PR Profiling: Add link for diff of IB and PR Vtune reports and add script to generate page with the diff of IB and PR sorted-CPU-RES reports#2787
Conversation
|
A new Pull Request was created by @gartung for branch master. @akritkbehera, @cmsbuild, @iarspider, @raoatifshad, @smuzaffar can you please review it and eventually sign? Thanks. |
|
cms-bot internal usage |
|
enable profiling |
|
Pull request #2787 was updated. |
1 similar comment
|
Pull request #2787 was updated. |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3d44c3/53842/summary.html Comparison SummarySummary:
|
|
Need to wait for the next IB to generate the vtune top-down json file before adding the comparison scripting. |
|
Pull request #2787 was updated. |
2 similar comments
|
Pull request #2787 was updated. |
|
Pull request #2787 was updated. |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3d44c3/53846/summary.html Comparison SummarySummary:
|
…filtering options
…tune event loop comparison
… top-down reports
…t Loop Compare Viewer.
|
Pull request #2787 was updated. |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3d44c3/53985/summary.html Comparison SummarySummary:
|
|
@makortel I generalized the script that generates the doEvent summary to include all produce functions. I had copilot look at all produce functions and find the parent function. It found that "EventSetup Producers" were not included in the parent function list. This should lead to no missing produce functions in the diff report unless a new producer was added/removed in the PR. |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3d44c3/53992/summary.html Comparison SummarySummary:
|
|
please test |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-3d44c3/53998/summary.html Comparison SummarySummary:
|
|
I think the EventSetup producers should not be added in "RECO event loop". The EventSetup producers are not run on every event, but on IOV boundaries (which is likely once per job in all test jobs). |
|
OK. Adding them did not make the number of IB vs PR functions reported the same. I will remove them from the report. |

No description provided.