-
Notifications
You must be signed in to change notification settings - Fork 608
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Switch more ci.yml jobs to optional based on paths changed. #17923
Conversation
# Common code likely enough to affect code paths unique to AMDGPU: | ||
"compiler/src/iree/compiler/GlobalOptimization/*", | ||
] | ||
|
||
# Jobs to run in presumbit if files under the corresponding path see changes. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I also want to change
build_test_runtime :: arm64
fromci.yml
andRegression Test / test_onnx :: amdgpu_vulkan
frompkgci_regression_test.yml
, but those jobs are under matrices that can't useif: contains(fromJson(needs.setup.outputs.enabled-jobs), 'job_name')At least the pkgci jobs are not blocking PR merges right now.
Definitely want build_test_runtime :: arm64
excluded too. That competes with build_test_all_arm64
for a single runner when LLVM integrate PRs are running 1h+ builds like https://github.com/iree-org/iree/actions/runs/9960816403/job/27520932492.
Strategies: https://stackoverflow.com/questions/65384420/how-do-i-make-a-github-action-matrix-element-conditional
0f3c715
to
5dd6eb9
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks.
Also need to look into the multiple runners off one machine because these queue times are getting longg 😞 |
Yep. These are the main levers we have for improving CI time:
Benefits from each of those will multiply as time savings. 2 and 3 will reduce $$$ costs while 1 will generally increase $$$ costs. 3 won't help if the runners themselves are unstable or go offline. |
…#17923) The `test_amd_mi250` and `test_amd_mi300` jobs in particular have been stuck with long queues and should not block merging core project PRs. This changes `ci.yml` to only run these jobs if files under code paths most likely to affect them are changed. The jobs can still be opted into via `ci-extra: job_name` git trailers in PR descriptions (see https://iree.dev/developers/general/contributing/#ci-behavior-manipulation). I also want to change `build_test_runtime :: arm64` from `ci.yml` and `Regression Test / test_onnx :: amdgpu_vulkan` from `pkgci_regression_test.yml`, but those jobs are under matrices that can't use ```yml if: contains(fromJson(needs.setup.outputs.enabled-jobs), 'job_name') ``` At least the pkgci jobs are not blocking PR merges right now. Signed-off-by: Lubo Litchev <lubol@google.com>
The
test_amd_mi250
andtest_amd_mi300
jobs in particular have been stuck with long queues and should not block merging core project PRs. This changesci.yml
to only run these jobs if files under code paths most likely to affect them are changed. The jobs can still be opted into viaci-extra: job_name
git trailers in PR descriptions (see https://iree.dev/developers/general/contributing/#ci-behavior-manipulation).I also want to change
build_test_runtime :: arm64
fromci.yml
andRegression Test / test_onnx :: amdgpu_vulkan
frompkgci_regression_test.yml
, but those jobs are under matrices that can't useAt least the pkgci jobs are not blocking PR merges right now.