36#ifndef MBEDTLS_BN_MUL_H
37#define MBEDTLS_BN_MUL_H
39#if !defined(MBEDTLS_CONFIG_FILE)
42#include MBEDTLS_CONFIG_FILE
52#if defined(MBEDTLS_HAVE_INT32)
54#define MBEDTLS_BYTES_TO_T_UINT_4(a, b, c, d) \
55 ((mbedtls_mpi_uint) (a) << 0) | \
56 ((mbedtls_mpi_uint) (b) << 8) | \
57 ((mbedtls_mpi_uint) (c) << 16) | \
58 ((mbedtls_mpi_uint) (d) << 24)
60#define MBEDTLS_BYTES_TO_T_UINT_2(a, b) \
61 MBEDTLS_BYTES_TO_T_UINT_4(a, b, 0, 0)
63#define MBEDTLS_BYTES_TO_T_UINT_8(a, b, c, d, e, f, g, h) \
64 MBEDTLS_BYTES_TO_T_UINT_4(a, b, c, d), \
65 MBEDTLS_BYTES_TO_T_UINT_4(e, f, g, h)
69#define MBEDTLS_BYTES_TO_T_UINT_8(a, b, c, d, e, f, g, h) \
70 ((mbedtls_mpi_uint) (a) << 0) | \
71 ((mbedtls_mpi_uint) (b) << 8) | \
72 ((mbedtls_mpi_uint) (c) << 16) | \
73 ((mbedtls_mpi_uint) (d) << 24) | \
74 ((mbedtls_mpi_uint) (e) << 32) | \
75 ((mbedtls_mpi_uint) (f) << 40) | \
76 ((mbedtls_mpi_uint) (g) << 48) | \
77 ((mbedtls_mpi_uint) (h) << 56)
79#define MBEDTLS_BYTES_TO_T_UINT_4(a, b, c, d) \
80 MBEDTLS_BYTES_TO_T_UINT_8(a, b, c, d, 0, 0, 0, 0)
82#define MBEDTLS_BYTES_TO_T_UINT_2(a, b) \
83 MBEDTLS_BYTES_TO_T_UINT_8(a, b, 0, 0, 0, 0, 0, 0)
88#if defined(MBEDTLS_HAVE_ASM)
95#if defined(__GNUC__) && \
96 ( !defined(__ARMCC_VERSION) || __ARMCC_VERSION >= 6000000 )
110#if defined(__GNUC__) && __GNUC__ < 5 && defined(__PIC__)
111#define MULADDC_CANNOT_USE_EBX
120#if defined(__i386__) && defined(__OPTIMIZE__) && !defined(MULADDC_CANNOT_USE_EBX)
122#define MULADDC_INIT \
124 "movl %%ebx, %0 \n\t" \
125 "movl %5, %%esi \n\t" \
126 "movl %6, %%edi \n\t" \
127 "movl %7, %%ecx \n\t" \
128 "movl %8, %%ebx \n\t"
130#define MULADDC_CORE \
133 "addl %%ecx, %%eax \n\t" \
134 "adcl $0, %%edx \n\t" \
135 "addl (%%edi), %%eax \n\t" \
136 "adcl $0, %%edx \n\t" \
137 "movl %%edx, %%ecx \n\t" \
140#if defined(MBEDTLS_HAVE_SSE2)
142#define MULADDC_HUIT \
143 "movd %%ecx, %%mm1 \n\t" \
144 "movd %%ebx, %%mm0 \n\t" \
145 "movd (%%edi), %%mm3 \n\t" \
146 "paddq %%mm3, %%mm1 \n\t" \
147 "movd (%%esi), %%mm2 \n\t" \
148 "pmuludq %%mm0, %%mm2 \n\t" \
149 "movd 4(%%esi), %%mm4 \n\t" \
150 "pmuludq %%mm0, %%mm4 \n\t" \
151 "movd 8(%%esi), %%mm6 \n\t" \
152 "pmuludq %%mm0, %%mm6 \n\t" \
153 "movd 12(%%esi), %%mm7 \n\t" \
154 "pmuludq %%mm0, %%mm7 \n\t" \
155 "paddq %%mm2, %%mm1 \n\t" \
156 "movd 4(%%edi), %%mm3 \n\t" \
157 "paddq %%mm4, %%mm3 \n\t" \
158 "movd 8(%%edi), %%mm5 \n\t" \
159 "paddq %%mm6, %%mm5 \n\t" \
160 "movd 12(%%edi), %%mm4 \n\t" \
161 "paddq %%mm4, %%mm7 \n\t" \
162 "movd %%mm1, (%%edi) \n\t" \
163 "movd 16(%%esi), %%mm2 \n\t" \
164 "pmuludq %%mm0, %%mm2 \n\t" \
