aes.c 8.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255
  1. /*
  2. * copyright (c) 2007 Michael Niedermayer <michaelni@gmx.at>
  3. *
  4. * some optimization ideas from aes128.c by Reimar Doeffinger
  5. *
  6. * This file is part of FFmpeg.
  7. *
  8. * FFmpeg is free software; you can redistribute it and/or
  9. * modify it under the terms of the GNU Lesser General Public
  10. * License as published by the Free Software Foundation; either
  11. * version 2.1 of the License, or (at your option) any later version.
  12. *
  13. * FFmpeg is distributed in the hope that it will be useful,
  14. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  16. * Lesser General Public License for more details.
  17. *
  18. * You should have received a copy of the GNU Lesser General Public
  19. * License along with FFmpeg; if not, write to the Free Software
  20. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  21. */
  22. #include "common.h"
  23. #include "aes.h"
  24. typedef union {
  25. uint64_t u64[2];
  26. uint32_t u32[4];
  27. uint8_t u8x4[4][4];
  28. uint8_t u8[16];
  29. } av_aes_block;
  30. typedef struct AVAES{
  31. // Note: round_key[16] is accessed in the init code, but this only
  32. // overwrites state, which does not matter (see also r7471).
  33. av_aes_block round_key[15];
  34. av_aes_block state[2];
  35. int rounds;
  36. }AVAES;
  37. const int av_aes_size= sizeof(AVAES);
  38. static const uint8_t rcon[10] = {
  39. 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36
  40. };
  41. static uint8_t sbox[256];
  42. static uint8_t inv_sbox[256];
  43. #if CONFIG_SMALL
  44. static uint32_t enc_multbl[1][256];
  45. static uint32_t dec_multbl[1][256];
  46. #else
  47. static uint32_t enc_multbl[4][256];
  48. static uint32_t dec_multbl[4][256];
  49. #endif
  50. static inline void addkey(av_aes_block *dst, const av_aes_block *src, const av_aes_block *round_key){
  51. dst->u64[0] = src->u64[0] ^ round_key->u64[0];
  52. dst->u64[1] = src->u64[1] ^ round_key->u64[1];
  53. }
  54. static void subshift(av_aes_block s0[2], int s, const uint8_t *box){
  55. av_aes_block *s1= (av_aes_block *)(s0[0].u8 - s);
  56. av_aes_block *s3= (av_aes_block *)(s0[0].u8 + s);
  57. s0[0].u8[0]=box[s0[1].u8[ 0]]; s0[0].u8[ 4]=box[s0[1].u8[ 4]]; s0[0].u8[ 8]=box[s0[1].u8[ 8]]; s0[0].u8[12]=box[s0[1].u8[12]];
  58. s1[0].u8[3]=box[s1[1].u8[ 7]]; s1[0].u8[ 7]=box[s1[1].u8[11]]; s1[0].u8[11]=box[s1[1].u8[15]]; s1[0].u8[15]=box[s1[1].u8[ 3]];
  59. s0[0].u8[2]=box[s0[1].u8[10]]; s0[0].u8[10]=box[s0[1].u8[ 2]]; s0[0].u8[ 6]=box[s0[1].u8[14]]; s0[0].u8[14]=box[s0[1].u8[ 6]];
  60. s3[0].u8[1]=box[s3[1].u8[13]]; s3[0].u8[13]=box[s3[1].u8[ 9]]; s3[0].u8[ 9]=box[s3[1].u8[ 5]]; s3[0].u8[ 5]=box[s3[1].u8[ 1]];
  61. }
  62. static inline int mix_core(uint32_t multbl[][256], int a, int b, int c, int d){
  63. #if CONFIG_SMALL
  64. #define ROT(x,s) ((x<<s)|(x>>(32-s)))
  65. return multbl[0][a] ^ ROT(multbl[0][b], 8) ^ ROT(multbl[0][c], 16) ^ ROT(multbl[0][d], 24);
  66. #else
  67. return multbl[0][a] ^ multbl[1][b] ^ multbl[2][c] ^ multbl[3][d];
  68. #endif
  69. }
  70. static inline void mix(av_aes_block state[2], uint32_t multbl[][256], int s1, int s3){
  71. uint8_t (*src)[4] = state[1].u8x4;
  72. state[0].u32[0] = mix_core(multbl, src[0][0], src[s1 ][1], src[2][2], src[s3 ][3]);
  73. state[0].u32[1] = mix_core(multbl, src[1][0], src[s3-1][1], src[3][2], src[s1-1][3]);
  74. state[0].u32[2] = mix_core(multbl, src[2][0], src[s3 ][1], src[0][2], src[s1 ][3]);
  75. state[0].u32[3] = mix_core(multbl, src[3][0], src[s1-1][1], src[1][2], src[s3-1][3]);
  76. }
  77. static inline void crypt(AVAES *a, int s, const uint8_t *sbox, uint32_t multbl[][256]){
  78. int r;
  79. for(r=a->rounds-1; r>0; r--){
  80. mix(a->state, multbl, 3-s, 1+s);
  81. addkey(&a->state[1], &a->state[0], &a->round_key[r]);
  82. }
  83. subshift(&a->state[0], s, sbox);
  84. }
  85. void av_aes_crypt(AVAES *a, uint8_t *dst_, const uint8_t *src_, int count, uint8_t *iv_, int decrypt){
  86. av_aes_block *dst = (av_aes_block *)dst_;
  87. const av_aes_block *src = (const av_aes_block *)src_;
  88. av_aes_block *iv = (av_aes_block *)iv_;
  89. while(count--){
  90. addkey(&a->state[1], src, &a->round_key[a->rounds]);
  91. if(decrypt) {
  92. crypt(a, 0, inv_sbox, dec_multbl);
  93. if(iv){
  94. addkey(&a->state[0], &a->state[0], iv);
  95. memcpy(iv, src, 16);
  96. }
  97. addkey(dst, &a->state[0], &a->round_key[0]);
  98. }else{
  99. if(iv) addkey(&a->state[1], &a->state[1], iv);
  100. crypt(a, 2, sbox, enc_multbl);
  101. addkey(dst, &a->state[0], &a->round_key[0]);
  102. if(iv) memcpy(iv, dst, 16);
  103. }
  104. src++;
  105. dst++;
  106. }
  107. }
  108. static void init_multbl2(uint8_t tbl[1024], const int c[4], const uint8_t *log8, const uint8_t *alog8, const uint8_t *sbox){
  109. int i, j;
  110. for(i=0; i<1024; i++){
  111. int x= sbox[i>>2];
  112. if(x) tbl[i]= alog8[ log8[x] + log8[c[i&3]] ];
  113. }
  114. #if !CONFIG_SMALL
  115. for(j=256; j<1024; j++)
  116. for(i=0; i<4; i++)
  117. tbl[4*j+i]= tbl[4*j + ((i-1)&3) - 1024];
  118. #endif
  119. }
  120. // this is based on the reference AES code by Paulo Barreto and Vincent Rijmen
  121. int av_aes_init(AVAES *a, const uint8_t *key, int key_bits, int decrypt) {
  122. int i, j, t, rconpointer = 0;
  123. uint8_t tk[8][4];
  124. int KC= key_bits>>5;
  125. int rounds= KC + 6;
  126. uint8_t log8[256];
  127. uint8_t alog8[512];
  128. if(!enc_multbl[FF_ARRAY_ELEMS(enc_multbl)-1][FF_ARRAY_ELEMS(enc_multbl[0])-1]){
  129. j=1;
  130. for(i=0; i<255; i++){
  131. alog8[i]=
  132. alog8[i+255]= j;
  133. log8[j]= i;
  134. j^= j+j;
  135. if(j>255) j^= 0x11B;
  136. }
  137. for(i=0; i<256; i++){
  138. j= i ? alog8[255-log8[i]] : 0;
  139. j ^= (j<<1) ^ (j<<2) ^ (j<<3) ^ (j<<4);
  140. j = (j ^ (j>>8) ^ 99) & 255;
  141. inv_sbox[j]= i;
  142. sbox [i]= j;
  143. }
  144. init_multbl2(dec_multbl[0], (const int[4]){0xe, 0x9, 0xd, 0xb}, log8, alog8, inv_sbox);
  145. init_multbl2(enc_multbl[0], (const int[4]){0x2, 0x1, 0x1, 0x3}, log8, alog8, sbox);
  146. }
  147. if(key_bits!=128 && key_bits!=192 && key_bits!=256)
  148. return -1;
  149. a->rounds= rounds;
  150. memcpy(tk, key, KC*4);
  151. for(t= 0; t < (rounds+1)*16;) {
  152. memcpy(a->round_key[0].u8+t, tk, KC*4);
  153. t+= KC*4;
  154. for(i = 0; i < 4; i++)
  155. tk[0][i] ^= sbox[tk[KC-1][(i+1)&3]];
  156. tk[0][0] ^= rcon[rconpointer++];
  157. for(j = 1; j < KC; j++){
  158. if(KC != 8 || j != KC>>1)
  159. for(i = 0; i < 4; i++) tk[j][i] ^= tk[j-1][i];
  160. else
  161. for(i = 0; i < 4; i++) tk[j][i] ^= sbox[tk[j-1][i]];
  162. }
  163. }
  164. if(decrypt){
  165. for(i=1; i<rounds; i++){
  166. av_aes_block tmp[3];
  167. memcpy(&tmp[2], &a->round_key[i], 16);
  168. subshift(&tmp[1], 0, sbox);
  169. mix(tmp, dec_multbl, 1, 3);
  170. memcpy(&a->round_key[i], &tmp[0], 16);
  171. }
  172. }else{
  173. for(i=0; i<(rounds+1)>>1; i++){
  174. for(j=0; j<16; j++)
  175. FFSWAP(int, a->round_key[i].u8[j], a->round_key[rounds-i].u8[j]);
  176. }
  177. }
  178. return 0;
  179. }
  180. #ifdef TEST
  181. #include "lfg.h"
  182. #include "log.h"
  183. int main(void){
  184. int i,j;
  185. AVAES ae, ad, b;
  186. uint8_t rkey[2][16]= {
  187. {0},
  188. {0x10, 0xa5, 0x88, 0x69, 0xd7, 0x4b, 0xe5, 0xa3, 0x74, 0xcf, 0x86, 0x7c, 0xfb, 0x47, 0x38, 0x59}};
  189. uint8_t pt[16], rpt[2][16]= {
  190. {0x6a, 0x84, 0x86, 0x7c, 0xd7, 0x7e, 0x12, 0xad, 0x07, 0xea, 0x1b, 0xe8, 0x95, 0xc5, 0x3f, 0xa3},
  191. {0}};
  192. uint8_t rct[2][16]= {
  193. {0x73, 0x22, 0x81, 0xc0, 0xa0, 0xaa, 0xb8, 0xf7, 0xa5, 0x4a, 0x0c, 0x67, 0xa0, 0xc4, 0x5e, 0xcf},
  194. {0x6d, 0x25, 0x1e, 0x69, 0x44, 0xb0, 0x51, 0xe0, 0x4e, 0xaa, 0x6f, 0xb4, 0xdb, 0xf7, 0x84, 0x65}};
  195. uint8_t temp[16];
  196. AVLFG prng;
  197. av_aes_init(&ae, "PI=3.141592654..", 128, 0);
  198. av_aes_init(&ad, "PI=3.141592654..", 128, 1);
  199. av_log_set_level(AV_LOG_DEBUG);
  200. av_lfg_init(&prng, 1);
  201. for(i=0; i<2; i++){
  202. av_aes_init(&b, rkey[i], 128, 1);
  203. av_aes_crypt(&b, temp, rct[i], 1, NULL, 1);
  204. for(j=0; j<16; j++)
  205. if(rpt[i][j] != temp[j])
  206. av_log(NULL, AV_LOG_ERROR, "%d %02X %02X\n", j, rpt[i][j], temp[j]);
  207. }
  208. for(i=0; i<10000; i++){
  209. for(j=0; j<16; j++){
  210. pt[j] = av_lfg_get(&prng);
  211. }
  212. {START_TIMER
  213. av_aes_crypt(&ae, temp, pt, 1, NULL, 0);
  214. if(!(i&(i-1)))
  215. av_log(NULL, AV_LOG_ERROR, "%02X %02X %02X %02X\n", temp[0], temp[5], temp[10], temp[15]);
  216. av_aes_crypt(&ad, temp, temp, 1, NULL, 1);
  217. STOP_TIMER("aes")}
  218. for(j=0; j<16; j++){
  219. if(pt[j] != temp[j]){
  220. av_log(NULL, AV_LOG_ERROR, "%d %d %02X %02X\n", i,j, pt[j], temp[j]);
  221. }
  222. }
  223. }
  224. return 0;
  225. }
  226. #endif