Hannes Tschofenig
Example program to test AES-GCM functionality. Used for a workshop
Embed:
(wiki syntax)
Show/hide line numbers
bn_mul.h
Go to the documentation of this file.
00001 /** 00002 * \file bn_mul.h 00003 * 00004 * \brief Multi-precision integer library 00005 * 00006 * Copyright (C) 2006-2010, Brainspark B.V. 00007 * 00008 * This file is part of PolarSSL (http://www.polarssl.org) 00009 * Lead Maintainer: Paul Bakker <polarssl_maintainer at polarssl.org> 00010 * 00011 * All rights reserved. 00012 * 00013 * This program is free software; you can redistribute it and/or modify 00014 * it under the terms of the GNU General Public License as published by 00015 * the Free Software Foundation; either version 2 of the License, or 00016 * (at your option) any later version. 00017 * 00018 * This program is distributed in the hope that it will be useful, 00019 * but WITHOUT ANY WARRANTY; without even the implied warranty of 00020 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 00021 * GNU General Public License for more details. 00022 * 00023 * You should have received a copy of the GNU General Public License along 00024 * with this program; if not, write to the Free Software Foundation, Inc., 00025 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. 00026 */ 00027 /* 00028 * Multiply source vector [s] with b, add result 00029 * to destination vector [d] and set carry c. 00030 * 00031 * Currently supports: 00032 * 00033 * . IA-32 (386+) . AMD64 / EM64T 00034 * . IA-32 (SSE2) . Motorola 68000 00035 * . PowerPC, 32-bit . MicroBlaze 00036 * . PowerPC, 64-bit . TriCore 00037 * . SPARC v8 . ARM v3+ 00038 * . Alpha . MIPS32 00039 * . C, longlong . C, generic 00040 */ 00041 #ifndef POLARSSL_BN_MUL_H 00042 #define POLARSSL_BN_MUL_H 00043 00044 #include "bignum.h" 00045 00046 #if defined(POLARSSL_HAVE_ASM) 00047 00048 #if defined(__GNUC__) 00049 #if defined(__i386__) 00050 00051 #define MULADDC_INIT \ 00052 asm( " \ 00053 movl %%ebx, %0; \ 00054 movl %5, %%esi; \ 00055 movl %6, %%edi; \ 00056 movl %7, %%ecx; \ 00057 movl %8, %%ebx; \ 00058 " 00059 00060 #define MULADDC_CORE \ 00061 " \ 00062 lodsl; \ 00063 mull %%ebx; \ 00064 addl %%ecx, %%eax; \ 00065 adcl $0, %%edx; \ 00066 addl (%%edi), %%eax; \ 00067 adcl $0, %%edx; \ 00068 movl %%edx, %%ecx; \ 00069 stosl; \ 00070 " 00071 00072 #if defined(POLARSSL_HAVE_SSE2) 00073 00074 #define MULADDC_HUIT \ 00075 " \ 00076 movd %%ecx, %%mm1; \ 00077 movd %%ebx, %%mm0; \ 00078 movd (%%edi), %%mm3; \ 00079 paddq %%mm3, %%mm1; \ 00080 movd (%%esi), %%mm2; \ 00081 pmuludq %%mm0, %%mm2; \ 00082 movd 4(%%esi), %%mm4; \ 00083 pmuludq %%mm0, %%mm4; \ 00084 movd 8(%%esi), %%mm6; \ 00085 pmuludq %%mm0, %%mm6; \ 00086 movd 12(%%esi), %%mm7; \ 00087 pmuludq %%mm0, %%mm7; \ 00088 paddq %%mm2, %%mm1; \ 00089 movd 4(%%edi), %%mm3; \ 00090 paddq %%mm4, %%mm3; \ 00091 movd 8(%%edi), %%mm5; \ 00092 paddq %%mm6, %%mm5; \ 00093 movd 12(%%edi), %%mm4; \ 00094 paddq %%mm4, %%mm7; \ 00095 movd %%mm1, (%%edi); \ 00096 movd 16(%%esi), %%mm2; \ 00097 pmuludq %%mm0, %%mm2; \ 00098 