165 "psrlq $32, %%mm1 \n\t" \
166 "movd 20(%%esi), %%mm4 \n\t" \
167 "pmuludq %%mm0, %%mm4 \n\t" \
168 "paddq %%mm3, %%mm1 \n\t" \
169 "movd 24(%%esi), %%mm6 \n\t" \
170 "pmuludq %%mm0, %%mm6 \n\t" \
171 "movd %%mm1, 4(%%edi) \n\t" \
172 "psrlq $32, %%mm1 \n\t" \
173 "movd 28(%%esi), %%mm3 \n\t" \
174 "pmuludq %%mm0, %%mm3 \n\t" \
175 "paddq %%mm5, %%mm1 \n\t" \
176 "movd 16(%%edi), %%mm5 \n\t" \
177 "paddq %%mm5, %%mm2 \n\t" \
178 "movd %%mm1, 8(%%edi) \n\t" \
179 "psrlq $32, %%mm1 \n\t" \
180 "paddq %%mm7, %%mm1 \n\t" \
181 "movd 20(%%edi), %%mm5 \n\t" \
182 "paddq %%mm5, %%mm4 \n\t" \
183 "movd %%mm1, 12(%%edi) \n\t" \
184 "psrlq $32, %%mm1 \n\t" \
185 "paddq %%mm2, %%mm1 \n\t" \
186 "movd 24(%%edi), %%mm5 \n\t" \
187 "paddq %%mm5, %%mm6 \n\t" \
188 "movd %%mm1, 16(%%edi) \n\t" \
189 "psrlq $32, %%mm1 \n\t" \
190 "paddq %%mm4, %%mm1 \n\t" \
191 "movd 28(%%edi), %%mm5 \n\t" \
192 "paddq %%mm5, %%mm3 \n\t" \
193 "movd %%mm1, 20(%%edi) \n\t" \
194 "psrlq $32, %%mm1 \n\t" \
195 "paddq %%mm6, %%mm1 \n\t" \
196 "movd %%mm1, 24(%%edi) \n\t" \
197 "psrlq $32, %%mm1 \n\t" \
198 "paddq %%mm3, %%mm1 \n\t" \
199 "movd %%mm1, 28(%%edi) \n\t" \
200 "addl $32, %%edi \n\t" \
201 "addl $32, %%esi \n\t" \
202 "psrlq $32, %%mm1 \n\t" \
203 "movd %%mm1, %%ecx \n\t"
205#define MULADDC_STOP \
207 "movl %4, %%ebx \n\t" \
208 "movl %%ecx, %1 \n\t" \
209 "movl %%edi, %2 \n\t" \
210 "movl %%esi, %3 \n\t" \
211 : "=m" (t), "=m" (c), "=m" (d), "=m" (s) \
212 : "m" (t), "m" (s), "m" (d), "m" (c), "m" (b) \
213 : "eax", "ebx", "ecx", "edx", "esi", "edi" \
218#define MULADDC_STOP \
219 "movl %4, %%ebx \n\t" \
220 "movl %%ecx, %1 \n\t" \
221 "movl %%edi, %2 \n\t" \
222 "movl %%esi, %3 \n\t" \
223 : "=m" (t), "=m" (c), "=m" (d), "=m" (s) \
224 : "m" (t), "m" (s), "m" (d), "m" (c), "m" (b) \
225 : "eax", "ebx", "ecx", "edx", "esi", "edi" \
230#if defined(__amd64__) || defined (__x86_64__)
232#define MULADDC_INIT \
236#define MULADDC_CORE \
237 "movq (%%rsi), %%rax\n" \
240 "addq %%rcx, %%rax\n" \
241 "movq %%r8, %%rcx\n" \
244 "addq %%rax, (%%rdi)\n" \
245 "adcq %%rdx, %%rcx\n" \
248#define MULADDC_STOP \
249 : "+c" (c), "+D" (d), "+S" (s), "+m" (*(uint64_t (*)[16]) d) \
250 : "b" (b), "m" (*(const uint64_t (*)[16]) s) \
251 : "rax", "rdx", "r8" \
256#if defined(__aarch64__)
258#define MULADDC_INIT \
261#define MULADDC_CORE \
262 "ldr x4, [%2], #8 \n\t" \
263 "ldr x5, [%1] \n\t" \
264 "mul x6, x4, %4 \n\t" \
265 "umulh x7, x4, %4 \n\t" \
266 "adds x5, x5, x6 \n\t" \
267 "adc x7, x7, xzr \n\t" \
268 "adds x5, x5, %0 \n\t" \
269 "adc %0, x7, xzr \n\t" \
