2 * copyright (c) 2007 Michael Niedermayer <michaelni@gmx.at>
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 * some optimization ideas from aes128.c by Reimar Doeffinger
27 uint8_t round_key[15][4][4];
32 const int av_aes_size= sizeof(AVAES);
34 static const uint8_t rcon[10] = {
35 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36
38 static uint8_t sbox[256];
39 static uint8_t inv_sbox[256];
41 static uint32_t enc_multbl[1][256];
42 static uint32_t dec_multbl[1][256];
44 static uint32_t enc_multbl[4][256];
45 static uint32_t dec_multbl[4][256];
48 static inline void addkey(uint64_t state[2], uint64_t round_key[2]){
49 state[0] ^= round_key[0];
50 state[1] ^= round_key[1];
53 static void subshift(uint8_t s0[4], uint8_t s1[4], uint8_t s2[4], uint8_t s3[4], uint8_t *box){
55 s0[0]=box[s0[ 0]]; s0[ 4]=box[s0[ 4]]; s0[ 8]=box[s0[ 8]]; s0[12]=box[s0[12]];
56 t=s1[0]; s1[0]=box[s1[ 4]]; s1[ 4]=box[s1[ 8]]; s1[ 8]=box[s1[12]]; s1[12]=box[t];
57 t=s2[0]; s2[0]=box[s2[ 8]]; s2[ 8]=box[ t]; t=s2[ 4]; s2[ 4]=box[s2[12]]; s2[12]=box[t];
58 t=s3[0]; s3[0]=box[s3[12]]; s3[12]=box[s3[ 8]]; s3[ 8]=box[s3[ 4]]; s3[ 4]=box[t];
61 #define ROT(x,s) ((x<<s)|(x>>(32-s)))
63 static inline void mix(uint8_t state[4][4], uint32_t multbl[4][256]){
67 ((uint32_t *)(state))[i] = multbl[0][state[i][0]] ^ ROT(multbl[0][state[i][1]], 8)
68 ^ROT(multbl[0][state[i][2]],16) ^ ROT(multbl[0][state[i][3]],24);
70 ((uint32_t *)(state))[i] = multbl[0][state[i][0]] ^ multbl[1][state[i][1]]
71 ^multbl[2][state[i][2]] ^ multbl[3][state[i][3]];
75 static inline void mix2(uint8_t state[4][4], uint32_t multbl[4][256], int s1, int s3){
76 int a = multbl[0][state[0][0]] ^ multbl[1][state[s1 ][1]]
77 ^multbl[2][state[2][2]] ^ multbl[3][state[s3 ][3]];
78 int b = multbl[0][state[1][0]] ^ multbl[1][state[s3-1][1]]
79 ^multbl[2][state[3][2]] ^ multbl[3][state[s1-1][3]];
80 int c = multbl[0][state[2][0]] ^ multbl[1][state[s3 ][1]]
81 ^multbl[2][state[0][2]] ^ multbl[3][state[s1 ][3]];
82 int d = multbl[0][state[3][0]] ^ multbl[1][state[s1-1][1]]
83 ^multbl[2][state[1][2]] ^ multbl[3][state[s3-1][3]];
85 ((uint32_t *)(state))[0]=a;
86 ((uint32_t *)(state))[1]=b;
87 ((uint32_t *)(state))[2]=c;
88 ((uint32_t *)(state))[3]=d;
91 static inline void crypt(AVAES *a, int s, uint8_t *sbox, uint32_t *multbl){
94 for(r=a->rounds; r>1; r--){
95 addkey(a->state, a->round_key[r]);
96 mix2(a->state, multbl, 3-s, 1+s);
98 addkey(a->state, a->round_key[1]);
99 subshift(a->state[0], a->state[0]+3-s, a->state[0]+2, a->state[0]+1+s, sbox);
100 addkey(a->state, a->round_key[0]);
103 static void aes_decrypt(AVAES *a){
104 crypt(a, 0, inv_sbox, dec_multbl);
107 static void aes_encrypt(AVAES *a){