psrlq $32, %%mm1; \ 00099 movd 20(%%esi), %%mm4; \ 00100 pmuludq %%mm0, %%mm4; \ 00101 paddq %%mm3, %%mm1; \ 00102 movd 24(%%esi), %%mm6; \ 00103 pmuludq %%mm0, %%mm6; \ 00104 movd %%mm1, 4(%%edi); \ 00105 psrlq $32, %%mm1; \ 00106 movd 28(%%esi), %%mm3; \ 00107 pmuludq %%mm0, %%mm3; \ 00108 paddq %%mm5, %%mm1; \ 00109 movd 16(%%edi), %%mm5; \ 00110 paddq %%mm5, %%mm2; \ 00111 movd %%mm1, 8(%%edi); \ 00112 psrlq $32, %%mm1; \ 00113 paddq %%mm7, %%mm1; \ 00114 movd 20(%%edi), %%mm5; \ 00115 paddq %%mm5, %%mm4; \ 00116 movd %%mm1, 12(%%edi); \ 00117 psrlq $32, %%mm1; \ 00118 paddq %%mm2, %%mm1; \ 00119 movd 24(%%edi), %%mm5; \ 00120 paddq %%mm5, %%mm6; \ 00121 movd %%mm1, 16(%%edi); \ 00122 psrlq $32, %%mm1; \ 00123 paddq %%mm4, %%mm1; \ 00124 movd 28(%%edi), %%mm5; \ 00125 paddq %%mm5, %%mm3; \ 00126 movd %%mm1, 20(%%edi); \ 00127 psrlq $32, %%mm1; \ 00128 paddq %%mm6, %%mm1; \ 00129 movd %%mm1, 24(%%edi); \ 00130 psrlq $32, %%mm1; \ 00131 paddq %%mm3, %%mm1; \ 00132 movd %%mm1, 28(%%edi); \ 00133 addl $32, %%edi; \ 00134 addl $32, %%esi; \ 00135 psrlq $32, %%mm1; \ 00136 movd %%mm1, %%ecx; \ 00137 " 00138 00139 #define MULADDC_STOP \ 00140 " \ 00141 emms; \ 00142 movl %4, %%ebx; \ 00143 movl %%ecx, %1; \ 00144 movl %%edi, %2; \ 00145 movl %%esi, %3; \ 00146 " \ 00147 : "=m" (t), "=m" (c), "=m" (d), "=m" (s) \ 00148 : "m" (t), "m" (s), "m" (d), "m" (c), "m" (b) \ 00149 : "eax", "ecx", "edx", "esi", "edi" \ 00150 ); 00151 00152 #else 00153 00154 #define MULADDC_STOP \ 00155 " \ 00156 movl %4, %%ebx; \ 00157 movl %%ecx, %1; \ 00158 movl %%edi, %2; \ 00159 movl %%esi, %3; \ 00160 " \ 00161 : "=m" (t), "=m" (c), "=m" (d), "=m" (s) \ 00162 : "m" (t), "m" (s), "m" (d), "m" (c), "m" (b) \ 00163 : "eax", "ecx", "edx", "esi", "edi" \ 00164 ); 00165 #endif /* SSE2 */ 00166 #endif /* i386 */ 00167 00168 #if defined(__amd64__) || defined (__x86_64__) 00169 00170 #define MULADDC_INIT \ 00171 asm( \ 00172 " \ 00173 movq %3, %%rsi; \ 00174 movq %4, %%rdi; \ 00175 movq %5, %%rcx; \ 00176 movq %6, %%rbx; \ 00177 xorq %%r8, %%r8; \ 00178 " 00179 00180 #define MULADDC_CORE \ 00181 " \ 00182 movq (%%rsi), %%rax; \ 00183 mulq %%rbx; \ 00184 addq $8, %%rsi; \ 00185 addq %%rcx, %%rax; \ 00186 movq %%r8, %%rcx; \ 00187 adcq $0, %%rdx; \ 00188 nop; \ 00189 addq %%rax, (%%rdi); \ 00190 adcq %%rdx, %%rcx; \ 00191 addq $8, %%rdi; \ 00192 " 00193 00194 #define MULADDC_STOP \ 00195 " \ 00196 movq %%rcx, %0; \ 00197 movq %%rdi, %1; \ 00198 movq %%rsi, %2; \ 00199 " \ 00200 : "=m" (c), "=m" (d), "=m" (s) \ 00201 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00202 : "rax", "rcx", "rdx", "rbx", "rsi", "rdi", "r8" \ 00203 ); 00204 00205 #endif /* AMD64 */ 00206 00207 #if defined(__mc68020__) || defined(__mcpu32__) 00208 00209 #define MULADDC_INIT \ 00210 asm( \ 00211 " \ 00212 movl %3, %%a2; \ 00213 movl %4, %%a3; \ 00214 movl %5, %%d3; \ 00215 movl %6, %%d2; \ 00216 moveq #0, %%d0; \ 00217 " 00218 00219 #define MULADDC_CORE \ 00220 " \ 00221 movel %%a2@+, %%d1; \ 00222 mulul %%d2, %%d4:%%d1; \ 00223 addl %%d3, %%d1; \ 00224 addxl %%d0, %%d4; \ 00225 moveq #0, %%d3; \ 00226 addl %%d1, %%a3@+; \ 00227 addxl %%d4, %%d3; \ 00228 " 00229 00230 #define MULADDC_STOP \ 00231 " \ 00232 movl %%d3, %0; \ 00233 movl %%a3, %1; \ 00234 movl %%a2, %2; \ 00235 " \ 00236 : "=m" (c), "=m" (d), "=m" (s) \ 00237 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00238 : "d0", "d1", "d2", "d3", "d4", "a2", "a3" \ 00239 ); 00240 00241 #define MULADDC_HUIT \ 00242 " \ 00243 movel %%a2@+, %%d1; \ 00244 mulul %%d2, %%d4:%%d1; \ 00245 addxl %%d3, %%d1; \ 00246 addxl %%d0, %%d4; \ 00247 addl %%d1, %%a3@+; \ 00248 movel %%a2@+, %%d1; \ 00249 mulul %%d2, %%d3:%%d1; \ 00250 addxl %%d4, %%d1; \ 00251 addxl %%d0, %%d3; \ 00252 addl %%d1, %%a3@+; \ 00253 movel %%a2@+, %%d1; \ 00254 mulul %%d2, %%d4:%%d1; \ 00255 addxl %%d3, %%d1; \ 00256 addxl %%d0, %%d4; \ 00257 addl %%d1, %%a3@+; \ 00258 movel %%a2@+, %%d1; \ 00259 mulul %%d2, %%d3:%%d1; \ 00260 addxl %%d4, %%d1; \ 00261 addxl %%d0, %%d3; \ 00262 addl %%d1, %%a3@+; \ 00263 movel %%a2@+, %%d1; \ 00264 mulul %%d2, %%d4:%%d1; \ 00265 addxl %%d3, %%d1; \ 00266 addxl %%d0, %%d4; \ 00267 addl %%d1, %%a3@+; \ 00268 movel %%a2@+, %%d1; \ 00269 mulul %%d2, %%d3:%%d1; \ 00270 addxl %%d4, %%d1; \ 00271 addxl %%d0, %%d3; \ 00272 addl %%d1, %%a3@+; \ 00273 movel %%a2@+, %%d1; \ 00274 mulul %%d2, %%d4:%%d1; \ 00275 addxl %%d3, %%d1; \ 00276 addxl %%d0, %%d4; \ 00277 addl %%d1, %%a3@+; \ 00278 movel %%a2@+, %%d1; \ 00279 mulul %%d2, %%d3:%%d1; \ 00280 addxl %%d4, %%d1; \ 00281 addxl %%d0, %%d3; \ 00282 addl %%d1, %%a3@+; \ 00283 addxl %%d0, %%d3; \ 00284 " 00285 00286 #endif /* MC68000 */ 00287 00288 #if defined(__powerpc__) || defined(__ppc__) 00289 #if defined(__powerpc64__) || defined(__ppc64__) 00290 00291 #if defined(__MACH__) && defined(__APPLE__) 00292 00293 #define MULADDC_INIT \ 00294 asm( \ 00295 " \ 00296 ld r3, %3; \ 00297 ld r4, %4; \ 00298 ld r5, %5; \ 00299 ld r6, %6; \ 00300 addi r3, r3, -8; \ 00301 addi r4, r4, -8; \ 00302 addic r5, r5, 0; \ 00303 " 00304 00305 #define MULADDC_CORE \ 00306 " \ 00307 ldu r7, 8(r3); \ 00308 mulld r8, r7, r6; \ 00309 mulhdu r9, r7, r6; \ 00310 adde r8, r8, r5; \ 00311 ld r7, 8(r4); \ 00312 addze r5, r9; \ 00313 addc r8, r8, r7; \ 00314 stdu r8, 8(r4); \ 00315 " 00316 00317 #define MULADDC_STOP \ 00318 " \ 00319 addze r5, r5; \ 00320 addi r4, r4, 8; \ 00321 addi r3, r3, 8; \ 00322 std r5, %0; \ 00323 std r4, %1; \ 00324 std r3, %2; \ 00325 " \ 00326 : "=m" (c), "=m" (d), "=m" (s) \ 00327 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00328 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \ 00329 ); 00330 00331 00332 #else /* __MACH__ && __APPLE__ */ 00333 00334 #define MULADDC_INIT \ 00335 asm( \ 00336 " \ 00337 ld %%r3, %3; \ 00338 ld %%r4, %4; \ 00339 ld %%r5, %5; \ 00340 ld %%r6, %6; \ 00341 addi %%r3, %%r3, -8; \ 00342 addi %%r4, %%r4, -8; \ 00343 addic %%r5, %%r5, 0; \ 00344 " 00345 00346 #define MULADDC_CORE \ 00347 " \ 00348 ldu %%r7, 8(%%r3); \ 00349 mulld %%r8, %%r7, %%r6; \ 00350 mulhdu %%r9, %%r7, %%r6; \ 00351 adde %%r8, %%r8, %%r5; \ 