270 "str x5, [%1], #8 \n\t"
272#define MULADDC_STOP \
273 : "+r" (c), "+r" (d), "+r" (s), "+m" (*(uint64_t (*)[16]) d) \
274 : "r" (b), "m" (*(const uint64_t (*)[16]) s) \
275 : "x4", "x5", "x6", "x7", "cc" \
280#if defined(__mc68020__) || defined(__mcpu32__)
282#define MULADDC_INIT \
284 "movl %3, %%a2 \n\t" \
285 "movl %4, %%a3 \n\t" \
286 "movl %5, %%d3 \n\t" \
287 "movl %6, %%d2 \n\t" \
288 "moveq #0, %%d0 \n\t"
290#define MULADDC_CORE \
291 "movel %%a2@+, %%d1 \n\t" \
292 "mulul %%d2, %%d4:%%d1 \n\t" \
293 "addl %%d3, %%d1 \n\t" \
294 "addxl %%d0, %%d4 \n\t" \
295 "moveq #0, %%d3 \n\t" \
296 "addl %%d1, %%a3@+ \n\t" \
297 "addxl %%d4, %%d3 \n\t"
299#define MULADDC_STOP \
300 "movl %%d3, %0 \n\t" \
301 "movl %%a3, %1 \n\t" \
302 "movl %%a2, %2 \n\t" \
303 : "=m" (c), "=m" (d), "=m" (s) \
304 : "m" (s), "m" (d), "m" (c), "m" (b) \
305 : "d0", "d1", "d2", "d3", "d4", "a2", "a3" \
308#define MULADDC_HUIT \
309 "movel %%a2@+, %%d1 \n\t" \
310 "mulul %%d2, %%d4:%%d1 \n\t" \
311 "addxl %%d3, %%d1 \n\t" \
312 "addxl %%d0, %%d4 \n\t" \
313 "addl %%d1, %%a3@+ \n\t" \
314 "movel %%a2@+, %%d1 \n\t" \
315 "mulul %%d2, %%d3:%%d1 \n\t" \
316 "addxl %%d4, %%d1 \n\t" \
317 "addxl %%d0, %%d3 \n\t" \
318 "addl %%d1, %%a3@+ \n\t" \
319 "movel %%a2@+, %%d1 \n\t" \
320 "mulul %%d2, %%d4:%%d1 \n\t" \
321 "addxl %%d3, %%d1 \n\t" \
322 "addxl %%d0, %%d4 \n\t" \
323 "addl %%d1, %%a3@+ \n\t" \
324 "movel %%a2@+, %%d1 \n\t" \
325 "mulul %%d2, %%d3:%%d1 \n\t" \
326 "addxl %%d4, %%d1 \n\t" \
327 "addxl %%d0, %%d3 \n\t" \
328 "addl %%d1, %%a3@+ \n\t" \
329 "movel %%a2@+, %%d1 \n\t" \
330 "mulul %%d2, %%d4:%%d1 \n\t" \
331 "addxl %%d3, %%d1 \n\t" \
332 "addxl %%d0, %%d4 \n\t" \
333 "addl %%d1, %%a3@+ \n\t" \
334 "movel %%a2@+, %%d1 \n\t" \
335 "mulul %%d2, %%d3:%%d1 \n\t" \
336 "addxl %%d4, %%d1 \n\t" \
337 "addxl %%d0, %%d3 \n\t" \
338 "addl %%d1, %%a3@+ \n\t" \
339 "movel %%a2@+, %%d1 \n\t" \
340 "mulul %%d2, %%d4:%%d1 \n\t" \
341 "addxl %%d3, %%d1 \n\t" \
342 "addxl %%d0, %%d4 \n\t" \
343 "addl %%d1, %%a3@+ \n\t" \
344 "movel %%a2@+, %%d1 \n\t" \
345 "mulul %%d2, %%d3:%%d1 \n\t" \
346 "addxl %%d4, %%d1 \n\t" \
347 "addxl %%d0, %%d3 \n\t" \
348 "addl %%d1, %%a3@+ \n\t" \
349 "addxl %%d0, %%d3 \n\t"
353#if defined(__powerpc64__) || defined(__ppc64__)
355#if defined(__MACH__) && defined(__APPLE__)
357#define MULADDC_INIT \
363 "addi r3, r3, -8 \n\t" \
364 "addi r4, r4, -8 \n\t" \
365 "addic r5, r5, 0 \n\t"
367#define MULADDC_CORE \
368 "ldu r7, 8(r3) \n\t" \
369 "mulld r8, r7, r6 \n\t" \
370 "mulhdu r9, r7, r6 \n\t" \
371 "adde r8, r8, r5 \n\t" \
