-
Update and commit the changes to the benchmark.
-
Create a zip file of the directory, e.g:
$ cd perf-automation/benchmarks $ zip -r unity-webgl-`git rev-parse --short=12 HEAD`.zip unity-webgl -
Figure out the sha256 hash and size:
$ sha256sum unity-webgl-<hash>.zip $ du -b unity-webgl-<hash>.zipNote: On OSX instead of 'du' you can use:
stat -f "%z bytes" unity-webgl-<hash>.zip -
Upload the zip as a Github release (click "Releases" on the gihub page). The tag and title should be unity-webgl-v1 (replace v1 with the next version number).
-
Update the appropriate fetch task with the url, sha256 hash and size. Benchmark fetches are found here:
https://searchfox.org/mozilla-central/source/taskcluster/ci/fetch/benchmarks.yml -
Update the local revision for running with the mach command here:
https://searchfox.org/mozilla-central/source/testing/raptor/mach_commands.py
-
Pull the code from this repo: https://github.com/jandem/matrix-react-bench
-
Build it by following the instructions in the README.
-
Copy the repo folder to
benchmarks/in this repo. -
Delete the folders and files:
benchmarks/matrix-react-bench/node_modules/.cachebenchmarks/matrix-react-bench/.gitbenchmarks/matrix-react-bench/.git-ignore
-
Commit the new code to the master branch in this repo, and follow the instructions for creating a new benchmark release.