3 This file is part of the ARM-Crypto-Lib.
4 Copyright (C) 2006-2010 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 "bmw_small.h"
33 #define SHL32(a,n) ((a)<<(n))
34 #define SHR32(a,n) ((a)>>(n))
35 #define ROTL32(a,n) (((a)<<(n))|((a)>>(32-(n))))
36 #define ROTR32(a,n) (((a)>>(n))|((a)<<(32-(n))))
45 void ctx_dump(const bmw_small_ctx_t* ctx){
47 cli_putstr("\r\n==== ctx dump ====");
49 cli_putstr("\r\n h[");
52 cli_hexdump_rev(&(ctx->h[i]), 4);
54 cli_putstr("\r\n counter = ");
55 cli_hexdump(&(ctx->counter), 4);
58 void dump_x(const uint32_t* q, uint8_t elements, char x){
60 cli_putstr("\r\n==== ");
62 cli_putstr(" dump ====");
63 for(i=0; i<elements;++i){
69 cli_hexdump_rev(&(q[i]), 4);
77 #define S32_0(x) ( (SHR32((x), 1)) ^ \
82 #define S32_1(x) ( (SHR32((x), 1)) ^ \
87 #define S32_2(x) ( (SHR32((x), 2)) ^ \
92 #define S32_3(x) ( (SHR32((x), 2)) ^ \
97 #define S32_4(x) ( (SHR32((x), 1)) ^ (x))
99 #define S32_5(x) ( (SHR32((x), 2)) ^ (x))
101 #define R32_1(x) (ROTL32((x), 3))
102 #define R32_2(x) (ROTL32((x), 7))
103 #define R32_3(x) (ROTL32((x), 13))
104 #define R32_4(x) (ROTL32((x), 16))
105 #define R32_5(x) (ROTR32((x), 13))
106 #define R32_6(x) (ROTR32((x), 9))
107 #define R32_7(x) (ROTR32((x), 5))
110 #include "f1_autogen.c"
113 void bmw_small_f0(uint32_t* q, uint32_t* h, const uint32_t* m){
132 q[ 0] = (h[ 5] - h[ 7] + h[10] + h[13] + h[14]);
133 q[ 3] = (h[ 0] - h[ 1] + h[ 8] - h[10] + h[13]);
134 q[ 6] = (h[ 4] - h[ 0] - h[ 3] - h[11] + h[13]);
135 q[ 9] = (h[ 0] - h[ 3] + h[ 6] - h[ 7] + h[14]);
136 q[12] = (h[ 1] + h[ 3] - h[ 6] - h[ 9] + h[10]);
137 q[15] = (h[12] - h[ 4] - h[ 6] - h[ 9] + h[13]);
138 q[ 2] = (h[ 0] + h[ 7] + h[ 9] - h[12] + h[15]);
139 q[ 5] = (h[ 3] - h[ 2] + h[10] - h[12] + h[15]);
140 q[ 8] = (h[ 2] - h[ 5] - h[ 6] + h[13] - h[15]);
141 q[11] = (h[ 8] - h[ 0] - h[ 2] - h[ 5] + h[ 9]);
142 q[14] = (h[ 3] - h[ 5] + h[ 8] - h[11] - h[12]);
143 q[ 1] = (h[ 6] - h[ 8] + h[11] + h[14] - h[15]);
144 q[ 4] = (h[ 1] + h[ 2] + h[ 9] - h[11] - h[14]);
145 q[ 7] = (h[ 1] - h[ 4] - h[ 5] - h[12] - h[14]);
146 q[10] = (h[ 8] - h[ 1] - h[ 4] - h[ 7] + h[15]);
147 q[13] = (h[ 2] + h[ 4] + h[ 7] + h[10] + h[11]);
149 q[ 0] = S32_0(q[ 0]) + (h[ 1] ^= m[ 1]);
150 q[ 1] = S32_1(q[ 1]) + (h[ 2] ^= m[ 2]);
151 q[ 2] = S32_2(q[ 2]) + (h[ 3] ^= m[ 3]);
152 q[ 3] = S32_3(q[ 3]) + (h[ 4] ^= m[ 4]);
153 q[ 4] = S32_4(q[ 4]) + (h[ 5] ^= m[ 5]);