372 "ld r7, 8(r4) \n\t" \
373 "addze r5, r9 \n\t" \
374 "addc r8, r8, r7 \n\t" \
375 "stdu r8, 8(r4) \n\t"
377#define MULADDC_STOP \
378 "addze r5, r5 \n\t" \
379 "addi r4, r4, 8 \n\t" \
380 "addi r3, r3, 8 \n\t" \
384 : "=m" (c), "=m" (d), "=m" (s) \
385 : "m" (s), "m" (d), "m" (c), "m" (b) \
386 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \
392#define MULADDC_INIT \
398 "addi %%r3, %%r3, -8 \n\t" \
399 "addi %%r4, %%r4, -8 \n\t" \
400 "addic %%r5, %%r5, 0 \n\t"
402#define MULADDC_CORE \
403 "ldu %%r7, 8(%%r3) \n\t" \
404 "mulld %%r8, %%r7, %%r6 \n\t" \
405 "mulhdu %%r9, %%r7, %%r6 \n\t" \
406 "adde %%r8, %%r8, %%r5 \n\t" \
407 "ld %%r7, 8(%%r4) \n\t" \
408 "addze %%r5, %%r9 \n\t" \
409 "addc %%r8, %%r8, %%r7 \n\t" \
410 "stdu %%r8, 8(%%r4) \n\t"
412#define MULADDC_STOP \
413 "addze %%r5, %%r5 \n\t" \
414 "addi %%r4, %%r4, 8 \n\t" \
415 "addi %%r3, %%r3, 8 \n\t" \
416 "std %%r5, %0 \n\t" \
417 "std %%r4, %1 \n\t" \
418 "std %%r3, %2 \n\t" \
419 : "=m" (c), "=m" (d), "=m" (s) \
420 : "m" (s), "m" (d), "m" (c), "m" (b) \
421 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \
426#elif defined(__powerpc__) || defined(__ppc__)
428#if defined(__MACH__) && defined(__APPLE__)
430#define MULADDC_INIT \
436 "addi r3, r3, -4 \n\t" \
437 "addi r4, r4, -4 \n\t" \
438 "addic r5, r5, 0 \n\t"
440#define MULADDC_CORE \
441 "lwzu r7, 4(r3) \n\t" \
442 "mullw r8, r7, r6 \n\t" \
443 "mulhwu r9, r7, r6 \n\t" \
444 "adde r8, r8, r5 \n\t" \
445 "lwz r7, 4(r4) \n\t" \
446 "addze r5, r9 \n\t" \
447 "addc r8, r8, r7 \n\t" \
448 "stwu r8, 4(r4) \n\t"
450#define MULADDC_STOP \
451 "addze r5, r5 \n\t" \
452 "addi r4, r4, 4 \n\t" \
453 "addi r3, r3, 4 \n\t" \
457 : "=m" (c), "=m" (d), "=m" (s) \
458 : "m" (s), "m" (d), "m" (c), "m" (b) \
459 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \
464#define MULADDC_INIT \
466 "lwz %%r3, %3 \n\t" \
467 "lwz %%r4, %4 \n\t" \
468 "lwz %%r5, %5 \n\t" \
469 "lwz %%r6, %6 \n\t" \
470 "addi %%r3, %%r3, -4 \n\t" \
471 "addi %%r4, %%r4, -4 \n\t" \
472 "addic %%r5, %%r5, 0 \n\t"
474#define MULADDC_CORE \
475 "lwzu %%r7, 4(%%r3) \n\t" \
476 "mullw %%r8, %%r7, %%r6 \n\t" \
477 "mulhwu %%r9, %%r7, %%r6 \n\t" \
478 "adde %%r8, %%r8, %%r5 \n\t" \
479 "lwz %%r7, 4(%%r4) \n\t" \
480 "addze %%r5, %%r9 \n\t" \
481 "addc %%r8, %%r8, %%r7 \n\t" \
482 "stwu %%r8, 4(%%r4) \n\t"
484#define MULADDC_STOP \
485 "addze %%r5, %%r5 \n\t" \
486 "addi %%r4, %%r4, 4 \n\t" \
487 "addi %%r3, %%r3, 4 \n\t" \
488 "stw %%r5, %0 \n\t" \
489 "stw %%r4, %1 \n\t" \
490 "stw %%r3, %2 \n\t" \
491 : "=m" (c), "=m" (d), "=m" (s) \
