Recover the Secret Components in a ForkCipher

Recently, a new cryptographic primitive has been proposed called ForkCiphers. This paper aims at proposing generic cryptanalysis against such constructions. We give method to apply existing decompositions againt the underlying block cipher $\mathcal{E}^{r}$ on forking variant xmlns:xlink="http://www.w3.org/1999/xlink">$\text{Fork}\mathcal{E}-(r-1)-r_{0}-(r+1-r_{0})$ . As application, we consider security of ForkSPN and ForkFN with secret inner functions. provide attack xmlns:xlink="http://www.w3.org/1999/xlink">$\text{ForkSPN}-2-r_{0}-(4-r_{0})$ based decomposition SASAS. And also extend Biryukov et al. Feistel networks in SAC 2015 get all unknown round functions xmlns:xlink="http://www.w3.org/1999/xlink">$\text{ForkFN}-r-r_{0}-r_{1}$ for xmlns:xlink="http://www.w3.org/1999/xlink">$r\leq 6$ xmlns:xlink="http://www.w3.org/1999/xlink">$r_{0}+r_{1}\leq 8$ Therefore, compared original cipher, version requires more iteration rounds resist recovery attack.