The incremental squeeze functions in the portable SHAKE XOF API, when attempting to squeeze more than RATE (168 for SHAKE128, 136 for SHAKE256) bytes, performed an additional permutation of the state before producing the first output block, thus discarding the first block of RATE bytes of valid XOF output.

This bug impacts users that rely on this XOF API to squeeze more than RATE bytes. It does not impact the use of libcrux-sha3 in libcrux-ml-kem or libcrux-ml-dsa.

Starting from version 0.0.8 the squeeze functions correctly output all blocks including the first block.