492 : "m" (s), "m" (d), "m" (c), "m" (b) \
493 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \
504#if 0 && defined(__sparc__)
505#if defined(__sparc64__)
507#define MULADDC_INIT \
509 "ldx %3, %%o0 \n\t" \
510 "ldx %4, %%o1 \n\t" \
514#define MULADDC_CORE \
515 "ld [%%o0], %%o4 \n\t" \
517 "ld [%%o1], %%o5 \n\t" \
518 "umul %%o3, %%o4, %%o4 \n\t" \
519 "addcc %%o4, %%o2, %%o4 \n\t" \
520 "rd %%y, %%g1 \n\t" \
521 "addx %%g1, 0, %%g1 \n\t" \
522 "addcc %%o4, %%o5, %%o4 \n\t" \
523 "st %%o4, [%%o1] \n\t" \
524 "addx %%g1, 0, %%o2 \n\t" \
527 #define MULADDC_STOP \
529 "stx %%o1, %1 \n\t" \
530 "stx %%o0, %2 \n\t" \
531 : "=m" (c), "=m" (d), "=m" (s) \
532 : "m" (s), "m" (d), "m" (c), "m" (b) \
533 : "g1", "o0", "o1", "o2", "o3", "o4", \
539#define MULADDC_INIT \
546#define MULADDC_CORE \
547 "ld [%%o0], %%o4 \n\t" \
549 "ld [%%o1], %%o5 \n\t" \
550 "umul %%o3, %%o4, %%o4 \n\t" \
551 "addcc %%o4, %%o2, %%o4 \n\t" \
552 "rd %%y, %%g1 \n\t" \
553 "addx %%g1, 0, %%g1 \n\t" \
554 "addcc %%o4, %%o5, %%o4 \n\t" \
555 "st %%o4, [%%o1] \n\t" \
556 "addx %%g1, 0, %%o2 \n\t" \
559#define MULADDC_STOP \
563 : "=m" (c), "=m" (d), "=m" (s) \
564 : "m" (s), "m" (d), "m" (c), "m" (b) \
565 : "g1", "o0", "o1", "o2", "o3", "o4", \
572#if defined(__microblaze__) || defined(microblaze)
574#define MULADDC_INIT \
580 "andi r7, r6, 0xffff \n\t" \
581 "bsrli r6, r6, 16 \n\t"
583#if(__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)
584#define MULADDC_LHUI \
585 "lhui r9, r3, 0 \n\t" \
586 "addi r3, r3, 2 \n\t" \
587 "lhui r8, r3, 0 \n\t"
589#define MULADDC_LHUI \
590 "lhui r8, r3, 0 \n\t" \
591 "addi r3, r3, 2 \n\t" \
592 "lhui r9, r3, 0 \n\t"
595#define MULADDC_CORE \
597 "addi r3, r3, 2 \n\t" \
598 "mul r10, r9, r6 \n\t" \
599 "mul r11, r8, r7 \n\t" \
600 "mul r12, r9, r7 \n\t" \
601 "mul r13, r8, r6 \n\t" \
602 "bsrli r8, r10, 16 \n\t" \
603 "bsrli r9, r11, 16 \n\t" \
604 "add r13, r13, r8 \n\t" \
605 "add r13, r13, r9 \n\t" \
606 "bslli r10, r10, 16 \n\t" \
607 "bslli r11, r11, 16 \n\t" \
608 "add r12, r12, r10 \n\t" \
609 "addc r13, r13, r0 \n\t" \
610 "add r12, r12, r11 \n\t" \
611 "addc r13, r13, r0 \n\t" \
612 "lwi r10, r4, 0 \n\t" \
613 "add r12, r12, r10 \n\t" \
614 "addc r13, r13, r0 \n\t" \
615 "add r12, r12, r5 \n\t" \
616 "addc r5, r13, r0 \n\t" \
617 "swi r12, r4, 0 \n\t" \
618 "addi r4, r4, 4 \n\t"
620#define MULADDC_STOP \
624 : "=m" (c), "=m" (d), "=m" (s) \
625 : "m" (s), "m" (d), "m" (c), "m" (b) \
626 : "r3", "r4", "r5", "r6", "r7", "r8", \
627 "r9", "r10", "r11", "r12", "r13" \
632#if defined(__tricore__)
634#define MULADDC_INIT \
