Bug in boundary condition PressureNonEquilibrium

While creating the app LaminarPipeFlow, I found that the boundary condition PressureNonEquilibriumCompressible is not working correctly. There is a sudden jump in the pressure next to the row, where the pressure is set by the boundary condition (MX and PX). The pressure is also a little too low (3.2e-6 instead of 3.3e-6) The correct values are displayed in the top left corner

For reference: This is what the density should look like (image created with the CPU sublibrary)

In the CPU-version of this bc, the pressure is only set for distributions that are normal to the boundary condition. In the GPU-version the pressure is set for all distributions. This probably causes the jump in the pressure.