Print Error message when a module requests a data product without first registering the request via 'consumes' #79
Conversation
…onding consumes call Each time a data product is gotten from the Event, LuminosityBlock or Run without having the module calling the corresponding consumes, mayConsume or consumesMany call we issue a MessageLogger warning. This required passing back from the Principal exactly which ProductHolderIndex was used to find the data. In the fully threaded framework this will be changed to throw an exception.
Temporarily have the registeredToConsume* functions only return false the first time an unregistered item is checked. This avoids having the checker issue a warning message each time. This is only temporary since this is not thread safe and in the threaded framework such a case would cause a fatal exception.
…as been registered by the EDConsumerBase since if it has not been registered we need to fail before any attempt is made to run the Producer in unscheduled mode.
…ssingConsumesWarning on 2013/07/11
|
@nclopezo, can you please test this as usual? Thanks. |
|
Hi, I took CMSSW_7_0_X_2013-07-11-0200 and pulled these changes but I got an scram p CMSSW_7_0_X_2013-07-11-0200 And the error that I got is the following: In file included from David Mendez On Thu, Jul 11, 2013 at 3:04 PM, Giulio Eulisse notifications@github.comwrote:
|
|
@nclopezo please format messages so that we get nice reports. Eg.: Have a look at: http://github.github.com/github-flavored-markdown/ for the syntax. @Dr15Jones can you look at the actual issue? |
|
This is the error: In file included from /build/dmendezl/CMSSW_7_0_X_
2013-07-11-0200/src/SimGeneral/MixingModule/src/PileUpEventPrincipal.cc:1:0:
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/SimGeneral/MixingModule/interface/PileUpEventPrincipal.h: In member function 'bool PileUpEventPrincipal::getByLabel(const edm::InputTag&, edm::Handle&) const':
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/SimGeneral/MixingModule/interface/PileUpEventPrincipal.h:53:95: error: no matching function for call to 'edm::EventPrincipal::getByLabel(edm::KindOfType, edm::TypeID, const edm::InputTag&) const'
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/SimGeneral/MixingModule/interface/PileUpEventPrincipal.h:53:95: note: candidates are:
In file included from /build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/FWCore/Framework/interface/EventPrincipal.h:22:0,
from /build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/SimGeneral/MixingModule/interface/PileUpEventPrincipal.h:7,
from /build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/SimGeneral/MixingModule/src/PileUpEventPrincipal.cc:1:
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/FWCore/Framework/interface/Principal.h:107:18: note: edm::BasicHandle edm::Principal::getByLabel(edm::KindOfType, const edm::TypeID&, const edm::InputTag&, const edm::EDConsumerBase*) const
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/FWCore/Framework/interface/Principal.h:107:18: note: candidate expects 4 arguments, 3 provided
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/FWCore/Framework/interface/Principal.h:112:18: note: edm::BasicHandle edm::Principal::getByLabel(edm::KindOfType, const edm::TypeID&, const string&, const string&, const string&, const edm::EDConsumerBase*) const
/build/dmendezl/CMSSW_7_0_X_2013-07-11-0200/src/FWCore/Framework/interface/Principal.h:112:18: note: candidate expects 6 arguments, 3 provided
|
|
The change I just made to my branch fixes the problem. Please try pulling again. |
|
The following categories have been signed by chrjones (a.k.a. @Dr15Jones on GitHub): Core @cms-git-visualization, @cms-git-core, @cms-git-geometry |
|
Hi, |
Print Error message when a module requests a data product without first registering the request via 'consumes'
update LCT reconstruction in ME1/a + Add GEM pad matching functions
Use the github status api to indicate the status of the pull request
…CMSSW_7_0_6 Updates to 7.0.X: electron ids, mc matching, fix multiloop
save post-fit snapshot in workspace for MultiDimFit and add ability to l...
added current datasets in 74X at 25ns
…heck Thank you! and also thanks for the independent PRs, this is very practical for the review.
Pu algorithm updates
Updated L1 information using dedicated MINIAOD collections
…DReview Two additional review updates:
* Fix to compiler warning and removal of hardoded value in MatchCalculator * Updates to fix names of pins to make unique * Migrate to new format of table for phi correction memory * Remove unused TETableBase class * Cleanup of VMRouterTable * Cleanup of LUT for bend consistency * Cleanup of LUT used in MEU for bend consistency * Cleanup code for disk bend table * cleanup of code to generate table used by TE * Cleanup of TP LUTs * Cleanup of LUTs for MatchCalculator cuts * Fix names of new LUT when writing to files * Cleanup of writing LUTs for the combined modules * Add code to write projection bend table for disk projections * More cleanup to remove hard coded layer and seed numbers * Cleanup of hardcoded numbers * Remove few more hardcoded constants: * Some further cleanup based on comments from Louise S. Co-authored-by: Anders Ryd <ryd@Anderss-MacBook-Pro.local> Co-authored-by: Louise Skinnari <louise.skinnari@cern.ch>
* Fix to compiler warning and removal of hardoded value in MatchCalculator * Updates to fix names of pins to make unique * Migrate to new format of table for phi correction memory * Remove unused TETableBase class * Cleanup of VMRouterTable * Cleanup of LUT for bend consistency * Cleanup of LUT used in MEU for bend consistency * Cleanup code for disk bend table * cleanup of code to generate table used by TE * Cleanup of TP LUTs * Cleanup of LUTs for MatchCalculator cuts * Fix names of new LUT when writing to files * Cleanup of writing LUTs for the combined modules * Add code to write projection bend table for disk projections * More cleanup to remove hard coded layer and seed numbers * Cleanup of hardcoded numbers * Remove few more hardcoded constants: * Some further cleanup based on comments from Louise S. Co-authored-by: Anders Ryd <ryd@Anderss-MacBook-Pro.local> Co-authored-by: Louise Skinnari <louise.skinnari@cern.ch>
Global Trigger data types & jets pattern files
…hetaTP Am 11 2 1 new analyzer theta tp
The threaded framework requires that all modules register what data they may request by calling 'consumes', 'mayConsume' or 'consumesMany' in the module's constructor. This change will catch the cases where this was not done and on the first such request it will print an Error message to the MessageLogger.
When we move to the threaded framework all such unregistered data requests will cause a fatal exception to be thrown.