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
|
;;
;; Copyright (c) 2012-2019, Intel Corporation
;;
;; Redistribution and use in source and binary forms, with or without
;; modification, are permitted provided that the following conditions are met:
;;
;; * Redistributions of source code must retain the above copyright notice,
;; this list of conditions and the following disclaimer.
;; * Redistributions in binary form must reproduce the above copyright
;; notice, this list of conditions and the following disclaimer in the
;; documentation and/or other materials provided with the distribution.
;; * Neither the name of Intel Corporation nor the names of its contributors
;; may be used to endorse or promote products derived from this software
;; without specific prior written permission.
;;
;; THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
;; AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
;; IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
;; DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
;; FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
;; DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
;; SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
;; CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
;; OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
;; OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
;;
%include "include/os.asm"
%include "job_aes_hmac.asm"
%include "include/memcpy.asm"
%include "include/const.inc"
%include "include/reg_sizes.asm"
; routine to do AES128 CNTR enc/decrypt "by4"
; XMM registers are clobbered. Saving/restoring must be done at a higher level
%ifndef AES_CNTR_128
%define AES_CNTR_128 aes_cntr_128_sse
%define AES_CNTR_BIT_128 aes_cntr_bit_128_sse
%endif
extern byteswap_const, set_byte15, ddq_add_1, ddq_add_2, ddq_add_3, ddq_add_4
%define CONCAT(a,b) a %+ b
%define MOVDQ movdqu
%define xdata0 xmm0
%define xdata1 xmm1
%define xpart xmm1
%define xdata2 xmm2
%define xdata3 xmm3
%define xdata4 xmm4
%define xdata5 xmm5
%define xdata6 xmm6
%define xdata7 xmm7
%define xcounter xmm8
%define xtmp xmm8
%define xbyteswap xmm9
%define xtmp2 xmm9
%define xkey0 xmm10
%define xtmp3 xmm10
%define xkey3 xmm11
%define xkey6 xmm12
%define xkey9 xmm13
%define xkeyA xmm14
%define xkeyB xmm15
%ifdef CNTR_CCM_SSE
%ifdef LINUX
%define job rdi
%define p_in rsi
%define p_keys rdx
%define p_out rcx
%define num_bytes r8
%define p_ivlen r9
%else ;; LINUX
%define job rcx
%define p_in rdx
%define p_keys r8
%define p_out r9
%define num_bytes r10
%define p_ivlen rax
%endif ;; LINUX
%define p_IV r11
%else ;; CNTR_CCM_SSE
%ifdef LINUX
%define p_in rdi
%define p_IV rsi
%define p_keys rdx
%define p_out rcx
%define num_bytes r8
%define num_bits r8
%define p_ivlen r9
%else ;; LINUX
%define p_in rcx
%define p_IV rdx
%define p_keys r8
%define p_out r9
%define num_bytes r10
%define num_bits r10
%define p_ivlen qword [rsp + 8*6]
%endif ;; LINUX
%endif ;; CNTR_CCM_SSE
%define tmp r11
%define flags r11
%define r_bits r12
%define tmp2 r13
%define mask r14
%macro do_aes_load 2
do_aes %1, %2, 1
%endmacro
%macro do_aes_noload 2
do_aes %1, %2, 0
%endmacro
; do_aes num_in_par load_keys
