In VTM2.0.1 deblocking is only considered on 8x8 grids. In this contribution, it is proposed to apply deblocking for 4xN and Nx4 block boundaries on not only 8x8 grids but also non-8x8 grids. In addition, deblocking processes are modified for 4xN and Nx4 block boundaries to allow parallel deblocking of all parallel block boundaries. Comparing with VTM2.0.1, simulation results reportedly show 0.00%, -0.15%, -0.26%, and -0.25% luma BD-rates for AI, RA, LB, and LP configurations, respectively, with negligible run-time changes. Comparing with VTM2.0.1 with ALF switched off, simulation results reportedly show -0.02%, -0.25%, -0.22%, and -0.39% luma BD-rates for AI, RA, LB, and LP, respectively, with negligible run-time changes. It is asserted that the subjective quality especially for low resolution sequences is improved.