[CD] fix xpu support packages version #138189
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/138189
Note: Links to docs will display an error until the docs builds have been completed. ❌ 5 New FailuresAs of commit e54fc4e with merge base 9c084cc ( NEW FAILURES - The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
drisspg
added
the
triaged
| if [ -n "$XPU_VERSION" ]; then | ||
| apt-get install -y intel-for-pytorch-gpu-dev-${XPU_VERSION} intel-pti-dev | ||
| apt-get install -y intel-for-pytorch-gpu-dev-${XPU_VERSION} intel-pti-dev-0.9 | ||
| else | ||
| apt-get install -y intel-for-pytorch-gpu-dev intel-pti-dev | ||
| apt-get install -y intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9 |
There was a problem hiding this comment.
rhel and sles always install intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9. May I know why ubuntu needs to install different intel-for-pytorch-gpu-dev based on XPU_VERSION?
There was a problem hiding this comment.
The ubuntu used for CI, which extend the XPU_VERSION param for future upgrade
| if [[ "${XPU_DRIVER_TYPE,,}" == "rolling" ]]; then | ||
| apt-get install -y intel-ocloc | ||
| fi |
There was a problem hiding this comment.
Why does the other two OSes not require rolling check?
There was a problem hiding this comment.
Because the installation between LTS and latest rolling driver on ubuntu are different now. Previously, the ocloc package included in igc package on ubuntu both for LTS and rolling. But in the latest ubuntu rolling driver, the ocloc has been extracted as a standalone package. For others 2 OSes, the ocloc is always a standalone package.
|
@EikanWang can you please elaborate why this had 2.5.1 milestone? |
|
HI @chuanqi129 this does not change the packaging of the wheel, but changes to Docker file. How does it fixes #135867 ? |
Hi @atalman, thanks for the review and comment. Because we the new PTI version is the patch release, the max.min version and so name didn't change, so we don't need update packaging of the wheel part. But if bundle/PTI release a new version in the future, which change the max.min version, we will update the packaging part. For now, we can fix the version and upgrade the version by PR in case the interrupt. Does it make sense to you? We need it for pytorch 2.5.1 release, because when we build the docker image for release wheel 2.5.1, we need to apply those changes to make sure we install right version packages in the docker image. CC: @malfet @EikanWang |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Merge failedReason: 2 jobs have failed, first few of them are: linux-binary-manywheel / manywheel-py3_9-cuda12_4-build / build, linux-binary-libtorch-cxx11-abi / libtorch-cpu-shared-with-deps-cxx11-abi-build / build Details for Dev Infra teamRaised by workflow job |
|
@pytorchmergebot merge -f "failures are irrelevant" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
@pytorchbot cherry-pick --onto release/2.5 -c critical |
Works for #114850 Pull Request resolved: #138189 Approved by: https://github.com/EikanWang, https://github.com/malfet, https://github.com/atalman (cherry picked from commit 2f1842f)
Cherry picking #138189The cherry pick PR is at #138694 and it is recommended to link a critical cherry pick PR with an issue. The following tracker issues are updated: Details for Dev Infra teamRaised by workflow job |
@EikanWang why is it critical? |
It would only be needed if there are plan to rebuild docker images for 2.5 branch, but there isn't at the moment |
|
2.5.1 is an emergency patch release to address known large regressions, moving this to 2.6.0 |
Works for #114850 Pull Request resolved: #138189 Approved by: https://github.com/EikanWang, https://github.com/malfet, https://github.com/atalman
8 participants
Works for #114850