wall tangential velocity components will have zero gradients (and zero value) along wall due to no slip and this from differential form of continuity equation (consider 2-D for simplicity)implies wall normal component will have zero gradient on wall (in addition to being zero for non-porous boundaries)...

how is this implemented in a cfd code..I read somewhere that while discretizing in Control Volume formulation the first derivative of the normal velocity on faces aligned with walls should be made zero...

I have faced some problems when I implemented this (which I will mention in a new thread)...anyone who have success doing this or something similar/more/more appropriate on this line...

in Control Volume formulation the first derivative of the normal velocity on faces aligned with walls should be made zero: --> where do you use this derived condition, in normal shear stress, or in pressure Poisson eq.? usually, noslip condition is enough.

in normal stress..

I agree with versi, no-slip condition is sufficient. Why not leave the gradient of tangential velocity as it is (probably not zero) in the normal stress. It seems to me no need to force it to be zero at all.