1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
|
/*************************************************
* Lowest Level MPI Algorithms Header File *
* (C) 1999-2006 Jack Lloyd *
* 2006 Luca Piccarreta *
*************************************************/
#ifndef BOTAN_MP_ASM_INTERNAL_H__
#define BOTAN_MP_ASM_INTERNAL_H__
#include "mp_asm.h"
namespace Botan {
extern "C" {
/*************************************************
* Word Addition *
*************************************************/
inline word word_add(word x, word y, word* carry)
{
word z = x + y;
word c1 = (z < x);
z += *carry;
*carry = c1 | (z < *carry);
return z;
}
/*************************************************
* Four Word Block Addition, Two Argument *
*************************************************/
inline word word4_addcarry(word x[4], word carry)
{
__asm {
mov edx,[x]
xor eax,eax
sub eax,[carry] //force CF=1 iff *carry==1
adc [edx],0
mov eax,[esi+4]
adc [edx+4],0
mov eax,[esi+8]
adc [edx+8],0
mov eax,[esi+12]
adc [edx+12],0
sbb eax,eax
neg eax
}
}
/*************************************************
* Four Word Block Addition, Two Argument *
*************************************************/
inline word word8_add2(word x[8], const word y[8], word carry)
{
__asm {
mov edx,[x]
mov esi,[y]
xor eax,eax
sub eax,[carry] //force CF=1 iff *carry==1
mov eax,[esi]
adc [edx],eax
mov eax,[esi+4]
adc [edx+4],eax
mov eax,[esi+8]
adc [edx+8],eax
mov eax,[esi+12]
adc [edx+12],eax
mov eax,[esi+16]
adc [edx+16],eax
mov eax,[esi+20]
adc [edx+20],eax
mov eax,[esi+24]
adc [edx+24],eax
mov eax,[esi+28]
adc [edx+28],eax
sbb eax,eax
neg eax
}
}
/*************************************************
* Four Word Block Addition, Three Argument *
*************************************************/
inline word word8_add3(word z[8], const word x[8], const word y[8], word carry)
{
__asm {
mov edi,[x]
mov esi,[y]
mov ebx,[z]
xor eax,eax
sub eax,[carry] //force CF=1 iff *carry==1
mov eax,[edi]
adc eax,[esi]
mov [ebx],eax
mov eax,[edi+4]
adc eax,[esi+4]
mov [ebx+4],eax
mov eax,[edi+8]
adc eax,[esi+8]
mov [ebx+8],eax
mov eax,[edi+12]
adc eax,[esi+12]
mov [ebx+12],eax
mov eax,[edi+16]
adc eax,[esi+16]
mov [ebx+16],eax
mov eax,[edi+20]
adc eax,[esi+20]
mov [ebx+20],eax
mov eax,[edi+24]
adc eax,[esi+24]
mov [ebx+24],eax
mov eax,[edi+28]
adc eax,[esi+28]
mov [ebx+28],eax
sbb eax,eax
neg eax
}
}
/*************************************************
* Word Subtraction *
*************************************************/
inline word word_sub(word x, word y, word* carry)
{
word t0 = x - y;
word c1 = (t0 > x);
word z = t0 - *carry;
*carry = c1 | (z > t0);
return z;
}
/*************************************************
* Four Word Block Subtraction, Two Argument *
*************************************************/
inline word word8_sub2(word x[8], const word y[8], word carry)
{
_asm {
mov edi,[x]
mov esi,[y]
xor eax,eax
sub eax,[carry] //force CF=1 iff *carry==1
mov eax,[edi]
sbb eax,[esi]
mov [edi],eax
mov eax,[edi+4]
sbb eax,[esi+4]
mov [edi+4],eax
mov eax,[edi+8]
sbb eax,[esi+8]
mov [edi+8],eax
mov eax,[edi+12]
sbb eax,[esi+12]
mov [edi+12],eax
mov eax,[edi+16]
sbb eax,[esi+16]
mov [edi+16],eax
mov eax,[edi+20]
sbb eax,[esi+20]
mov [edi+20],eax
mov eax,[edi+24]
sbb eax,[esi+24]
mov [edi+24],eax
mov eax,[edi+28]
sbb eax,[esi+28]
mov [edi+28],eax
sbb eax,eax
neg eax
}
}
/*************************************************
* Four Word Block Subtraction, Three Argument *
*************************************************/
__forceinline word word8_sub3(word z[8], const word x[8],
const word y[8], word carry)
{
__asm {
mov edi,[x]
mov esi,[y]
xor eax,eax
sub eax,[carry] //force CF=1 iff *carry==1
mov ebx,[z]
mov eax,[edi]
sbb eax,[esi]
mov [ebx],eax
mov eax,[edi+4]
sbb eax,[esi+4]
mov [ebx+4],eax
mov eax,[edi+8]
sbb eax,[esi+8]
mov [ebx+8],eax
mov eax,[edi+12]
sbb eax,[esi+12]
mov [ebx+12],eax
mov eax,[edi+16]
sbb eax,[esi+16]
mov [ebx+16],eax
mov eax,[edi+20]
sbb eax,[esi+20]
mov [ebx+20],eax
mov eax,[edi+24]
sbb eax,[esi+24]
mov [ebx+24],eax
mov eax,[edi+28]
sbb eax,[esi+28]
mov [ebx+28],eax
sbb eax,eax
neg eax
}
}
/*************************************************
* Four Word Block Linear Multiplication *
*************************************************/
inline word word8_linmul2(word x[8], word y, word carry)
{
__asm
{
mov esi,[x]
mov eax,[esi] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,[carry] //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi],eax //load a
mov eax,[esi+4] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+4],eax //load a
mov eax,[esi+8] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+8],eax //load a
mov eax,[esi+12] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+12],eax //load a