; This increments p_in, but not p_out
%macro do_aes 3
%define %%by %1
%define %%cntr_type %2
%define %%load_keys %3
%if (%%load_keys)
movdqa xkey0, [p_keys + 0*16]
%endif
movdqa xdata0, xcounter
pshufb xdata0, xbyteswap
%assign i 1
%rep (%%by - 1)
movdqa CONCAT(xdata,i), xcounter
paddd CONCAT(xdata,i), [rel CONCAT(ddq_add_,i)]
pshufb CONCAT(xdata,i), xbyteswap
%assign i (i + 1)
%endrep
movdqa xkeyA, [p_keys + 1*16]
pxor xdata0, xkey0
%ifidn %%cntr_type, CNTR_BIT
paddq xcounter, [rel CONCAT(ddq_add_,%%by)]
%else
paddd xcounter, [rel CONCAT(ddq_add_,%%by)]
%endif
%assign i 1
%rep (%%by - 1)
pxor CONCAT(xdata,i), xkey0
%assign i (i + 1)
%endrep
movdqa xkeyB, [p_keys + 2*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyA ; key 1
%assign i (i+1)
%endrep
%if (%%load_keys)
movdqa xkey3, [p_keys + 3*16]
%endif
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyB ; key 2
%assign i (i+1)
%endrep
add p_in, 16*%%by
movdqa xkeyB, [p_keys + 4*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkey3 ; key 3
%assign i (i+1)
%endrep
movdqa xkeyA, [p_keys + 5*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyB ; key 4
%assign i (i+1)
%endrep
%if (%%load_keys)
movdqa xkey6, [p_keys + 6*16]
%endif
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyA ; key 5
%assign i (i+1)
%endrep
movdqa xkeyA, [p_keys + 7*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkey6 ; key 6
%assign i (i+1)
%endrep
movdqa xkeyB, [p_keys + 8*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyA ; key 7
%assign i (i+1)
%endrep
%if (%%load_keys)
movdqa xkey9, [p_keys + 9*16]
%endif
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkeyB ; key 8
%assign i (i+1)
%endrep
movdqa xkeyB, [p_keys + 10*16]
%assign i 0
%rep %%by
aesenc CONCAT(xdata,i), xkey9 ; key 9
%assign i (i+1)
%endrep
%assign i 0
%rep %%by
aesenclast CONCAT(xdata,i), xkeyB ; key 10
%assign i (i+1)
%endrep
%assign i 0
%rep (%%by / 2)
%assign j (i+1)
MOVDQ xkeyA, [p_in + i*16 - 16*%%by]
MOVDQ xkeyB, [p_in + j*16 - 16*%%by]
pxor CONCAT(xdata,i), xkeyA
pxor CONCAT(xdata,j), xkeyB
%assign i (i+2)
%endrep
%if (i < %%by)
MOVDQ xkeyA, [p_in + i*16 - 16*%%by]
pxor CONCAT(xdata,i), xkeyA
%endif
%ifidn %%cntr_type, CNTR_BIT
;; check if this is the end of the message
mov tmp, num_bytes
and tmp, ~(%%by*16)
jnz %%skip_preserve
;; Check if there is a partial byte
or r_bits, r_bits
jz %%skip_preserve
%assign idx (%%by - 1)
;; Load output to get last partial byte
movdqu xtmp, [p_out + idx * 16]
;; Save RCX in temporary GP register
mov tmp, rcx
mov mask, 0xff
mov cl, BYTE(r_bits)
shr mask, cl ;; e.g. 3 remaining bits -> mask = 00011111
mov rcx, tmp
movq xtmp2, mask
pslldq xtmp2, 15
;; At this point, xtmp2 contains a mask with all 0s, but with some ones
;; in the partial byte
;; Clear all the bits that do not need to be preserved from the output
pand xtmp, xtmp2
;; Clear all bits from the input that are not to be ciphered
pandn xtmp2, CONCAT(xdata, idx)
por xtmp2, xtmp
movdqa CONCAT(xdata, idx), xtmp2
%%skip_preserve:
%endif
%assign i 0
%rep %%by
MOVDQ [p_out + i*16], CONCAT(xdata,i)
%assign i (i+1)
%endrep
%endmacro
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
section .text
;; Macro performing AES-CTR.
;;
%macro DO_CNTR 1
%define %%CNTR_TYPE %1 ; [in] Type of CNTR operation to do (CNTR/CNTR_BIT/CCM)
%ifidn %%CNTR_TYPE, CCM
mov p_in, [job + _src]
add p_in, [job + _cipher_start_src_offset_in_bytes]
mov p_ivlen, [job + _iv_len_in_bytes]
mov num_bytes, [job + _msg_len_to_cipher_in_bytes]
mov p_keys, [job + _aes_enc_key_expanded]
mov p_out, [job + _dst]
movdqa xbyteswap, [rel byteswap_const]
;; Prepare IV ;;
;; Byte 0: flags with L'
;; Calculate L' = 15 - Nonce length - 1 = 14 - IV length
mov flags, 14
sub flags, p_ivlen
movd xcounter, DWORD(flags)
;; Bytes 1 - 13: Nonce (7 - 13 bytes long)
;; Bytes 1 - 7 are always copied (first 7 bytes)
mov p_IV, [job + _iv]
pinsrb xcounter, [p_IV], 1
pinsrw xcounter, [p_IV + 1], 1
pinsrd xcounter, [p_IV + 3], 1
cmp p_ivlen, 7
je _finish_nonce_move
cmp p_ivlen, 8
je _iv_length_8
cmp p_ivlen, 9
je _iv_length_9
cmp p_ivlen, 10
je _iv_length_10
cmp p_ivlen, 11
je _iv_length_11
cmp p_ivlen, 12
je _iv_length_12
;; Bytes 8 - 13
_iv_length_13:
pinsrb xcounter, [p_IV + 12], 13
_iv_length_12:
pinsrb xcounter, [p_IV + 11], 12
_iv_length_11:
pinsrd xcounter, [p_IV + 7], 2
jmp _finish_nonce_move
_iv_length_10:
pinsrb xcounter, [p_IV + 9], 10
_iv_length_9:
pinsrb xcounter, [p_IV + 8], 9
_iv_length_8:
pinsrb xcounter, [p_IV + 7], 8
_finish_nonce_move:
; last byte = 1
por xcounter, [rel set_byte15]
%else ;; CNTR/CNTR_BIT
%ifndef LINUX
mov num_bytes, [rsp + 8*5] ; arg5
%endif
%ifidn %%CNTR_TYPE, CNTR_BIT
push r12
push r13
push r14
%endif
movdqa xbyteswap, [rel byteswap_const]
%ifidn %%CNTR_TYPE, CNTR
test p_ivlen, 16
jnz %%iv_is_16_bytes
; Read 12 bytes: Nonce + ESP IV. Then pad with block counter 0x00000001
mov DWORD(tmp), 0x01000000
pinsrq xcounter, [p_IV], 0
pinsrd xcounter, [p_IV + 8], 2
pinsrd xcounter, DWORD(tmp), 3
%else ;; CNTR_BIT
; Read 16 byte IV: Nonce + 8-byte block counter (BE)
movdqu xcounter, [p_IV]
%endif
%endif ;; CNTR/CNTR_BIT/CCM
%%bswap_iv:
pshufb xcounter, xbyteswap
;; calculate len
;; convert bits to bytes (message length in bits for CNTR_BIT)
%ifidn %%CNTR_TYPE, CNTR_BIT
mov r_bits, num_bits
add num_bits, 7
shr num_bits, 3 ; "num_bits" and "num_bytes" registers are the same
and r_bits, 7 ; Check if there are remainder bits (0-7)
%endif
mov tmp, num_bytes
and tmp, 3*16
jz %%chk ; x4 > or < 15 (not 3 lines)
; 1 <= tmp <= 3
cmp tmp, 2*16
jg %%eq3
je %%eq2
%%eq1:
do_aes_load 1, %%CNTR_TYPE ; 1 block
add p_out, 1*16
jmp %%chk
%%eq2:
do_aes_load 2, %%CNTR_TYPE ; 2 blocks
add p_out, 2*16
jmp %%chk
%%eq3:
do_aes_load 3, %%CNTR_TYPE ; 3 blocks
add p_out, 3*16
; fall through to chk
%%chk:
and num_bytes, ~(3*16)
jz %%do_return2
cmp num_bytes, 16
jb %%last
; process multiples of 4 blocks
