In cryptography, the bricklayer function is a part of a round function that can be decomposed into identical independent Boolean operations on the partitioned pieces of its input data, so called bundles. The term was introduced by Daemen and Rijmen in 2001.
If the underlying function transforming the bundle is nonlinear, it is traditionally called an S-box. If the function is linear, Daemen and Rijmen use for it the term D-box (after diffusion).