Skip to content

Guess=Core broken (HeH+ test case) #2

Description

@mbanck

I've debugged the report by Noel O'Boyle in http://lists.nongnu.org/archive/html/freeon-users/2014-01/msg00000.html For me, only the HeH+ test case failed.

It appears NLink never gets initialized for the Guess=Core case in QCTC.F90, as ElectronsOn is set to .false. in this case and thus MakeRhoList() is never called. This leads to bogus NLink values like "11129" being passed to density::collate(), where the assert around line 211 (SUM(NZ(1:NExpt)).NE.NLink) fails.

If I set NLink to 0 in case MakeRhoList() is skipped in QCTC.F90 , the test case passes successfully (though I am not sure this is the correct fix). If I change the HeH+ test case to use Guess=Superpos as starting guess, it does not crash, either.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions