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
|
;; rdrand.asm - written and placed in public domain by Jeffrey Walton and Uri Blumenthal.
;; Copyright assigned to the Crypto++ project.
;; This ASM file provides RDRAND and RDSEED to downlevel Unix and Linux tool
;; chains. You will need a modern Nasm, however. You can also use it in place
;; of intrinsics. The routines below run a little faster than the intrinsic
;; based routines.
;; nasm -f elf32 rdrand.s -DX86 -g -o rdrand-x86.o
;; nasm -f elfx32 rdrand.s -DX32 -g -o rdrand-x32.o
;; nasm -f elf64 rdrand.s -DX64 -g -o rdrand-x64.o
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; C/C++ Function prototypes
;; X86, X32 and X64:
;; extern "C" void NASM_RDRAND_GenerateBlock(byte* ptr, size_t size);
;; extern "C" void NASM_RDSEED_GenerateBlock(byte* ptr, size_t size);
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%ifdef X86 ;; Set via the command line
%define arg1 [esp+04h]
%define arg2 [esp+08h]
%define buffer ecx
%define bsize edx
%define lsize dl ;; Used for tail bytes, 1-byte constants
%define MWSIZE 04h ;; machine word size
%elifdef X32 ;; Set via the command line
%define buffer edi ;; Linux ABI
%define bsize esi ;; Linux ABI
%define lsize si
%define MWSIZE 04h ;; machine word size
%elifdef X64 ;; Set via the command line
%ifdef CYGWIN ;; Cygwin follows Windows ABI here, not Linux ABI
%define buffer rcx ;; Windows ABI
%define bsize rdx ;; Windows ABI
%define lsize dx ;; Used for tail bytes, 2-byte constants
%else
%define buffer rdi ;; Linux ABI
%define bsize rsi ;; Linux ABI
%define lsize si ;; Used for tail bytes, 2-byte constants
%endif
%define MWSIZE 08h ;; machine word size
%else
%error Missing or unknown architecture
%endif
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; Fixups
%ifdef DARWIN
%define NASM_RDRAND_GenerateBlock _NASM_RDRAND_GenerateBlock
%define NASM_RDSEED_GenerateBlock _NASM_RDSEED_GenerateBlock
%endif
%ifdef CYGWIN
%ifdef X86
%define NASM_RDRAND_GenerateBlock _NASM_RDRAND_GenerateBlock
%define NASM_RDSEED_GenerateBlock _NASM_RDSEED_GenerateBlock
%endif
%endif
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%ifdef X86 ;; Set via the command line
global NASM_RDRAND_GenerateBlock
section .text
align 8
NASM_RDRAND_GenerateBlock:
.Load_Arguments:
mov buffer, arg1
mov bsize, arg2
;; A block of 16-bytes appears to be optimal. Adding
;; more rdrand calls degrades performance.
cmp bsize, 16
jb .GenerateBlock_4
.GenerateBlock_16:
.Call_RDRAND_EAX_4:
rdrand eax
jnc .Call_RDRAND_EAX_4
mov [buffer+0], eax
.Call_RDRAND_EAX_3:
rdrand eax
jnc .Call_RDRAND_EAX_3
mov [buffer+4], eax
.Call_RDRAND_EAX_2:
rdrand eax
jnc .Call_RDRAND_EAX_2
mov [buffer+8], eax
.Call_RDRAND_EAX_1:
rdrand eax
jnc .Call_RDRAND_EAX_1
mov [buffer+12], eax
sub bsize, 16
add buffer, 16
cmp bsize, 16
jae .GenerateBlock_16
;; Fewer than 16 bytes remain
.GenerateBlock_4:
cmp lsize, 0
je .GenerateBlock_Return
.Call_RDRAND_EAX_0:
rdrand eax
jnc .Call_RDRAND_EAX_0
cmp lsize, MWSIZE
jb .Partial_Machine_Word
.Full_Machine_Word:
mov [buffer], eax
add buffer, MWSIZE
sub lsize, MWSIZE
;; Continue
jmp .GenerateBlock_4
;; 1,2,3 bytes remain
.Partial_Machine_Word:
;; Test bit 1 to see if size is at least 2
test lsize, 2
jz .Bit_1_Not_Set
mov [buffer], ax
shr eax, 16
add buffer, 2
.Bit_1_Not_Set:
;; Test bit 0 to see if size is at least 1
test lsize, 1
jz .Bit_0_Not_Set
mov [buffer], al
.Bit_0_Not_Set:
;; We've hit all the bits
.GenerateBlock_Return:
xor eax, eax
ret
%endif ;; X86
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%ifdef X64 or X32 ;; Set via the command line
global NASM_RDRAND_GenerateBlock
section .text
align 16
NASM_RDRAND_GenerateBlock:
;; No need for Load_Arguments due to fastcall
;; A block of 32-bytes appears to be optimal. Adding
;; more rdrand calls degrades performance.
