-
Notifications
You must be signed in to change notification settings - Fork 4.6k
Follow up to the Alpaka integration in CMSSW #43853
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
Conversation
|
cms-bot internal usage |
|
assign heterogeneous |
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-43853/38691 ERROR: Build errors found during clang-tidy run. |
64e5704 to
a884a4d
Compare
|
test parameters:
|
|
please test |
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-43853/38692
|
|
Pull request #43853 was updated. @jfernan2, @francescobrivio, @subirsarkar, @consuegs, @perrotta, @srimanob, @saumyaphor4252, @mandrenguyen can you please check and sign again. |
|
+Upgrade |
|
+alca |
|
+db |
|
are the tests stuck? |
No @mandrenguyen , tests are running. Jenkins is very busy today that is why tests are waiting for free nodes |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-536c0d/37363/summary.html Comparison SummarySummary:
GPU Comparison SummarySummary:
|
|
+1 |
|
This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @sextonkennedy, @rappoccio, @antoniovilela (and backports should be raised in the release meeting by the corresponding L2) |
|
+1 |
|
Yes, #43879 . |
PR description:
HeterogeneousCore/AlpakaInterfaceRewrite the work division utilities:
alpakatoolsutilities.Rewrite the
zeroAndInitkernel using thealpakatoolsutilities.Rename
elements_with_stridetouniform_elementsin user code.RecoLocalTracker/SiPixelClusterizerImplement fixes that were lost during the CUDA-to-Alpaka migration.
Collect the updates from #39711, #42618, #42977 and #43052, that were missing from the Alpaka implementation:
invalid ROC number error (
errorType=36);errorType=27);errorType=26,27,30with the legacy code;printfstatements at compile time.RecoTracker/PixelSeedingDo not call the fishbone for events with pixel hits only in BPIX1 (reimplement #35638).
RecoTracker/MeasurementDetAdd a check on the pixel ROC range, to avoid overflows in the (unsigned) range differences.
DataFormats/TrackingRecHitSoAExtend
TrackingRecHitscollections for testing and their unit tests.CondFormats/SiPixelObjectsClean up includes and dependencies, and avoid including framework headers in the device compilation.
Geometry/CommonTopologiesUpdate comments to pixel topologies.
Other changes
Fix include guards after files were moved to different packages.
Clean up include files.
Simplify nested namespace declarations.
PR validation:
Unit tests pass.
Validated running a recent HLT menu over 20k events without issues, both on GPU and on CPU, on data and MC: the alpaka implementation running on CPU or GPU gives the same results as the CUDA implementation, within the same non-reproducibility margins (1 event or up to 1% for high-rate triggers).
Backport status
The backport to CMSSW_14_0_X for 2024 data taking is in #43879.