636 "ld.a %%a2, %3 \n\t" \
637 "ld.a %%a3, %4 \n\t" \
638 "ld.w %%d4, %5 \n\t" \
639 "ld.w %%d1, %6 \n\t" \
640 "xor %%d5, %%d5 \n\t"
642#define MULADDC_CORE \
643 "ld.w %%d0, [%%a2+] \n\t" \
644 "madd.u %%e2, %%e4, %%d0, %%d1 \n\t" \
645 "ld.w %%d0, [%%a3] \n\t" \
646 "addx %%d2, %%d2, %%d0 \n\t" \
647 "addc %%d3, %%d3, 0 \n\t" \
648 "mov %%d4, %%d3 \n\t" \
649 "st.w [%%a3+], %%d2 \n\t"
651#define MULADDC_STOP \
652 "st.w %0, %%d4 \n\t" \
653 "st.a %1, %%a3 \n\t" \
654 "st.a %2, %%a2 \n\t" \
655 : "=m" (c), "=m" (d), "=m" (s) \
656 : "m" (s), "m" (d), "m" (c), "m" (b) \
657 : "d0", "d1", "e2", "d4", "a2", "a3" \
673#if defined(__GNUC__) && !defined(__OPTIMIZE__)
674#define MULADDC_CANNOT_USE_R7
677#if defined(__arm__) && !defined(MULADDC_CANNOT_USE_R7)
679#if defined(__thumb__) && !defined(__thumb2__)
681#define MULADDC_INIT \
687 "lsr r7, r3, #16 \n\t" \
689 "lsl r7, r3, #16 \n\t" \
690 "lsr r7, r7, #16 \n\t" \
693#define MULADDC_CORE \
694 "ldmia r0!, {r6} \n\t" \
695 "lsr r7, r6, #16 \n\t" \
696 "lsl r6, r6, #16 \n\t" \
697 "lsr r6, r6, #16 \n\t" \
706 "lsr r3, r6, #16 \n\t" \
707 "add r5, r5, r3 \n\t" \
708 "lsr r3, r7, #16 \n\t" \
709 "add r5, r5, r3 \n\t" \
710 "add r4, r4, r2 \n\t" \
713 "lsl r3, r6, #16 \n\t" \
714 "add r4, r4, r3 \n\t" \
716 "lsl r3, r7, #16 \n\t" \
717 "add r4, r4, r3 \n\t" \
719 "ldr r3, [r1] \n\t" \
720 "add r4, r4, r3 \n\t" \
722 "stmia r1!, {r4} \n\t"
724#define MULADDC_STOP \
728 : "=m" (c), "=m" (d), "=m" (s) \
729 : "m" (s), "m" (d), "m" (c), "m" (b) \
730 : "r0", "r1", "r2", "r3", "r4", "r5", \
731 "r6", "r7", "r8", "r9", "cc" \
734#elif (__ARM_ARCH >= 6) && \
735 defined (__ARM_FEATURE_DSP) && (__ARM_FEATURE_DSP == 1)
737#define MULADDC_INIT \
740#define MULADDC_CORE \
741 "ldr r0, [%0], #4 \n\t" \
742 "ldr r1, [%1] \n\t" \
743 "umaal r1, %2, %3, r0 \n\t" \
744 "str r1, [%1], #4 \n\t"
746#define MULADDC_STOP \
747 : "=r" (s), "=r" (d), "=r" (c) \
748 : "r" (b), "0" (s), "1" (d), "2" (c) \
749 : "r0", "r1", "memory" \
754#define MULADDC_INIT \
761#define MULADDC_CORE \
762 "ldr r4, [r0], #4 \n\t" \
764 "ldr r6, [r1] \n\t" \
765 "umlal r2, r5, r3, r4 \n\t" \
766 "adds r7, r6, r2 \n\t" \
767 "adc r2, r5, #0 \n\t" \
768 "str r7, [r1], #4 \n\t"
770#define MULADDC_STOP \
774 : "=m" (c), "=m" (d), "=m" (s) \
775 : "m" (s), "m" (d), "m" (c), "m" (b) \
776 : "r0", "r1", "r2", "r3", "r4", "r5", \
784#if defined(__alpha__)
786#define MULADDC_INIT \
793#define MULADDC_CORE \
794 "ldq $6, 0($1) \n\t" \
795 "addq $1, 8, $1 \n\t" \
796 "mulq $6, $4, $7 \n\t" \
797 "umulh $6, $4, $6 \n\t" \
798 "addq $7, $3, $7 \n\t" \