movdqa xkey0, [p_keys + 0*16]
movdqa xkey3, [p_keys + 3*16]
movdqa xkey6, [p_keys + 6*16]
movdqa xkey9, [p_keys + 9*16]
align 32
%%main_loop2:
; num_bytes is a multiple of 4 blocks + partial bytes
do_aes_noload 4, %%CNTR_TYPE
add p_out, 4*16
sub num_bytes, 4*16
cmp num_bytes, 4*16
jae %%main_loop2
; Check if there is a partial block
or num_bytes, num_bytes
jnz %%last
%%do_return2:
%ifidn %%CNTR_TYPE, CCM
mov rax, job
or dword [rax + _status], STS_COMPLETED_AES
%endif
%ifidn %%CNTR_TYPE, CNTR_BIT
pop r14
pop r13
pop r12
%endif
ret
%%last:
; load partial block into XMM register
simd_load_sse_15_1 xpart, p_in, num_bytes
%%final_ctr_enc:
; Encryption of a single partial block
pshufb xcounter, xbyteswap
movdqa xdata0, xcounter
pxor xdata0, [p_keys + 16*0]
%assign i 1
%rep 9
aesenc xdata0, [p_keys + 16*i]
%assign i (i+1)
%endrep
; created keystream
aesenclast xdata0, [p_keys + 16*i]
; xor keystream with the message (scratch)
pxor xdata0, xpart
%ifidn %%CNTR_TYPE, CNTR_BIT
;; Check if there is a partial byte
or r_bits, r_bits
jz %%store_output
;; Load output to get last partial byte
simd_load_sse_15_1 xtmp, p_out, num_bytes
;; Save RCX in temporary GP register
mov tmp, rcx
mov mask, 0xff
%ifidn r_bits, rcx
%error "r_bits cannot be mapped to rcx!"
%endif
mov cl, BYTE(r_bits)
shr mask, cl ;; e.g. 3 remaining bits -> mask = 00011111
mov rcx, tmp
movq xtmp2, mask
;; Get number of full bytes in last block of 16 bytes
mov tmp, num_bytes
dec tmp
XPSLLB xtmp2, tmp, xtmp3, tmp2
;; At this point, xtmp2 contains a mask with all 0s, but with some ones
;; in the partial byte
;; Clear all the bits that do not need to be preserved from the output
pand xtmp, xtmp2
;; Clear the bits from the input that are not to be ciphered
pandn xtmp2, xdata0
por xtmp2, xtmp
movdqa xdata0, xtmp2
%endif
%%store_output:
; copy result into the output buffer
simd_store_sse_15 p_out, xdata0, num_bytes, tmp, rax
jmp %%do_return2
%%iv_is_16_bytes:
; Read 16 byte IV: Nonce + ESP IV + block counter (BE)
movdqu xcounter, [p_IV]
jmp %%bswap_iv
%endmacro
align 32
%ifdef CNTR_CCM_SSE
; JOB_AES_HMAC * aes_cntr_ccm_128_sse(JOB_AES_HMAC *job)
; arg 1 : job
MKGLOBAL(AES_CNTR_CCM_128,function,internal)
AES_CNTR_CCM_128:
DO_CNTR CCM
%else
;; aes_cntr_128_sse(void *in, void *IV, void *keys, void *out, UINT64 num_bytes, UINT64 iv_len)
MKGLOBAL(AES_CNTR_128,function,internal)
AES_CNTR_128:
DO_CNTR CNTR
;; aes_cntr_bit_128_sse(void *in, void *IV, void *keys, void *out, UINT64 num_bits, UINT64 iv_len)
MKGLOBAL(AES_CNTR_BIT_128,function,internal)
AES_CNTR_BIT_128:
DO_CNTR CNTR_BIT
%endif ;; CNTR_CCM_SSE
%ifdef LINUX
section .note.GNU-stack noalloc noexec nowrite progbits
%endif
|