154 q[ 5] = S32_0(q[ 5]) + (h[ 6] ^= m[ 6]);
155 q[ 6] = S32_1(q[ 6]) + (h[ 7] ^= m[ 7]);
156 q[ 7] = S32_2(q[ 7]) + (h[ 8] ^= m[ 8]);
157 q[ 8] = S32_3(q[ 8]) + (h[ 9] ^= m[ 9]);
158 q[ 9] = S32_4(q[ 9]) + (h[10] ^= m[10]);
159 q[10] = S32_0(q[10]) + (h[11] ^= m[11]);
160 q[11] = S32_1(q[11]) + (h[12] ^= m[12]);
161 q[12] = S32_2(q[12]) + (h[13] ^= m[13]);
162 q[13] = S32_3(q[13]) + (h[14] ^= m[14]);
163 q[14] = S32_4(q[14]) + (h[15] ^= m[15]);
164 q[15] = S32_0(q[15]) + (h[ 0] ^= m[ 0]);
168 void bmw_small_f2(uint32_t* h, uint32_t* q, const uint32_t* m){
170 xl = q[16] ^ q[17] ^ q[18] ^ q[19] ^ q[20] ^ q[21] ^ q[22] ^ q[23];
171 xh = xl ^ q[24] ^ q[25] ^ q[26] ^ q[27] ^ q[28] ^ q[29] ^ q[30] ^ q[31];
173 cli_putstr("\r\n XL = ");
174 cli_hexdump_rev(&xl, 4);
175 cli_putstr("\r\n XH = ");
176 cli_hexdump_rev(&xh, 4);
179 h[0] = (SHL32(xh, 5) ^ SHR32(q[16], 5) ^ m[ 0]) + (xl ^ q[24] ^ q[ 0]);
180 h[1] = (SHR32(xh, 7) ^ SHL32(q[17], 8) ^ m[ 1]) + (xl ^ q[25] ^ q[ 1]);
181 h[2] = (SHR32(xh, 5) ^ SHL32(q[18], 5) ^ m[ 2]) + (xl ^ q[26] ^ q[ 2]);
182 h[3] = (SHR32(xh, 1) ^ SHL32(q[19], 5) ^ m[ 3]) + (xl ^ q[27] ^ q[ 3]);
183 h[4] = (SHR32(xh, 3) ^ q[20] ^ m[ 4]) + (xl ^ q[28] ^ q[ 4]);
184 h[5] = (SHL32(xh, 6) ^ SHR32(q[21], 6) ^ m[ 5]) + (xl ^ q[29] ^ q[ 5]);
185 h[6] = (SHR32(xh, 4) ^ SHL32(q[22], 6) ^ m[ 6]) + (xl ^ q[30] ^ q[ 6]);
186 h[7] = (SHR32(xh,11) ^ SHL32(q[23], 2) ^ m[ 7]) + (xl ^ q[31] ^ q[ 7]);
188 h[ 8] = ROTL32(h[4], 9) + (xh ^ q[24] ^ m[ 8]) + (SHL32(xl, 8) ^ q[23] ^ q[ 8]);
189 h[ 9] = ROTL32(h[5], 10) + (xh ^ q[25] ^ m[ 9]) + (SHR32(xl, 6) ^ q[16] ^ q[ 9]);
190 h[10] = ROTL32(h[6], 11) + (xh ^ q[26] ^ m[10]) + (SHL32(xl, 6) ^ q[17] ^ q[10]);
191 h[11] = ROTL32(h[7], 12) + (xh ^ q[27] ^ m[11]) + (SHL32(xl, 4) ^ q[18] ^ q[11]);
192 h[12] = ROTL32(h[0], 13) + (xh ^ q[28] ^ m[12]) + (SHR32(xl, 3) ^ q[19] ^ q[12]);
193 h[13] = ROTL32(h[1], 14) + (xh ^ q[29] ^ m[13]) + (SHR32(xl, 4) ^ q[20] ^ q[13]);
194 h[14] = ROTL32(h[2], 15) + (xh ^ q[30] ^ m[14]) + (SHR32(xl, 7) ^ q[21] ^ q[14]);
195 h[15] = ROTL32(h[3], 16) + (xh ^ q[31] ^ m[15]) + (SHR32(xl, 2) ^ q[22] ^ q[15]);
198 void bmw_small_nextBlock(bmw_small_ctx_t* ctx, const void* block){
200 dump_x(block, 16, 'M');
201 bmw_small_f0(q, ctx->h, block);
203 bmw_small_f1(q, block, ctx->h);
204 dump_x(q+16, 16, 'Q');
205 bmw_small_f2(ctx->h, q, block);
210 void bmw_small_lastBlock(bmw_small_ctx_t* ctx, const void* block, uint16_t length_b){
212 while(length_b >= BMW_SMALL_BLOCKSIZE){
213 bmw_small_nextBlock(ctx, block);
214 length_b -= BMW_SMALL_BLOCKSIZE;
215 block = (uint8_t*)block + BMW_SMALL_BLOCKSIZE_B;
217 memset(buffer, 0, 64);
218 memcpy(buffer, block, (length_b+7)/8);
219 buffer[length_b>>3] |= 0x80 >> (length_b&0x07);
220 if(length_b+1>64*8-64){
221 bmw_small_nextBlock(ctx, buffer);
222 memset(buffer, 0, 64-8);
225 *((uint64_t*)&(buffer[64-8])) = (uint64_t)(ctx->counter*512LL)+(uint64_t)length_b;
226 bmw_small_nextBlock(ctx, buffer);
229 memset(buffer, 0xaa, 64);
231 buffer[i*4] = i+0xa0;
233 // dump_x(buffer, 16, 'A');
234 dump_x(ctx->h, 16, 'M');
235 bmw_small_f0(q, (uint32_t*)buffer, ctx->h);
236 dump_x(buffer, 16, 'a');
238 bmw_small_f1(q, ctx->h, (uint32_t*)buffer);
240 bmw_small_f2((uint32_t*)buffer, q, ctx->h);
241 memcpy(ctx->h, buffer, 64);
244 void bmw224_init(bmw224_ctx_t* ctx){
246 ctx->h[0] = 0x00010203;
248 ctx->h[i] = ctx->h[i-1]+ 0x04040404;
254 void bmw256_init(bmw256_ctx_t* ctx){
256 ctx->h[0] = 0x40414243;
258 ctx->h[i] = ctx->h[i-1]+ 0x04040404;
264 void bmw224_nextBlock(bmw224_ctx_t* ctx, const void* block){
265 bmw_small_nextBlock(ctx, block);
268 void bmw256_nextBlock(bmw256_ctx_t* ctx, const void* block){
269 bmw_small_nextBlock(ctx, block);
272 void bmw224_lastBlock(bmw224_ctx_t* ctx, const void* block, uint16_t length_b){
273 bmw_small_lastBlock(ctx, block, length_b);
276 void bmw256_lastBlock(bmw256_ctx_t* ctx, const void* block, uint16_t length_b){
277 bmw_small_lastBlock(ctx, block, length_b);
280 void bmw224_ctx2hash(void* dest, const bmw224_ctx_t* ctx){
281 memcpy(dest, &(ctx->h[9]), 224/8);
284 void bmw256_ctx2hash(void* dest, const bmw256_ctx_t* ctx){
285 memcpy(dest, &(ctx->h[8]), 256/8);
288 void bmw224(void* dest, const void* msg, uint32_t length_b){
291 while(length_b>=BMW_SMALL_BLOCKSIZE){
292 bmw_small_nextBlock(&ctx, msg);
293 length_b -= BMW_SMALL_BLOCKSIZE;
294 msg = (uint8_t*)msg + BMW_SMALL_BLOCKSIZE_B;
296 bmw_small_lastBlock(&ctx, msg, length_b);
297 bmw224_ctx2hash(dest, &ctx);
300 void bmw256(void* dest, const void* msg, uint32_t length_b){
303 while(length_b>=BMW_SMALL_BLOCKSIZE){
304 bmw_small_nextBlock(&ctx, msg);
305 length_b -= BMW_SMALL_BLOCKSIZE;
306 msg = (uint8_t*)msg + BMW_SMALL_BLOCKSIZE_B;
308 bmw_small_lastBlock(&ctx, msg, length_b);
309 bmw256_ctx2hash(dest, &ctx);