108 crypt(a, 2, sbox, enc_multbl);
111 static void init_multbl2(uint8_t tbl[1024], int c[4], uint8_t *log8, uint8_t *alog8, uint8_t *sbox){
113 for(i=0; i<1024; i++){
115 if(x) tbl[i]= alog8[ log8[x] + log8[c[i&3]] ];
118 for(j=256; j<1024; j++)
120 tbl[4*j+i]= tbl[4*j + ((i-1)&3) - 1024];
124 // this is based on the reference AES code by Paulo Barreto and Vincent Rijmen
125 int av_aes_init(AVAES *a, uint8_t *key, int key_bits, int decrypt) {
126 int i, j, t, rconpointer = 0;
133 if(!enc_multbl[4][1023]){
135 for(i=0; i<255; i++){
142 for(i=0; i<256; i++){
143 j= i ? alog8[255-log8[i]] : 0;
144 j ^= (j<<1) ^ (j<<2) ^ (j<<3) ^ (j<<4);
145 j = (j ^ (j>>8) ^ 99) & 255;
149 init_multbl2(dec_multbl[0], (int[4]){0xe, 0x9, 0xd, 0xb}, log8, alog8, inv_sbox);
150 init_multbl2(enc_multbl[0], (int[4]){0x2, 0x1, 0x1, 0x3}, log8, alog8, sbox);
153 if(key_bits!=128 && key_bits!=192 && key_bits!=256)
158 memcpy(tk, key, KC*4);
160 for(t= 0; t < (rounds+1)*4;) {
161 memcpy(a->round_key[0][t], tk, KC*4);
164 for(i = 0; i < 4; i++)
165 tk[0][i] ^= sbox[tk[KC-1][(i+1)&3]];
166 tk[0][0] ^= rcon[rconpointer++];
168 for(j = 1; j < KC; j++){
169 if(KC != 8 || j != KC>>1)
170 for(i = 0; i < 4; i++) tk[j][i] ^= tk[j-1][i];
172 for(i = 0; i < 4; i++) tk[j][i] ^= sbox[tk[j-1][i]];
177 for(i=1; i<rounds; i++){
178 subshift(a->round_key[i][0], a->round_key[i][0]+3, a->round_key[i][0]+2, a->round_key[i][0]+1, sbox);
179 mix2(a->round_key[i], dec_multbl, 1, 3);
182 for(i=0; i<(rounds+1)>>1; i++){
184 FFSWAP(int, a->round_key[i][0][j], a->round_key[rounds-i][0][j]);
197 uint8_t rkey[2][16]= {
199 {0x10, 0xa5, 0x88, 0x69, 0xd7, 0x4b, 0xe5, 0xa3, 0x74, 0xcf, 0x86, 0x7c, 0xfb, 0x47, 0x38, 0x59}};
200 uint8_t pt[16], rpt[2][16]= {
201 {0x6a, 0x84, 0x86, 0x7c, 0xd7, 0x7e, 0x12, 0xad, 0x07, 0xea, 0x1b, 0xe8, 0x95, 0xc5, 0x3f, 0xa3},
203 uint8_t rct[2][16]= {
204 {0x73, 0x22, 0x81, 0xc0, 0xa0, 0xaa, 0xb8, 0xf7, 0xa5, 0x4a, 0x0c, 0x67, 0xa0, 0xc4, 0x5e, 0xcf},
205 {0x6d, 0x25, 0x1e, 0x69, 0x44, 0xb0, 0x51, 0xe0, 0x4e, 0xaa, 0x6f, 0xb4, 0xdb, 0xf7, 0x84, 0x65}};
207 av_aes_init(&ae, "PI=3.141592654..", 128, 0);
208 av_aes_init(&ad, "PI=3.141592654..", 128, 1);
209 av_log_level= AV_LOG_DEBUG;
212 av_aes_init(&b, rkey[i], 128, 1);
213 memcpy(b.state, rct[i], 16);
216 if(rpt[i][j] != b.state[0][j])
217 av_log(NULL, AV_LOG_ERROR, "%d %02X %02X\n", j, rpt[i][j], b.state[0][j]);
220 for(i=0; i<10000; i++){
224 memcpy(ae.state, pt, 16);
228 av_log(NULL, AV_LOG_ERROR, "%02X %02X %02X %02X\n", ae.state[0][0], ae.state[1][1], ae.state[2][2], ae.state[3][3]);
229 memcpy(ad.state, ae.state, 16);
233 if(pt[j] != ad.state[0][j]){
234 av_log(NULL, AV_LOG_ERROR, "%d %d %02X %02X\n", i,j, pt[j], ad.state[0][j]);