floating invalid calling DGEQRF with IFORT and MKL

I have implemented an orthogonalization method using the DGEQRF subroutine to compute a QR factorization. The compilation and subsequently execution works well using gfortran/MKL but  when I compile and run it with  Ifort/MKL I get the following error:

forrtl: error (65): floating invalid 

I have checked for uninitialized variables or NaN values, and the problem seems to be an arithmetic operation inside DGEQRF.

Issue description:

Ifort version: ifort (IFORT) 19.0.1.144 20181018 

MKL version: parallel_studio_xe_2019_update1_cluster_edition

The source code can be found at: https://github.com/NLESC-JCER/Fortran_Davidson

Line that cause the error: https://github.com/NLESC-JCER/Fortran_Davidson/blob/master/src/davidson....

I compile the code using cmake like: 

 cmake -H. -Bbuild -DCMAKE_Fortran_COMPILER=ifort DCMAKE_BUILD_TYPE=Debug

 cmake --build build

I run the resulting binary like:

./bin/main

Finally, The orthogonalization subroutine is

  subroutine lapack_qr(basis)
    !> Orthoghonalize the basis using the QR factorization.
    !> QR factorization of the M-by-N (M>N) matrx A=Q*R in the form where
    !> Q is square M-by-M matrix and R is an upper triangular M-by-N matrix.
    !> The equality A=Q*R can be re-written also as a product Q1*R1 where Q1
    !> is a rectangular M-by-N submatrix of the matrix Q and R1 is M-by-M
    !> submatrix of the R. Let us note that columns of Q1 are orthonormal
    !> (they are orthogonal to each other and have norms equal to 1).
    !> The equality A=Q1*R1 can be treated as every column of A is a linear
    !> combination of Q1 columns, i.e. they span the same linear space.
    !> In other words, columns of Q1 is the result of ortogonalization of columns A.
    !> DGEQRF does not not compute Q directly, DORGQR must be call subsequently.
    
    !> \param basis
    !> \return orthogonal basis    

    implicit none
    real(dp), dimension(:, :), intent(inout) :: basis
    real(dp), dimension(:), allocatable :: work ! workspace, see lapack documentation
    real(dp), dimension(size(basis, 2)) :: tau ! see DGEQRF documentation
    integer :: info, lwork, m, n

    ! Matrix shape
    m = size(basis, 1)
    n = size(basis, 2)

    ! 1. Call the QR decomposition
    ! 1.1 Query size of the workspace (Check lapack documentation)
    allocate(work(1))
    call DGEQRF(m, n, basis, m, tau, work, -1, info)

    ! 1.2 Allocate memory for the workspace
    lwork = max(1, int(work(1)))
    deallocate(work)
    allocate(work(lwork))

    ! 1.3 Call QR factorization
    call DGEQRF(m, n, basis, m, tau, work, lwork, info)
    deallocate(work)
    
    ! 2. Generates an orthonormal matrix
    ! 2.1 Query size of the workspace (Check lapack documentation)
    allocate(work(1))
    call DORGQR(m, n, min(m, n), basis, m, tau, work, -1, info)

    ! 2.2 Allocate memory fo the workspace
    lwork = max(1, int(work(1)))
    deallocate(work)
    allocate(work(lwork))

    ! 2.3 compute the matrix Q
    call DORGQR(m, n, min(m, n), basis, m, tau, work, lwork, info)
    
    ! release memory
    deallocate(work)
    
  end subroutine lapack_qr

I really appreciate any help you can provide.

Best,

Felipe Z.