799 "cmpult $7, $3, $3 \n\t" \
800 "ldq $5, 0($2) \n\t" \
801 "addq $7, $5, $7 \n\t" \
802 "cmpult $7, $5, $5 \n\t" \
803 "stq $7, 0($2) \n\t" \
804 "addq $2, 8, $2 \n\t" \
805 "addq $6, $3, $3 \n\t" \
806 "addq $5, $3, $3 \n\t"
808#define MULADDC_STOP \
812 : "=m" (c), "=m" (d), "=m" (s) \
813 : "m" (s), "m" (d), "m" (c), "m" (b) \
814 : "$1", "$2", "$3", "$4", "$5", "$6", "$7" \
818#if defined(__mips__) && !defined(__mips64)
820#define MULADDC_INIT \
827#define MULADDC_CORE \
828 "lw $14, 0($10) \n\t" \
829 "multu $13, $14 \n\t" \
830 "addi $10, $10, 4 \n\t" \
833 "addu $14, $12, $14 \n\t" \
834 "lw $15, 0($11) \n\t" \
835 "sltu $12, $14, $12 \n\t" \
836 "addu $15, $14, $15 \n\t" \
837 "sltu $14, $15, $14 \n\t" \
838 "addu $12, $12, $9 \n\t" \
839 "sw $15, 0($11) \n\t" \
840 "addu $12, $12, $14 \n\t" \
841 "addi $11, $11, 4 \n\t"
843#define MULADDC_STOP \
847 : "=m" (c), "=m" (d), "=m" (s) \
848 : "m" (s), "m" (d), "m" (c), "m" (b) \
849 : "$9", "$10", "$11", "$12", "$13", "$14", "$15", "lo", "hi" \
855#if (defined(_MSC_VER) && defined(_M_IX86)) || defined(__WATCOMC__)
857#define MULADDC_INIT \
863#define MULADDC_CORE \
868 __asm add eax, [edi] \
873#if defined(MBEDTLS_HAVE_SSE2)
875#define EMIT __asm _emit
877#define MULADDC_HUIT \
878 EMIT 0x0F EMIT 0x6E EMIT 0xC9 \
879 EMIT 0x0F EMIT 0x6E EMIT 0xC3 \
880 EMIT 0x0F EMIT 0x6E EMIT 0x1F \
881 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \
882 EMIT 0x0F EMIT 0x6E EMIT 0x16 \
883 EMIT 0x0F EMIT 0xF4 EMIT 0xD0 \
884 EMIT 0x0F EMIT 0x6E EMIT 0x66 EMIT 0x04 \
885 EMIT 0x0F EMIT 0xF4 EMIT 0xE0 \
886 EMIT 0x0F EMIT 0x6E EMIT 0x76 EMIT 0x08 \
887 EMIT 0x0F EMIT 0xF4 EMIT 0xF0 \
888 EMIT 0x0F EMIT 0x6E EMIT 0x7E EMIT 0x0C \
889 EMIT 0x0F EMIT 0xF4 EMIT 0xF8 \
890 EMIT 0x0F EMIT 0xD4 EMIT 0xCA \
891 EMIT 0x0F EMIT 0x6E EMIT 0x5F EMIT 0x04 \
892 EMIT 0x0F EMIT 0xD4 EMIT 0xDC \
893 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x08 \
894 EMIT 0x0F EMIT 0xD4 EMIT 0xEE \
895 EMIT 0x0F EMIT 0x6E EMIT 0x67 EMIT 0x0C \
896 EMIT 0x0F EMIT 0xD4 EMIT 0xFC \
897 EMIT 0x0F EMIT 0x7E EMIT 0x0F \
898 EMIT 0x0F EMIT 0x6E EMIT 0x56 EMIT 0x10 \
899 EMIT 0x0F EMIT 0xF4 EMIT 0xD0 \
900 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
901 EMIT 0x0F EMIT 0x6E EMIT 0x66 EMIT 0x14 \
902 EMIT 0x0F EMIT 0xF4 EMIT 0xE0 \
903 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \
904 EMIT 0x0F EMIT 0x6E EMIT 0x76 EMIT 0x18 \
905 EMIT 0x0F EMIT 0xF4 EMIT 0xF0 \
906 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x04 \
907 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
