Details

    • Type: Bug
    • Status: Closed (View Workflow)
    • Priority: High
    • Resolution: Fixed
    • Affects Version/s: 6.02/02
    • Fix Version/s: None
    • Component/s: Cling
    • Labels:
      None
    • Environment:

      Linux

      Description

      From emails between Marcin and Peter:

      This is a bit crazy: ph_cl_phiCalo is actually a name of a method argument. It does not show up in any dictionary.

      What's more interesting is that ROOT crashes when I try to #include the header in which this method is defined. This could be related.
      (btw, this is exactly the header where we found (and fixed) namespace clash long time ago)

      To reproduce the crash:

      # python
      > import ROOT
      > ROOT.gInterpreter.ProcessLine('#include "egammaMVACalib.h"')

      I'm attaching the header.

      Cheers,
      Marcin

      > On 16.12.2014 19:58, Peter Van Gemmeren wrote:
      >> I am investigating one of our Reconstruction test failures that could potentially point to Cling: http://atlas-computing.web.cern.ch/atlas-computing/links/distDirectory/nightlies/otherWebArea/nicos_web_areaLCG464BS6G48RecDbg/NICOS_TestLog_rel_2/Reconstruction_RecExample_RecExRecoTest_21__RecExRecoTest_TestConfiguration__calomax__a.html fails with:
      >>
      >>
      >>

      input_line_348:5:1: error: expected unqualified-id
       
      extern int __Cling_Autoloading_Map;
       
       ^
       
      input_line_348:5:1: error: expected ')'
       
      lookup.type.by.name.file:1:9: note: to match this '('
       
       fmod(abs(ph_cl_phiCalo
       
               ^
       
      python: /mnt/build/dev4/x86_64-slc6-gcc48-dbg-build/projects/ROOT-v6-02-00-patches/src/ROOT/v6-02-00-patches/core/meta/src/TCling.cxx:1527: virtual void TCling::RegisterModule(const char*, const char**, const char**, const char*, const char*, void (*)(), const FwdDeclArgsToKeepCollection_t&, const char**): Assertion `cling::Interpreter::kSuccess == compRes && "The forward declarations could not be compiled"' failed.

      >>
      >>
      >> I was trying to find the offending line in a Dict_gen.cpp, but didn┬┐t. It seems the only occurrence of ph_cl_phiCalo is in egammaMVACalib (http://acode-browser.usatlas.bnl.gov/lxr/ident?_i=ph_cl_phiCalo).
      >>
      >>
      >>
      >> Do you have any idea or debugging advice?
      >>
      >>
      >>
      >> Thanks, Peter
      >>
      >>

        Attachments

          Activity

            People

            • Assignee:
              axel Axel Naumann
              Reporter:
              axel Axel Naumann
            • Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: