3 This file is part of the AVR-Crypto-Lib.
4 Copyright (C) 2009 Daniel Otte (daniel.otte@rub.de)
6 This program is free software: you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation, either version 3 of the License, or
9 (at your option) any later version.
11 This program 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
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with this program. If not, see <http://www.gnu.org/licenses/>.
22 * \email daniel.otte@rub.de
24 * \license GPLv3 or later
30 #include <avr/pgmspace.h>
31 #include "bmw_small.h"
34 #define SHL32(a,n) ((a)<<(n))
35 #define SHR32(a,n) ((a)>>(n))
36 #define ROTL32(a,n) (((a)<<(n))|((a)>>(32-(n))))
37 #define ROTR32(a,n) (((a)>>(n))|((a)<<(32-(n))))
46 void ctx_dump(const bmw_small_ctx_t* ctx){
48 cli_putstr_P(PSTR("\r\n==== ctx dump ===="));
50 cli_putstr_P(PSTR("\r\n h["));
52 cli_putstr_P(PSTR("] = "));
53 cli_hexdump_rev(&(ctx->h[i]), 4);
55 cli_putstr_P(PSTR("\r\n counter = "));
56 cli_hexdump(&(ctx->counter), 4);
59 void dump_x(const uint32_t* q, uint8_t elements, char x){
61 cli_putstr_P(PSTR("\r\n==== "));
63 cli_putstr_P(PSTR(" dump ===="));
64 for(i=0; i<elements;++i){
65 cli_putstr_P(PSTR("\r\n "));
67 cli_putstr_P(PSTR("["));
69 cli_putstr_P(PSTR("] = "));
70 cli_hexdump_rev(&(q[i]), 4);
78 uint32_t bmw_small_s0(uint32_t x){
87 uint32_t bmw_small_s1(uint32_t x){
96 uint32_t bmw_small_s2(uint32_t x){
105 uint32_t bmw_small_s3(uint32_t x){
114 uint32_t bmw_small_s4(uint32_t x){
121 uint32_t bmw_small_s5(uint32_t x){
128 uint32_t bmw_small_r1(uint32_t x){
134 uint32_t bmw_small_r2(uint32_t x){
140 uint32_t bmw_small_r3(uint32_t x){
146 uint32_t bmw_small_r4(uint32_t x){
152 uint32_t bmw_small_r5(uint32_t x){
158 uint32_t bmw_small_r6(uint32_t x){
164 uint32_t bmw_small_r7(uint32_t x){
170 #define K 0x05555555L
172 uint32_t k_lut[] PROGMEM = {
173 16L*K, 17L*K, 18L*K, 19L*K, 20L*K, 21L*K, 22L*K, 23L*K,
174 24L*K, 25L*K, 26L*K, 27L*K, 28L*K, 29L*K, 30L*K, 31L*K
177 /* same as above but precomputed to avoid compiler warnings */
180 uint32_t k_lut[] PROGMEM = {
181 0x55555550L, 0x5aaaaaa5L, 0x5ffffffaL,
182 0x6555554fL, 0x6aaaaaa4L, 0x6ffffff9L,
183 0x7555554eL, 0x7aaaaaa3L, 0x7ffffff8L,
184 0x8555554dL, 0x8aaaaaa2L, 0x8ffffff7L,
185 0x9555554cL, 0x9aaaaaa1L, 0x9ffffff6L,
189 uint32_t bmw_small_expand1(uint8_t j, const uint32_t* q, const void* m){
190 uint32_t(*s[])(uint32_t) = {bmw_small_s1, bmw_small_s2, bmw_small_s3, bmw_small_s0};
193 /* r = 0x05555555*(j+16); */
194 r = pgm_read_dword(k_lut+j);
198 r += ((uint32_t*)m)[j];
199 r += ((uint32_t*)m)[j+3];
200 r -= ((uint32_t*)m)[j+10];
204 uint32_t bmw_small_expand2(uint8_t j, const uint32_t* q, const void* m){
205 uint32_t(*rf[])(uint32_t) = {bmw_small_r1, bmw_small_r2, bmw_small_r3,
206 bmw_small_r4, bmw_small_r5, bmw_small_r6,
210 /* r = 0x05555555*(j+16); */
211 r = pgm_read_dword(k_lut+j);
212 for(i=0; i<14; i+=2){
215 for(i=0; i<14; i+=2){
216 r += rf[i/2](q[j+i+1]);
218 r += bmw_small_s5(q[j+14]);
219 r += bmw_small_s4(q[j+15]);
220 r += ((uint32_t*)m)[j];
221 r += ((uint32_t*)m)[(j+3)%16];
222 r -= ((uint32_t*)m)[(j+10)%16];
227 uint8_t f0_lut[] PROGMEM = {
228 5<<1, ( 7<<1)+1, (10<<1)+0, (13<<1)+0, (14<<1)+0,
229 6<<1, ( 8<<1)+1, (11<<1)+0, (14<<1)+0, (15<<1)+1,
230 0<<1, ( 7<<1)+0, ( 9<<1)+0, (12<<1)+1, (15<<1)+0,
231 0<<1, ( 1<<1)+1, ( 8<<1)+0, (10<<1)+1, (13<<1)+0,
232 1<<1, ( 2<<1)+0, ( 9<<1)+0, (11<<1)+1, (14<<1)+1,
233 3<<1, ( 2<<1)+1, (10<<1)+0, (12<<1)+1, (15<<1)+0,
234 4<<1, ( 0<<1)+1, ( 3<<1)+1, (11<<1)+1, (13<<1)+0,
235 1<<1, ( 4<<1)+1, ( 5<<1)+1, (12<<1)+1, (14<<1)+1,
236 2<<1, ( 5<<1)+1, ( 6<<1)+1, (13<<1)+0, (15<<1)+1,
237 0<<1, ( 3<<1)+1, ( 6<<1)+0, ( 7<<1)+1, (14<<1)+0,
238 8<<1, ( 1<<1)+1, ( 4<<1)+1, ( 7<<1)+1, (15<<1)+0,
239 8<<1, ( 0<<1)+1, ( 2<<1)+1, ( 5<<1)+1, ( 9<<1)+0,
240 1<<1, ( 3<<1)+0, ( 6<<1)+1, ( 9<<1)+1, (10<<1)+0,
241 2<<1, ( 4<<1)+0, ( 7<<1)+0, (10<<1)+0, (11<<1)+0,
242 3<<1, ( 5<<1)+1, ( 8<<1)+0, (11<<1)+1, (12<<1)+1,
243 12<<1, ( 4<<1)+1, ( 6<<1)+1, ( 9<<1)+1, (13<<1)+0
246 void bmw_small_f0(uint32_t* q, uint32_t* h, const void* m){
247 uint8_t i,j=-1,v,sign,l=0;
248 uint32_t(*s[])(uint32_t)={ bmw_small_s0, bmw_small_s1, bmw_small_s2,
249 bmw_small_s3, bmw_small_s4 };
251 h[i] ^= ((uint32_t*)m)[i];
254 // memset(q, 0, 4*16);
255 for(i=0; i<5*16; ++i){
256 v = pgm_read_byte(f0_lut+i);
278 void bmw_small_f0(uint32_t* q, uint32_t* h, const void* m){
280 uint32_t(*s[])(uint32_t)={ bmw_small_s0, bmw_small_s1, bmw_small_s2,
281 bmw_small_s3, bmw_small_s4 };
283 h[i] ^= ((uint32_t*)m)[i];
286 q[ 0] = (h[ 5] - h[ 7] + h[10] + h[13] + h[14]);
287 q[ 1] = (h[ 6] - h[ 8] + h[11] + h[14] - h[15]);
288 q[ 2] = (h[ 0] + h[ 7] + h[ 9] - h[12] + h[15]);
289 q[ 3] = (h[ 0] - h[ 1] + h[ 8] - h[10] + h[13]);
290 q[ 4] = (h[ 1] + h[ 2] + h[ 9] - h[11] - h[14]);
291 q[ 5] = (h[ 3] - h[ 2] + h[10] - h[12] + h[15]);
292 q[ 6] = (h[ 4] - h[ 0] - h[ 3] - h[11] + h[13]);
293 q[ 7] = (h[ 1] - h[ 4] - h[ 5] - h[12] - h[14]);
294 q[ 8] = (h[ 2] - h[ 5] - h[ 6] + h[13] - h[15]);
295 q[ 9] = (h[ 0] - h[ 3] + h[ 6] - h[ 7] + h[14]);
296 q[10] = (h[ 8] - h[ 1] - h[ 4] - h[ 7] + h[15]);
297 q[11] = (h[ 8] - h[ 0] - h[ 2] - h[ 5] + h[ 9]);
298 q[12] = (h[ 1] + h[ 3] - h[ 6] - h[ 9] + h[10]);
299 q[13] = (h[ 2] + h[ 4] + h[ 7] + h[10] + h[11]);
300 q[14] = (h[ 3] - h[ 5] + h[ 8] - h[11] - h[12]);
301 q[15] = (h[12] - h[ 4] - h[ 6] - h[ 9] + h[13]);
309 void bmw_small_f1(uint32_t* q, const void* m){
311 q[16] = bmw_small_expand1(0, q, m);
312 q[17] = bmw_small_expand1(1, q, m);
314 q[16+i] = bmw_small_expand2(i, q, m);
318 void bmw_small_f2(uint32_t* h, const uint32_t* q, const void* m){
329 cli_putstr_P(PSTR("\r\n XL = "));
330 cli_hexdump_rev(&xl, 4);
331 cli_putstr_P(PSTR("\r\n XH = "));
332 cli_hexdump_rev(&xh, 4);
335 h[0] ^= SHL32(xh, 5) ^ SHR32(q[16], 5);
336 h[1] ^= SHR32(xh, 7) ^ SHL32(q[17], 8);
337 h[2] ^= SHR32(xh, 5) ^ SHL32(q[18], 5);
338 h[3] ^= SHR32(xh, 1) ^ SHL32(q[19], 5);
339 h[4] ^= SHR32(xh, 3) ^ q[20];
340 h[5] ^= SHL32(xh, 6) ^ SHR32(q[21], 6);
341 h[6] ^= SHR32(xh, 4) ^ SHL32(q[22], 6);
342 h[7] ^= SHR32(xh,11) ^ SHL32(q[23], 2);
344 h[i] += xl ^ q[24+i] ^ q[i];
347 h[8+i] ^= xh ^ q[24+i];
348 h[8+i] += ROTL32(h[(4+i)%8],i+9);
350 h[ 8] += SHL32(xl, 8) ^ q[23] ^ q[ 8];
351 h[ 9] += SHR32(xl, 6) ^ q[16] ^ q[ 9];
352 h[10] += SHL32(xl, 6) ^ q[17] ^ q[10];
353 h[11] += SHL32(xl, 4) ^ q[18] ^ q[11];
354 h[12] += SHR32(xl, 3) ^ q[19] ^ q[12];
355 h[13] += SHR32(xl, 4) ^ q[20] ^ q[13];
356 h[14] += SHR32(xl, 7) ^ q[21] ^ q[14];
357 h[15] += SHR32(xl, 2) ^ q[22] ^ q[15];
360 void bmw_small_nextBlock(bmw_small_ctx_t* ctx, const void* block){
362 dump_x(block, 16, 'M');
363 bmw_small_f0(q, ctx->h, block);
365 bmw_small_f1(q, block);
367 bmw_small_f2(ctx->h, q, block);
372 void bmw_small_lastBlock(bmw_small_ctx_t* ctx, const void* block, uint16_t length_b){
374 while(length_b >= BMW_SMALL_BLOCKSIZE){
375 bmw_small_nextBlock(ctx, block);
376 length_b -= BMW_SMALL_BLOCKSIZE;
377 block = (uint8_t*)block + BMW_SMALL_BLOCKSIZE_B;
379 memset(buffer, 0, 64);
380 memcpy(buffer, block, (length_b+7)/8);
381 buffer[length_b>>3] |= 0x80 >> (length_b&0x07);
382 if(length_b+1>64*8-64){
383 bmw_small_nextBlock(ctx, buffer);
384 memset(buffer, 0, 64-8);
387 *((uint64_t*)&(buffer[64-8])) = (uint64_t)(ctx->counter*512LL)+(uint64_t)length_b;
388 bmw_small_nextBlock(ctx, buffer);
391 void bmw224_init(bmw224_ctx_t* ctx){
393 ctx->h[0] = 0x00010203;
395 ctx->h[i] = ctx->h[i-1]+ 0x04040404;
398 ctx->h[13] = 0x24353637;
404 void bmw256_init(bmw256_ctx_t* ctx){
406 ctx->h[0] = 0x40414243;
408 ctx->h[i] = ctx->h[i-1]+ 0x04040404;
414 void bmw224_nextBlock(bmw224_ctx_t* ctx, const void* block){
415 bmw_small_nextBlock(ctx, block);
418 void bmw256_nextBlock(bmw256_ctx_t* ctx, const void* block){
419 bmw_small_nextBlock(ctx, block);
422 void bmw224_lastBlock(bmw224_ctx_t* ctx, const void* block, uint16_t length_b){
423 bmw_small_lastBlock(ctx, block, length_b);
426 void bmw256_lastBlock(bmw256_ctx_t* ctx, const void* block, uint16_t length_b){
427 bmw_small_lastBlock(ctx, block, length_b);
430 void bmw224_ctx2hash(void* dest, const bmw224_ctx_t* ctx){
431 memcpy(dest, &(ctx->h[9]), 224/8);
434 void bmw256_ctx2hash(void* dest, const bmw256_ctx_t* ctx){
435 memcpy(dest, &(ctx->h[8]), 256/8);
438 void bmw224(void* dest, const void* msg, uint32_t length_b){
441 while(length_b>=BMW_SMALL_BLOCKSIZE){
442 bmw_small_nextBlock(&ctx, msg);
443 length_b -= BMW_SMALL_BLOCKSIZE;
444 msg = (uint8_t*)msg + BMW_SMALL_BLOCKSIZE_B;
446 bmw_small_lastBlock(&ctx, msg, length_b);
447 bmw224_ctx2hash(dest, &ctx);
450 void bmw256(void* dest, const void* msg, uint32_t length_b){
453 while(length_b>=BMW_SMALL_BLOCKSIZE){
454 bmw_small_nextBlock(&ctx, msg);
455 length_b -= BMW_SMALL_BLOCKSIZE;
456 msg = (uint8_t*)msg + BMW_SMALL_BLOCKSIZE_B;
458 bmw_small_lastBlock(&ctx, msg, length_b);
459 bmw256_ctx2hash(dest, &ctx);