00352 ld %%r7, 8(%%r4); \ 00353 addze %%r5, %%r9; \ 00354 addc %%r8, %%r8, %%r7; \ 00355 stdu %%r8, 8(%%r4); \ 00356 " 00357 00358 #define MULADDC_STOP \ 00359 " \ 00360 addze %%r5, %%r5; \ 00361 addi %%r4, %%r4, 8; \ 00362 addi %%r3, %%r3, 8; \ 00363 std %%r5, %0; \ 00364 std %%r4, %1; \ 00365 std %%r3, %2; \ 00366 " \ 00367 : "=m" (c), "=m" (d), "=m" (s) \ 00368 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00369 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \ 00370 ); 00371 00372 #endif /* __MACH__ && __APPLE__ */ 00373 00374 #else /* PPC32 */ 00375 00376 #if defined(__MACH__) && defined(__APPLE__) 00377 00378 #define MULADDC_INIT \ 00379 asm( \ 00380 " \ 00381 lwz r3, %3; \ 00382 lwz r4, %4; \ 00383 lwz r5, %5; \ 00384 lwz r6, %6; \ 00385 addi r3, r3, -4; \ 00386 addi r4, r4, -4; \ 00387 addic r5, r5, 0; \ 00388 " 00389 00390 #define MULADDC_CORE \ 00391 " \ 00392 lwzu r7, 4(r3); \ 00393 mullw r8, r7, r6; \ 00394 mulhwu r9, r7, r6; \ 00395 adde r8, r8, r5; \ 00396 lwz r7, 4(r4); \ 00397 addze r5, r9; \ 00398 addc r8, r8, r7; \ 00399 stwu r8, 4(r4); \ 00400 " 00401 00402 #define MULADDC_STOP \ 00403 " \ 00404 addze r5, r5; \ 00405 addi r4, r4, 4; \ 00406 addi r3, r3, 4; \ 00407 stw r5, %0; \ 00408 stw r4, %1; \ 00409 stw r3, %2; \ 00410 " \ 00411 : "=m" (c), "=m" (d), "=m" (s) \ 00412 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00413 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \ 00414 ); 00415 00416 #else /* __MACH__ && __APPLE__ */ 00417 00418 #define MULADDC_INIT \ 00419 asm( \ 00420 " \ 00421 lwz %%r3, %3; \ 00422 lwz %%r4, %4; \ 00423 lwz %%r5, %5; \ 00424 lwz %%r6, %6; \ 00425 addi %%r3, %%r3, -4; \ 00426 addi %%r4, %%r4, -4; \ 00427 addic %%r5, %%r5, 0; \ 00428 " 00429 00430 #define MULADDC_CORE \ 00431 " \ 00432 lwzu %%r7, 4(%%r3); \ 00433 mullw %%r8, %%r7, %%r6; \ 00434 mulhwu %%r9, %%r7, %%r6; \ 00435 adde %%r8, %%r8, %%r5; \ 00436 lwz %%r7, 4(%%r4); \ 00437 addze %%r5, %%r9; \ 00438 addc %%r8, %%r8, %%r7; \ 00439 stwu %%r8, 4(%%r4); \ 00440 " 00441 00442 #define MULADDC_STOP \ 00443 " \ 00444 addze %%r5, %%r5; \ 00445 addi %%r4, %%r4, 4; \ 00446 addi %%r3, %%r3, 4; \ 00447 stw %%r5, %0; \ 00448 stw %%r4, %1; \ 00449 stw %%r3, %2; \ 00450 " \ 00451 : "=m" (c), "=m" (d), "=m" (s) \ 00452 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00453 : "r3", "r4", "r5", "r6", "r7", "r8", "r9" \ 00454 ); 00455 00456 #endif /* __MACH__ && __APPLE__ */ 00457 00458 #endif /* PPC32 */ 00459 #endif /* PPC64 */ 00460 00461 #if defined(__sparc__) && defined(__sparc64__) 00462 00463 #define MULADDC_INIT \ 00464 asm( \ 00465 " \ 00466 ldx %3, %%o0; \ 00467 ldx %4, %%o1; \ 00468 ld %5, %%o2; \ 00469 ld %6, %%o3; \ 00470 " 00471 00472 #define MULADDC_CORE \ 00473 " \ 00474 ld [%%o0], %%o4; \ 00475 inc 4, %%o0; \ 00476 ld [%%o1], %%o5; \ 00477 umul %%o3, %%o4, %%o4; \ 00478 addcc %%o4, %%o2, %%o4; \ 00479 rd %%y, %%g1; \ 00480 addx %%g1, 0, %%g1; \ 00481 addcc %%o4, %%o5, %%o4; \ 00482 st %%o4, [%%o1]; \ 00483 addx %%g1, 0, %%o2; \ 00484 inc 4, %%o1; \ 00485 " 00486 