-
Notifications
You must be signed in to change notification settings - Fork 4.3k
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
Fix uploading the EventSetup conditions to multiple CUDA devices [11.3.x] #34950
Fix uploading the EventSetup conditions to multiple CUDA devices [11.3.x] #34950
Conversation
Add a default constructor that stores the current devices, without changing it. Add a set() method to change the current device, without affecting the stored value for the original device.
Associate to the correct CUDA device the events used to track if the conditions have been transferred to each device.
type bugfix |
A new Pull Request was created by @fwyzard (Andrea Bocci) for CMSSW_11_3_X. It involves the following packages:
@makortel, @cmsbuild, @fwyzard can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
backport #34948 |
please test |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-fbd004/17885/summary.html Comparison SummarySummary:
|
+heterogeneous |
This pull request is fully signed and it will be integrated in one of the next CMSSW_11_3_X IBs (tests are also fine) and once validation in the development release cycle CMSSW_12_1_X is complete. This pull request will now be reviewed by the release team before it's merged. @perrotta, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2) |
hold |
Pull request has been put on hold by @makortel |
please test |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-fbd004/17950/summary.html Comparison SummarySummary:
|
unhold |
@cmsbuild, please test workflow 136.885501 Just to make sure |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-fbd004/17952/summary.html Comparison Summary@slava77 comparisons for the following workflows were not done due to missing matrix map:
Summary:
|
+1 |
This pull request is fully signed and it will be integrated in one of the next CMSSW_11_3_X IBs (tests are also fine) and once validation in the development release cycle CMSSW_12_1_X is complete. This pull request will now be reviewed by the release team before it's merged. @perrotta, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2) |
+1
|
PR description:
When multiple CUDA devices are available, each CUDA EventSetup object uses a different CUDA event to keep track if the conditions have been uploaded to each device. However, all events were associated to the default device, instead of the correct one, leading to a runtime error when multiple devices were used.
These changes associate to the correct CUDA device the events used to track if the conditions have been transferred.
PR validation:
See #34948 .
If this PR is a backport please specify the original PR and why you need to backport that PR:
Backport of #34948 .
Would be useful to have, if there are plans to use this release cycle for more online tests.