908 EMIT 0x0F EMIT 0x6E EMIT 0x5E EMIT 0x1C \
909 EMIT 0x0F EMIT 0xF4 EMIT 0xD8 \
910 EMIT 0x0F EMIT 0xD4 EMIT 0xCD \
911 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x10 \
912 EMIT 0x0F EMIT 0xD4 EMIT 0xD5 \
913 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x08 \
914 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
915 EMIT 0x0F EMIT 0xD4 EMIT 0xCF \
916 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x14 \
917 EMIT 0x0F EMIT 0xD4 EMIT 0xE5 \
918 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x0C \
919 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
920 EMIT 0x0F EMIT 0xD4 EMIT 0xCA \
921 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x18 \
922 EMIT 0x0F EMIT 0xD4 EMIT 0xF5 \
923 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x10 \
924 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
925 EMIT 0x0F EMIT 0xD4 EMIT 0xCC \
926 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x1C \
927 EMIT 0x0F EMIT 0xD4 EMIT 0xDD \
928 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x14 \
929 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
930 EMIT 0x0F EMIT 0xD4 EMIT 0xCE \
931 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x18 \
932 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
933 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \
934 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x1C \
935 EMIT 0x83 EMIT 0xC7 EMIT 0x20 \
936 EMIT 0x83 EMIT 0xC6 EMIT 0x20 \
937 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \
938 EMIT 0x0F EMIT 0x7E EMIT 0xC9
940#define MULADDC_STOP \
941 EMIT 0x0F EMIT 0x77 \
948#define MULADDC_STOP \
958#if !defined(MULADDC_CORE)
959#if defined(MBEDTLS_HAVE_UDBL)
961#define MULADDC_INIT \
964 mbedtls_mpi_uint r0, r1;
966#define MULADDC_CORE \
967 r = *(s++) * (mbedtls_t_udbl) b; \
968 r0 = (mbedtls_mpi_uint) r; \
969 r1 = (mbedtls_mpi_uint)( r >> biL ); \
970 r0 += c; r1 += (r0 < c); \
971 r0 += *d; r1 += (r0 < *d); \
974#define MULADDC_STOP \
978#define MULADDC_INIT \
980 mbedtls_mpi_uint s0, s1, b0, b1; \
981 mbedtls_mpi_uint r0, r1, rx, ry; \
982 b0 = ( b << biH ) >> biH; \
985#define MULADDC_CORE \
986 s0 = ( *s << biH ) >> biH; \
987 s1 = ( *s >> biH ); s++; \
988 rx = s0 * b1; r0 = s0 * b0; \
989 ry = s1 * b0; r1 = s1 * b1; \
990 r1 += ( rx >> biH ); \
991 r1 += ( ry >> biH ); \
992 rx <<= biH; ry <<= biH; \
993 r0 += rx; r1 += (r0 < rx); \
994 r0 += ry; r1 += (r0 < ry); \
995 r0 += c; r1 += (r0 < c); \
996 r0 += *d; r1 += (r0 < *d); \
999#define MULADDC_STOP \
Multi-precision integer library.
Configuration options (set of defines)