00487 #define MULADDC_STOP \ 00488 " \ 00489 st %%o2, %0; \ 00490 stx %%o1, %1; \ 00491 stx %%o0, %2; \ 00492 " \ 00493 : "=m" (c), "=m" (d), "=m" (s) \ 00494 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00495 : "g1", "o0", "o1", "o2", "o3", "o4", \ 00496 "o5" \ 00497 ); 00498 #endif /* SPARCv9 */ 00499 00500 #if defined(__sparc__) && !defined(__sparc64__) 00501 00502 #define MULADDC_INIT \ 00503 asm( \ 00504 " \ 00505 ld %3, %%o0; \ 00506 ld %4, %%o1; \ 00507 ld %5, %%o2; \ 00508 ld %6, %%o3; \ 00509 " 00510 00511 #define MULADDC_CORE \ 00512 " \ 00513 ld [%%o0], %%o4; \ 00514 inc 4, %%o0; \ 00515 ld [%%o1], %%o5; \ 00516 umul %%o3, %%o4, %%o4; \ 00517 addcc %%o4, %%o2, %%o4; \ 00518 rd %%y, %%g1; \ 00519 addx %%g1, 0, %%g1; \ 00520 addcc %%o4, %%o5, %%o4; \ 00521 st %%o4, [%%o1]; \ 00522 addx %%g1, 0, %%o2; \ 00523 inc 4, %%o1; \ 00524 " 00525 00526 #define MULADDC_STOP \ 00527 " \ 00528 st %%o2, %0; \ 00529 st %%o1, %1; \ 00530 st %%o0, %2; \ 00531 " \ 00532 : "=m" (c), "=m" (d), "=m" (s) \ 00533 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00534 : "g1", "o0", "o1", "o2", "o3", "o4", \ 00535 "o5" \ 00536 ); 00537 00538 #endif /* SPARCv8 */ 00539 00540 #if defined(__microblaze__) || defined(microblaze) 00541 00542 #define MULADDC_INIT \ 00543 asm( \ 00544 " \ 00545 lwi r3, %3; \ 00546 lwi r4, %4; \ 00547 lwi r5, %5; \ 00548 lwi r6, %6; \ 00549 andi r7, r6, 0xffff; \ 00550 bsrli r6, r6, 16; \ 00551 " 00552 00553 #define MULADDC_CORE \ 00554 " \ 00555 lhui r8, r3, 0; \ 00556 addi r3, r3, 2; \ 00557 lhui r9, r3, 0; \ 00558 addi r3, r3, 2; \ 00559 mul r10, r9, r6; \ 00560 mul r11, r8, r7; \ 00561 mul r12, r9, r7; \ 00562 mul r13, r8, r6; \ 00563 bsrli r8, r10, 16; \ 00564 bsrli r9, r11, 16; \ 00565 add r13, r13, r8; \ 00566 add r13, r13, r9; \ 00567 bslli r10, r10, 16; \ 00568 bslli r11, r11, 16; \ 00569 add r12, r12, r10; \ 00570 addc r13, r13, r0; \ 00571 add r12, r12, r11; \ 00572 addc r13, r13, r0; \ 00573 lwi r10, r4, 0; \ 00574 add r12, r12, r10; \ 00575 addc r13, r13, r0; \ 00576 add r12, r12, r5; \ 00577 addc r5, r13, r0; \ 00578 swi r12, r4, 0; \ 00579 addi r4, r4, 4; \ 00580 " 00581 00582 #define MULADDC_STOP \ 00583 " \ 00584 swi r5, %0; \ 00585 swi r4, %1; \ 00586 swi r3, %2; \ 00587 " \ 00588 : "=m" (c), "=m" (d), "=m" (s) \ 00589 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00590 : "r3", "r4" "r5", "r6", "r7", "r8", \ 00591 "r9", "r10", "r11", "r12", "r13" \ 00592 ); 00593 00594 #endif /* MicroBlaze */ 00595 00596 #if defined(__tricore__) 00597 00598 #define MULADDC_INIT \ 00599 asm( \ 00600 " \ 00601 ld.a %%a2, %3; \ 00602 ld.a %%a3, %4; \ 00603 ld.w %%d4, %5; \ 00604 ld.w %%d1, %6; \ 00605 xor %%d5, %%d5; \ 00606 " 00607 00608 #define MULADDC_CORE \ 00609 " \ 00610 ld.w %%d0, [%%a2+]; \ 00611 madd.u %%e2, %%e4, %%d0, %%d1; \ 00612 ld.w %%d0, [%%a3]; \ 00613 addx %%d2, %%d2, %%d0; \ 00614 addc %%d3, %%d3, 0; \ 00615 mov %%d4, %%d3; \ 00616 st.w [%%a3+], %%d2; \ 00617 " 00618 00619 #define MULADDC_STOP \ 00620 " \ 00621 st.w %0, %%d4; \ 00622 st.a %1, %%a3; \ 