mov eax,[esi+16] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+16],eax //load a
mov eax,[esi+20] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+20],eax //load a
mov eax,[esi+24] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [esi+24],eax //load a
mov eax,[esi+28] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov [esi+28],eax //load a
mov eax,edx //store carry
}
}
/*************************************************
* Eight Word Block Linear Multiplication *
*************************************************/
__forceinline word word8_muladd(word z[8], const word x[8],
word y, word carry)
{
__asm
{
mov esi,[x]
mov ebx,[y]
mov edi,[z]
mov eax,[esi] //load a
mul ebx //edx(hi):eax(lo)=a*b
add eax,[carry] //sum lo carry
adc edx,0 //sum hi carry
add eax,[edi] //sum lo z
adc edx,0 //sum hi z
mov ecx,edx //carry for next block = hi z
mov [edi],eax //save lo z
mov eax,[esi+4]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+4]
adc edx,0
mov ecx,edx
mov [edi+4],eax
mov eax,[esi+8]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+8]
adc edx,0
mov ecx,edx
mov [edi+8],eax
mov eax,[esi+12]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+12]
adc edx,0
mov ecx,edx
mov [edi+12],eax
mov eax,[esi+16]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+16]
adc edx,0
mov ecx,edx
mov [edi+16],eax
mov eax,[esi+20]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+20]
adc edx,0
mov ecx,edx
mov [edi+20],eax
mov eax,[esi+24]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+24]
adc edx,0
mov ecx,edx
mov [edi+24],eax
mov eax,[esi+28]
mul ebx
add eax,ecx
adc edx,0
add eax,[edi+28]
adc edx,0
mov [edi+28],eax
mov eax,edx
}
}
__forceinline word word8_linmul3(word z[4], const word x[4], word y, word carry)
{
__asm
{
#if 0
//it's slower!!!
mov edx,[z]
mov eax,[x]
movd mm7,[y]
movd mm0,[eax]
movd mm1,[eax+4]
movd mm2,[eax+8]
pmuludq mm0,mm7
pmuludq mm1,mm7
pmuludq mm2,mm7
movd mm6,[carry]
paddq mm0,mm6
movd [edx],mm0
psrlq mm0,32
paddq mm1,mm0
movd [edx+4],mm1
movd mm3,[eax+12]
psrlq mm1,32
paddq mm2,mm1
movd [edx+8],mm2
pmuludq mm3,mm7
movd mm4,[eax+16]
psrlq mm2,32
paddq mm3,mm2
movd [edx+12],mm3
pmuludq mm4,mm7
movd mm5,[eax+20]
psrlq mm3,32
paddq mm4,mm3
movd [edx+16],mm4
pmuludq mm5,mm7
movd mm0,[eax+24]
psrlq mm4,32
paddq mm5,mm4
movd [edx+20],mm5
pmuludq mm0,mm7
movd mm1,[eax+28]
psrlq mm5,32
paddq mm0,mm5
movd [edx+24],mm0
pmuludq mm1,mm7
psrlq mm0,32
paddq mm1,mm0
movd [edx+28],mm1
psrlq mm1,32
movd eax,mm1
emms
#else
mov edi,[z]
mov esi,[x]
mov eax,[esi] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,[carry] //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi],eax //load a
mov eax,[esi+4] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+4],eax //load a
mov eax,[esi+8] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+8],eax //load a
mov eax,[esi+12] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+12],eax //load a
mov eax,[esi+16] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+16],eax //load a
mov eax,[esi+20] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+20],eax //load a
mov eax,[esi+24] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov ecx,edx //store carry
mov [edi+24],eax //load a
mov eax,[esi+28] //load a
mul [y] //edx(hi):eax(lo)=a*b
add eax,ecx //sum lo carry
adc edx,0 //sum hi carry
mov [edi+28],eax //load a
mov eax,edx //store carry
#endif
}
}
/*************************************************
* Eight Word Block Multiply/Add *
*************************************************/
inline word word8_madd3(word z[8], const word x[8], word y, word carry)
{
z[0] = word_madd3(x[0], y, z[0], &carry);
z[1] = word_madd3(x[1], y, z[1], &carry);
z[2] = word_madd3(x[2], y, z[2], &carry);
z[3] = word_madd3(x[3], y, z[3], &carry);
z[4] = word_madd3(x[4], y, z[4], &carry);
z[5] = word_madd3(x[5], y, z[5], &carry);
z[6] = word_madd3(x[6], y, z[6], &carry);
z[7] = word_madd3(x[7], y, z[7], &carry);
return carry;
}
/*************************************************
* Multiply-Add Accumulator *
*************************************************/
inline void word3_muladd(word* w2, word* w1, word* w0, word a, word b)
{
dword z = (dword)a * b + (*w0);
*w0 = (word)z; //lo
word t1 = (word)(z >> BOTAN_MP_WORD_BITS); //hi
*w1 += t1; //w1+=lo
*w2 += (*w1 < t1) ? 1 : 0; //w2+=carry
}
/*************************************************
* Multiply-Add Accumulator *
*************************************************/
inline void word3_muladd_2(word* w2, word* w1, word* w0, word a, word b)
{
dword z = (dword)a * b;
word t0 = (word)z;
word t1 = (word)(z >> BOTAN_MP_WORD_BITS);
*w0 += t0;
*w1 += t1 + ((*w0 < t0) ? 1 : 0);
*w2 += (*w1 < t1) ? 1 : 0;
*w0 += t0;
*w1 += t1 + ((*w0 < t0) ? 1 : 0);
*w2 += (*w1 < t1) ? 1 : 0;
}
}
}
#endif
|
