[Paraview] Paraview 5.4.1 FTBS on Debian/Testing with gcc 7.2 when -DPARAVIEW_USE_VISITBRIDGE=ON

[Ben Boeckel]

On Thu, Oct 12, 2017 at 11:41:46 +0200, Christophe Trophime wrote:
> /usr/include/x86_64-linux-gnu/qt5/QtWidgets -isystem /usr/include/x86_64-linux-gnu/qt5/QtGui -isystem /usr/include/x86_64-linux-gnu/qt5/QtCore -isystem /usr/lib/x86_64-linux-gnu/qt5/mkspecs/linux-g++-64 -g -O2 -fdebug-prefix-map=/build/paraview-5.4.1=. -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -fPIC -std=c++11 -o CMakeFiles/vtkPVVTKExtensionsRenderingPythonD.dir/vtkZlibImageCompressorPython.cxx.o -c /build/paraview-5.4.1/obj-x86_64-linux-gnu/VTK/Wrapping/Python/vtkZlibImageCompressorPython.cxx 
> /build/paraview-5.4.1/Utilities/VisItBridge/databases/readers/Vs/VsStaggeredField.C: In constructor 'VsStaggeredField<TYPE>::VsStaggeredField(size_t)': 
> /build/paraview-5.4.1/Utilities/VisItBridge/databases/readers/Vs/VsStaggeredField.C:30:24: error: 'class std::valarray<int>' has no member named 'clear' 
> this->oriCellDimProd.clear(); 
> ^~~~~ 

Hmm, my browser history says I searched for this back in July… looking
around there finds:

    https://gitlab.kitware.com/paraview/paraview/issues/17456

which seems to have dropped off our radar. Is the Vs reader necessary?
If not, feel free to disable it, otherwise the reader will need fixed
(it seems it has never worked though since it did this since its
addition to visit upstream).

> extra_flags += -Wno-dev \ 
> -DCMAKE_C_COMPILER=mpicc \ 
> -DCMAKE_CXX_COMPILER=mpic++ \ 

This probably isn't related, but this is not the typical way to use MPI
with CMake projects. They discover and use the information from the
wrappers themselves when they find MPI.

--Ben