00623 st.a %2, %%a2; \ 00624 " \ 00625 : "=m" (c), "=m" (d), "=m" (s) \ 00626 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00627 : "d0", "d1", "e2", "d4", "a2", "a3" \ 00628 ); 00629 00630 #endif /* TriCore */ 00631 00632 #if defined(__arm__) 00633 00634 #if defined(__thumb__) && !defined(__thumb2__) 00635 00636 #define MULADDC_INIT \ 00637 asm( \ 00638 " \ 00639 ldr r0, %3; \ 00640 ldr r1, %4; \ 00641 ldr r2, %5; \ 00642 ldr r3, %6; \ 00643 lsr r7, r3, #16; \ 00644 mov r9, r7; \ 00645 lsl r7, r3, #16; \ 00646 lsr r7, r7, #16; \ 00647 mov r8, r7; \ 00648 " 00649 00650 #define MULADDC_CORE \ 00651 " \ 00652 ldmia r0!, {r6}; \ 00653 lsr r7, r6, #16; \ 00654 lsl r6, r6, #16; \ 00655 lsr r6, r6, #16; \ 00656 mov r4, r8; \ 00657 mul r4, r6; \ 00658 mov r3, r9; \ 00659 mul r6, r3; \ 00660 mov r5, r9; \ 00661 mul r5, r7; \ 00662 mov r3, r8; \ 00663 mul r7, r3; \ 00664 lsr r3, r6, #16; \ 00665 add r5, r5, r3; \ 00666 lsr r3, r7, #16; \ 00667 add r5, r5, r3; \ 00668 add r4, r4, r2; \ 00669 mov r2, #0; \ 00670 adc r5, r2; \ 00671 lsl r3, r6, #16; \ 00672 add r4, r4, r3; \ 00673 adc r5, r2; \ 00674 lsl r3, r7, #16; \ 00675 add r4, r4, r3; \ 00676 adc r5, r2; \ 00677 ldr r3, [r1]; \ 00678 add r4, r4, r3; \ 00679 adc r2, r5; \ 00680 stmia r1!, {r4}; \ 00681 " 00682 00683 #define MULADDC_STOP \ 00684 " \ 00685 str r2, %0; \ 00686 str r1, %1; \ 00687 str r0, %2; \ 00688 " \ 00689 : "=m" (c), "=m" (d), "=m" (s) \ 00690 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00691 : "r0", "r1", "r2", "r3", "r4", "r5", \ 00692 "r6", "r7", "r8", "r9", "cc" \ 00693 ); 00694 00695 #else 00696 00697 #define MULADDC_INIT \ 00698 asm( \ 00699 " \ 00700 ldr r0, %3; \ 00701 ldr r1, %4; \ 00702 ldr r2, %5; \ 00703 ldr r3, %6; \ 00704 " 00705 00706 #define MULADDC_CORE \ 00707 " \ 00708 ldr r4, [r0], #4; \ 00709 mov r5, #0; \ 00710 ldr r6, [r1]; \ 00711 umlal r2, r5, r3, r4; \ 00712 adds r7, r6, r2; \ 00713 adc r2, r5, #0; \ 00714 str r7, [r1], #4; \ 00715 " 00716 00717 #define MULADDC_STOP \ 00718 " \ 00719 str r2, %0; \ 00720 str r1, %1; \ 00721 str r0, %2; \ 00722 " \ 00723 : "=m" (c), "=m" (d), "=m" (s) \ 00724 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00725 : "r0", "r1", "r2", "r3", "r4", "r5", \ 00726 "r6", "r7", "cc" \ 00727 ); 00728 00729 #endif /* Thumb */ 00730 00731 #endif /* ARMv3 */ 00732 00733 #if defined(__alpha__) 00734 00735 #define MULADDC_INIT \ 00736 asm( \ 00737 " \ 00738 ldq $1, %3; \ 00739 ldq $2, %4; \ 00740 ldq $3, %5; \ 00741 ldq $4, %6; \ 00742 " 00743 00744 #define MULADDC_CORE \ 00745 " \ 00746 ldq $6, 0($1); \ 00747 addq $1, 8, $1; \ 00748 mulq $6, $4, $7; \ 00749 umulh $6, $4, $6; \ 00750 addq $7, $3, $7; \ 00751 cmpult $7, $3, $3; \ 00752 ldq $5, 0($2); \ 00753 addq $7, $5, $7; \ 00754 cmpult $7, $5, $5; \ 00755 stq $7, 0($2); \ 00756 addq $2, 8, $2; \ 00757 addq $6, $3, $3; \ 00758 addq $5, $3, $3; \ 00759 " 00760 00761 #define MULADDC_STOP \ 00762 " \ 00763 stq $3, %0; \ 00764 stq $2, %1; \ 00765 stq $1, %2; \ 00766 " \ 00767 : "=m" (c), "=m" (d), "=m" (s) \ 00768 