skinny128.c 8.05 KB
Newer Older
Alexandre Adomnicai committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
/******************************************************************************
* Fixsliced implementation of SKINNY-128-384.
* Two blocks are processed in parallel.
*
* This implementation doesn't compute the ShiftRows operation. Some masks and
* shifts are applied during the MixColumns operation so that the proper bits
* are XORed together. Moreover, the row permutation within the MixColumns 
* is omitted, as well as the bit permutation at the end of the Sbox. The rows
* are synchronized with the classical after only 4 rounds. However, the Sbox
* permutation requires 8 rounds for a synchronization. To limit the impact
* on code size, we compute the permutation every 4 rounds. Therefore, this
* implementation relies on a "QUADRUPLE_ROUND" routine.
*
* For more details, see the paper at: https://
*
* @author	Alexandre Adomnicai, Nanyang Technological University,
*			alexandre.adomnicai@ntu.edu.sg
*
* @date		May 2020
******************************************************************************/
#include <stdio.h>
#include <string.h>
#include "skinny128.h"
#include "tk_schedule.h"

/****************************************************************************
* The MixColumns operation for rounds i such that (i % 4) == 0.
****************************************************************************/
void mixcolumns_0(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],24) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,30);
		tmp = ROR(state[i],16) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,4);
		tmp = ROR(state[i],8) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,2);
	}
}

/****************************************************************************
* The MixColumns operation for rounds i such that (i % 4) == 1.
****************************************************************************/
void mixcolumns_1(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],16) & 0x30303030;
		state[i] ^= ROR(tmp,30);
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,28);
		tmp = ROR(state[i],16) & 0x30303030;
		state[i] ^= ROR(tmp,2);
	}
}

/****************************************************************************
* The MixColumns operation for rounds i such that (i % 4) == 2.
****************************************************************************/
void mixcolumns_2(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],8) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,6);
		tmp = ROR(state[i],16) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,28);
		tmp = ROR(state[i],24) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,2);
	}
}

/****************************************************************************
* The MixColumns operation for rounds i such that (i % 4) == 3.
****************************************************************************/
void mixcolumns_3(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,30);
		tmp = state[i] & 0x30303030;
		state[i] ^= ROR(tmp,4);
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,26);
	}
}

/****************************************************************************
* The inverse MixColumns oepration for rounds i such that (i % 4) == 0
****************************************************************************/
void inv_mixcolumns_0(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],8) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,2);
		tmp = ROR(state[i],16) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,4);
		tmp = ROR(state[i],24) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,30);
	}
}

/****************************************************************************
* The inverse MixColumns oepration for rounds i such that (i % 4) == 1
****************************************************************************/
void inv_mixcolumns_1(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],16) & 0x30303030;
		state[i] ^= ROR(tmp,2);
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,28);
		tmp = ROR(state[i],16) & 0x30303030;
		state[i] ^= ROR(tmp,30);
	}
}

/****************************************************************************
* The inverse MixColumns oepration for rounds i such that (i % 4) == 2
****************************************************************************/
void inv_mixcolumns_2(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = ROR(state[i],24) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,2);
		tmp = ROR(state[i],16) & 0x0c0c0c0c;
		state[i] ^= ROR(tmp,28);
		tmp = ROR(state[i],8) & 0xc0c0c0c0;
		state[i] ^= ROR(tmp,6);
	}
}

/****************************************************************************
* The inverse MixColumns oepration for rounds i such that (i % 4) == 3
****************************************************************************/
void inv_mixcolumns_3(u32* state) {
	u32 tmp;
	for(int i = 0; i < 8; i++) {
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,26);
		tmp = state[i] & 0x30303030;
		state[i] ^= ROR(tmp,4);
		tmp = state[i] & 0x03030303;
		state[i] ^= ROR(tmp,30);
	}
}

/****************************************************************************
* Adds the tweakey (including the round constants) to the state.
****************************************************************************/
void add_tweakey(u32* state, const u32* rtk1, const u32* rtk2_3) {
	state[0] ^= rtk1[0] ^ rtk2_3[0];
	state[1] ^= rtk1[1] ^ rtk2_3[1]; 
	state[2] ^= rtk1[2] ^ rtk2_3[2];
	state[3] ^= rtk1[3] ^ rtk2_3[3];
	state[4] ^= rtk1[4] ^ rtk2_3[4];
	state[5] ^= rtk1[5] ^ rtk2_3[5];
	state[6] ^= rtk1[6] ^ rtk2_3[6];
	state[7] ^= rtk1[7] ^ rtk2_3[7];
}

/****************************************************************************
* Encryption of 2 blocks in parallel using SKINNY-128-384.
* The input parameters 'rtk1' and 'rtk2_3' are given seperately to avoid
* unnecessary recomputations of the entire tk schedule during SKINNY-AEAD-M1.
****************************************************************************/
void skinny128_384_encrypt(u8* ctext, u8* ctext_bis, const u8* ptext, 
					const u8* ptext_bis, const tweakey tk) {
	u32 state[8];
	packing(state, ptext, ptext_bis);
	QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3);
	QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+32);
	QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+64);
	QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+96);
	QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+128);
	QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+160);
	QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+192);
	QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+224);
	QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+256);
	QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+288);
	QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+320);
	QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+352);
	QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+384);
	QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+416);
	unpacking(ctext, ctext_bis, state);
}

/****************************************************************************
* Decryption of 2 blocks in parallel using SKINNY-128-384.
* The input parameters 'rtk1' and 'rtk2_3' are given seperately to avoid
* unnecessary recomputations of the entire tk schedule during SKINNY-AEAD-M1.
****************************************************************************/
void skinny128_384_decrypt(u8* ptext, u8* ptext_bis, const u8* ctext, 
					const u8* ctext_bis, const tweakey tk) {
	u32 state[8];
	packing(state, ctext, ctext_bis);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+416);
	INV_QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+384);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+352);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+320);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+288);
	INV_QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+256);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+224);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+192);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+160);
	INV_QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3+128);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+96, 	tk.rtk2_3+96);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+64, 	tk.rtk2_3+64);
	INV_QUADRUPLE_ROUND(state, tk.rtk1+32, 	tk.rtk2_3+32);
	INV_QUADRUPLE_ROUND(state, tk.rtk1, 	tk.rtk2_3);
	unpacking(ptext, ptext_bis, state);
}