add a nan protection to avoid crash in fastjet call from pvsorter #14846
Conversation
|
A new Pull Request was created by @arizzi for CMSSW_8_0_X. It involves the following packages: CommonTools/RecoAlgos @cmsbuild, @cvuosalo, @slava77, @davidlange6 can you please review it and eventually sign? Thanks. cms-bot commands are list here #13028 |
|
@cmsbuild please test |
|
The tests are being triggered in jenkins. |
|
@arizzi I'm still expecting 81X PR |
| if(c->pt()!=0) { | ||
| scale=(c->pt()-c->bestTrack()->ptError())/c->pt(); | ||
| } | ||
| if(isnan(scale)) { |
There was a problem hiding this comment.
The static analyzer warns: "std::isnan does not work when fast-math is used. Please use edm::isNotFinite from 'FWCore/Utilities/interface/isNotFinite.h'".
Please fix this line.
|
@arizzi: When might this PR get updated? |
|
@arizzi: When will the 81X version of this PR be submitted? |
cmsbuild
added
tests-pending
comparison-pending
and removed
comparison-available
tests-approved
labels
|
Pull request #14846 was updated. @cmsbuild, @cvuosalo, @slava77, @davidlange6 can you please check and sign again. |
|
81X PR made (just cherry picking the two commits, no other test performed) |
|
@cmsbuild please test |
|
The tests are being triggered in jenkins. |
|
+1 Bug fix to prevent crash due to a NAN in FastJet call from PV sorter. There should be no change in monitored quantities. #14945 is the 81X version of this PR. The code fix is satisfactory, and Jenkins tests against baseline CMSSW_8_0_X_2016-06-20-1100 show no significant differences, as expected. |
cmsbuild
added
reconstruction-approved
fully-signed
and removed
pending-signatures
reconstruction-pending
labels
|
This pull request is fully signed and it will be integrated in one of the next CMSSW_8_0_X IBs (tests are also fine). This pull request requires discussion in the ORP meeting before it's merged. @slava77, @davidlange6, @Degano, @smuzaffar |
|
+1 |
this should address
https://hypernews.cern.ch/HyperNews/CMS/get/recoDevelopment/1451/2/1/1/1.html
by adding a protection on "nan" being passed to fastjet.
the nan seems to be coming from ptError and should be further investigated