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00769 : "$1", "$2", "$3", "$4", "$5", "$6", "$7" \ 00770 ); 00771 #endif /* Alpha */ 00772 00773 #if defined(__mips__) && !defined(__mips64__) 00774 00775 #define MULADDC_INIT \ 00776 asm( \ 00777 " \ 00778 lw $10, %3; \ 00779 lw $11, %4; \ 00780 lw $12, %5; \ 00781 lw $13, %6; \ 00782 " 00783 00784 #define MULADDC_CORE \ 00785 " \ 00786 lw $14, 0($10); \ 00787 multu $13, $14; \ 00788 addi $10, $10, 4; \ 00789 mflo $14; \ 00790 mfhi $9; \ 00791 addu $14, $12, $14; \ 00792 lw $15, 0($11); \ 00793 sltu $12, $14, $12; \ 00794 addu $15, $14, $15; \ 00795 sltu $14, $15, $14; \ 00796 addu $12, $12, $9; \ 00797 sw $15, 0($11); \ 00798 addu $12, $12, $14; \ 00799 addi $11, $11, 4; \ 00800 " 00801 00802 #define MULADDC_STOP \ 00803 " \ 00804 sw $12, %0; \ 00805 sw $11, %1; \ 00806 sw $10, %2; \ 00807 " \ 00808 : "=m" (c), "=m" (d), "=m" (s) \ 00809 : "m" (s), "m" (d), "m" (c), "m" (b) \ 00810 : "$9", "$10", "$11", "$12", "$13", "$14", "$15" \ 00811 ); 00812 00813 #endif /* MIPS */ 00814 #endif /* GNUC */ 00815 00816 #if (defined(_MSC_VER) && defined(_M_IX86)) || defined(__WATCOMC__) 00817 00818 #define MULADDC_INIT \ 00819 __asm mov esi, s \ 00820 __asm mov edi, d \ 00821 __asm mov ecx, c \ 00822 __asm mov ebx, b 00823 00824 #define MULADDC_CORE \ 00825 __asm lodsd \ 00826 __asm mul ebx \ 00827 __asm add eax, ecx \ 00828 __asm adc edx, 0 \ 00829 __asm add eax, [edi] \ 00830 __asm adc edx, 0 \ 00831 __asm mov ecx, edx \ 00832 __asm stosd 00833 00834 #if defined(POLARSSL_HAVE_SSE2) 00835 00836 #define EMIT __asm _emit 00837 00838 #define MULADDC_HUIT \ 00839 EMIT 0x0F EMIT 0x6E EMIT 0xC9 \ 00840 EMIT 0x0F EMIT 0x6E EMIT 0xC3 \ 00841 EMIT 0x0F EMIT 0x6E EMIT 0x1F \ 00842 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \ 00843 EMIT 0x0F EMIT 0x6E EMIT 0x16 \ 00844 EMIT 0x0F EMIT 0xF4 EMIT 0xD0 \ 00845 EMIT 0x0F EMIT 0x6E EMIT 0x66 EMIT 0x04 \ 00846 EMIT 0x0F EMIT 0xF4 EMIT 0xE0 \ 00847 EMIT 0x0F EMIT 0x6E EMIT 0x76 EMIT 0x08 \ 00848 EMIT 0x0F EMIT 0xF4 EMIT 0xF0 \ 00849 EMIT 0x0F EMIT 0x6E EMIT 0x7E EMIT 0x0C \ 00850 EMIT 0x0F EMIT 0xF4 EMIT 0xF8 \ 00851 EMIT 0x0F EMIT 0xD4 EMIT 0xCA \ 00852 EMIT 0x0F EMIT 0x6E EMIT 0x5F EMIT 0x04 \ 00853 EMIT 0x0F EMIT 0xD4 EMIT 0xDC \ 00854 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x08 \ 00855 EMIT 0x0F EMIT 0xD4 EMIT 0xEE \ 00856 EMIT 0x0F EMIT 0x6E EMIT 0x67 EMIT 0x0C \ 00857 EMIT 0x0F EMIT 0xD4 EMIT 0xFC \ 00858 EMIT 0x0F EMIT 0x7E EMIT 0x0F \ 00859 EMIT 0x0F EMIT 0x6E EMIT 0x56 EMIT 0x10 \ 00860 EMIT 0x0F EMIT 0xF4 EMIT 0xD0 \ 00861 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00862 EMIT 0x0F EMIT 0x6E EMIT 0x66 EMIT 0x14 \ 00863 EMIT 0x0F EMIT 0xF4 EMIT 0xE0 \ 00864 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \ 00865 EMIT 0x0F EMIT 0x6E EMIT 0x76 EMIT 0x18 \ 00866 EMIT 0x0F EMIT 0xF4 EMIT 0xF0 \ 00867 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x04 \ 00868 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00869 EMIT 0x0F EMIT 0x6E EMIT 0x5E EMIT 0x1C \ 00870 EMIT 0x0F EMIT 