuning scripts create a build directory per GPU, which will conflict with one executable path introduced here. Work with @gonidelis to make sure all tuning and benchmarking scripts work with the change.
There was a problem hiding this comment.
Docs and benchmarking scripts should be fixed now.
8c3b2ee to
5af0941
Compare
This comment has been minimized.
This comment has been minimized.
⏱️ CCCL compile-time benchmark comparison: Public headers compile-time benchResult: 0 regression row(s), 2 improvement row(s) above threshold.
Artifacts: reports and traces Direct file processing
🟢 Direct file processing — Improvements
|
🥳 CI Workflow Results🟩 Finished in 4h 28m: Pass: 100%/531 | Total: 8d 12h | Max: 2h 16m | Hits: 58%/1117894See results here. |
Description
closes #10726
Checklist