cmp bsize, 32
jb .GenerateBlock_8
.GenerateBlock_32:
.Call_RDRAND_RAX_4:
rdrand rax
jnc .Call_RDRAND_RAX_4
mov [buffer+0], rax
.Call_RDRAND_RAX_3:
rdrand rax
jnc .Call_RDRAND_RAX_3
mov [buffer+8], rax
.Call_RDRAND_RAX_2:
rdrand rax
jnc .Call_RDRAND_RAX_2
mov [buffer+16], rax
.Call_RDRAND_RAX_1:
rdrand rax
jnc .Call_RDRAND_RAX_1
mov [buffer+24], rax
sub bsize, 32
add buffer, 32
cmp bsize, 32
jae .GenerateBlock_32
;; Fewer than 32 bytes remain
.GenerateBlock_8:
cmp lsize, 0
je .GenerateBlock_Return
.Call_RDRAND_RAX_0:
rdrand rax
jnc .Call_RDRAND_RAX_0
cmp lsize, MWSIZE
jb .Partial_Machine_Word
.Full_Machine_Word:
mov [buffer], rax
add buffer, MWSIZE
sub lsize, MWSIZE
;; Continue
jmp .GenerateBlock_8
;; 1,2,3,4,5,6,7 bytes remain
.Partial_Machine_Word:
;; Test bit 2 to see if size is at least 4
test lsize, 4
jz .Bit_2_Not_Set
mov [buffer], eax
shr rax, 32
add buffer, 4
.Bit_2_Not_Set:
;; Test bit 1 to see if size is at least 2
test lsize, 2
jz .Bit_1_Not_Set
mov [buffer], ax
shr eax, 16
add buffer, 2
.Bit_1_Not_Set:
;; Test bit 0 to see if size is at least 1
test lsize, 1
jz .Bit_0_Not_Set
mov [buffer], al
.Bit_0_Not_Set:
;; We've hit all the bits
.GenerateBlock_Return:
xor rax, rax
ret
%endif ;; X64
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%ifdef X86 ;; Set via the command line
global NASM_RDSEED_GenerateBlock
section .text
align 8
NASM_RDSEED_GenerateBlock:
.Load_Arguments:
mov buffer, arg1
mov bsize, arg2
;; A block of 16-bytes appears to be optimal. Adding
;; more rdrand calls degrades performance.
cmp bsize, 16
jb .GenerateBlock_4
.GenerateBlock_16:
.Call_RDSEED_EAX_4:
rdseed eax
jnc .Call_RDSEED_EAX_4
mov [buffer+0], eax
.Call_RDSEED_EAX_3:
rdseed eax
jnc .Call_RDSEED_EAX_3
mov [buffer+4], eax
.Call_RDSEED_EAX_2:
rdseed eax
jnc .Call_RDSEED_EAX_2
mov [buffer+8], eax
.Call_RDSEED_EAX_1:
rdseed eax
jnc .Call_RDSEED_EAX_1
mov [buffer+12], eax
sub bsize, 16
add buffer, 16
cmp bsize, 16
jae .GenerateBlock_16
;; Fewer than 16 bytes remain
.GenerateBlock_4:
cmp lsize, 0
je .GenerateBlock_Return
.Call_RDSEED_EAX_0:
rdseed eax
jnc .Call_RDSEED_EAX_0
cmp lsize, MWSIZE
jb .Partial_Machine_Word
.Full_Machine_Word:
mov [buffer], eax
add buffer, MWSIZE
sub lsize, MWSIZE
;; Continue
jmp .GenerateBlock_4
;; 1,2,3 bytes remain
.Partial_Machine_Word:
;; Test bit 1 to see if size is at least 2
test lsize, 2
jz .Bit_1_Not_Set
mov [buffer], ax
shr eax, 16
add buffer, 2
.Bit_1_Not_Set:
;; Test bit 0 to see if size is at least 1
test lsize, 1
jz .Bit_0_Not_Set
mov [buffer], al
.Bit_0_Not_Set:
;; We've hit all the bits
.GenerateBlock_Return:
xor eax, eax
ret
%endif ;; X86
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%ifdef X64 or X32 ;; Set via the command line
global NASM_RDSEED_GenerateBlock
section .text
align 16
NASM_RDSEED_GenerateBlock:
;; No need for Load_Arguments due to fastcall
;; A block of 32-bytes appears to be optimal. Adding
;; more rdrand calls degrades performance.
cmp bsize, 32
jb .GenerateBlock_8
.GenerateBlock_32:
.Call_RDSEED_RAX_4:
rdseed rax
jnc .Call_RDSEED_RAX_4
mov [buffer+0], rax
.Call_RDSEED_RAX_3:
rdseed rax
jnc .Call_RDSEED_RAX_3
mov [buffer+8], rax
.Call_RDSEED_RAX_2:
rdseed rax
jnc .Call_RDSEED_RAX_2
mov [buffer+16], rax
.Call_RDSEED_RAX_1:
rdseed rax
jnc .Call_RDSEED_RAX_1
mov [buffer+24], rax
sub bsize, 32
add buffer, 32
cmp bsize, 32
jae .GenerateBlock_32
;; Fewer than 32 bytes remain
.GenerateBlock_8:
cmp lsize, 0
je .GenerateBlock_Return
.Call_RDSEED_RAX_0:
rdseed rax
jnc .Call_RDSEED_RAX_0
cmp lsize, MWSIZE
jb .Partial_Machine_Word
.Full_Machine_Word:
mov [buffer], rax
add buffer, MWSIZE
sub lsize, MWSIZE
;; Continue
jmp .GenerateBlock_8
;; 1,2,3,4,5,6,7 bytes remain
.Partial_Machine_Word:
;; Test bit 2 to see if size is at least 4
test lsize, 4
jz .Bit_2_Not_Set
mov [buffer], eax
shr rax, 32
add buffer, 4
.Bit_2_Not_Set:
;; Test bit 1 to see if size is at least 2
test lsize, 2
jz .Bit_1_Not_Set
mov [buffer], ax
shr eax, 16
add buffer, 2
.Bit_1_Not_Set:
;; Test bit 0 to see if size is at least 1
test lsize, 1
jz .Bit_0_Not_Set
mov [buffer], al
.Bit_0_Not_Set:
;; We've hit all the bits
.GenerateBlock_Return:
xor rax, rax
ret
%endif ;; X64
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
|