0xF4 EMIT 0xD8 \ 00871 EMIT 0x0F EMIT 0xD4 EMIT 0xCD \ 00872 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x10 \ 00873 EMIT 0x0F EMIT 0xD4 EMIT 0xD5 \ 00874 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x08 \ 00875 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00876 EMIT 0x0F EMIT 0xD4 EMIT 0xCF \ 00877 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x14 \ 00878 EMIT 0x0F EMIT 0xD4 EMIT 0xE5 \ 00879 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x0C \ 00880 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00881 EMIT 0x0F EMIT 0xD4 EMIT 0xCA \ 00882 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x18 \ 00883 EMIT 0x0F EMIT 0xD4 EMIT 0xF5 \ 00884 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x10 \ 00885 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00886 EMIT 0x0F EMIT 0xD4 EMIT 0xCC \ 00887 EMIT 0x0F EMIT 0x6E EMIT 0x6F EMIT 0x1C \ 00888 EMIT 0x0F EMIT 0xD4 EMIT 0xDD \ 00889 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x14 \ 00890 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00891 EMIT 0x0F EMIT 0xD4 EMIT 0xCE \ 00892 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x18 \ 00893 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00894 EMIT 0x0F EMIT 0xD4 EMIT 0xCB \ 00895 EMIT 0x0F EMIT 0x7E EMIT 0x4F EMIT 0x1C \ 00896 EMIT 0x83 EMIT 0xC7 EMIT 0x20 \ 00897 EMIT 0x83 EMIT 0xC6 EMIT 0x20 \ 00898 EMIT 0x0F EMIT 0x73 EMIT 0xD1 EMIT 0x20 \ 00899 EMIT 0x0F EMIT 0x7E EMIT 0xC9 00900 00901 #define MULADDC_STOP \ 00902 EMIT 0x0F EMIT 0x77 \ 00903 __asm mov c, ecx \ 00904 __asm mov d, edi \ 00905 __asm mov s, esi \ 00906 00907 #else 00908 00909 #define MULADDC_STOP \ 00910 __asm mov c, ecx \ 00911 __asm mov d, edi \ 00912 __asm mov s, esi \ 00913 00914 #endif /* SSE2 */ 00915 #endif /* MSVC */ 00916 00917 #endif /* POLARSSL_HAVE_ASM */ 00918 00919 #if !defined(MULADDC_CORE) 00920 #if defined(POLARSSL_HAVE_UDBL) 00921 00922 #define MULADDC_INIT \ 00923 { \ 00924 t_udbl r; \ 00925 t_uint r0, r1; 00926 00927 #define MULADDC_CORE \ 00928 r = *(s++) * (t_udbl) b; \ 00929 r0 = r; \ 00930 r1 = r >> biL; \ 00931 r0 += c; r1 += (r0 < c); \ 00932 r0 += *d; r1 += (r0 < *d); \ 00933 c = r1; *(d++) = r0; 00934 00935 #define MULADDC_STOP \ 00936 } 00937 00938 #else 00939 #define MULADDC_INIT \ 00940 { \ 00941 t_uint s0, s1, b0, b1; \ 00942 t_uint r0, r1, rx, ry; \ 00943 b0 = ( b << biH ) >> biH; \ 00944 b1 = ( b >> biH ); 00945 00946 #define MULADDC_CORE \ 00947 s0 = ( *s << biH ) >> biH; \ 00948 s1 = ( *s >> biH ); s++; \ 00949 rx = s0 * b1; r0 = s0 * b0; \ 00950 ry = s1 * b0; r1 = s1 * b1; \ 00951 r1 += ( rx >> biH ); \ 00952 r1 += ( ry >> biH ); \ 00953 rx <<= biH; ry <<= biH; \ 00954 r0 += rx; r1 += (r0 < rx); \ 00955 r0 += ry; r1 += (r0 < ry); \ 00956 r0 += c; r1 += (r0 < c); \ 00957 r0 += *d; r1 += (r0 < *d); \ 00958 c = r1; *(d++) = r0; 00959 00960 #define MULADDC_STOP \ 00961 } 00962 00963 #endif /* C (generic) */ 00964 #endif /* C (longlong) */ 00965 00966 #endif /* bn_mul.h */ 00967 00968
Generated on Tue Jul 12 2022 19:40:15 by
1.7.2