1 /* SPDX-License-Identifier: GPL-2.0-or-later *
2 /*
3 * x86_64/AVX2 assembler optimized version of
4 *
5 * Copyright © 2012-2013 Jussi Kivilinna <juss
6 *
7 * Based on AVX assembler implementation of Se
8 * Copyright © 2012 Johannes Goetzfried
9 * <Johannes.Goetzfried@informatik.stud.un
10 */
11
12 #include <linux/linkage.h>
13 #include <asm/frame.h>
14 #include "glue_helper-asm-avx2.S"
15
16 .file "serpent-avx2-asm_64.S"
17
18 .section .rodata.cst16.bswap128_mask, "
19 .align 16
20 .Lbswap128_mask:
21 .byte 15, 14, 13, 12, 11, 10, 9, 8, 7,
22
23 .text
24
25 #define CTX %rdi
26
27 #define RNOT %ymm0
28 #define tp %ymm1
29
30 #define RA1 %ymm2
31 #define RA2 %ymm3
32 #define RB1 %ymm4
33 #define RB2 %ymm5
34 #define RC1 %ymm6
35 #define RC2 %ymm7
36 #define RD1 %ymm8
37 #define RD2 %ymm9
38 #define RE1 %ymm10
39 #define RE2 %ymm11
40
41 #define RK0 %ymm12
42 #define RK1 %ymm13
43 #define RK2 %ymm14
44 #define RK3 %ymm15
45
46 #define RK0x %xmm12
47 #define RK1x %xmm13
48 #define RK2x %xmm14
49 #define RK3x %xmm15
50
51 #define S0_1(x0, x1, x2, x3, x4) \
52 vpor x0, x3, tp; \
53 vpxor x3, x0, x0; \
54 vpxor x2, x3, x4; \
55 vpxor RNOT, x4, x4; \
56 vpxor x1, tp, x3; \
57 vpand x0, x1, x1; \
58 vpxor x4, x1, x1; \
59 vpxor x0, x2, x2;
60 #define S0_2(x0, x1, x2, x3, x4) \
61 vpxor x3, x0, x0; \
62 vpor x0, x4, x4; \
63 vpxor x2, x0, x0; \
64 vpand x1, x2, x2; \
65 vpxor x2, x3, x3; \
66 vpxor RNOT, x1, x1; \
67 vpxor x4, x2, x2; \
68 vpxor x2, x1, x1;
69
70 #define S1_1(x0, x1, x2, x3, x4) \
71 vpxor x0, x1, tp; \
72 vpxor x3, x0, x0; \
73 vpxor RNOT, x3, x3; \
74 vpand tp, x1, x4; \
75 vpor tp, x0, x0; \
76 vpxor x2, x3, x3; \
77 vpxor x3, x0, x0; \
78 vpxor x3, tp, x1;
79 #define S1_2(x0, x1, x2, x3, x4) \
80 vpxor x4, x3, x3; \
81 vpor x4, x1, x1; \
82 vpxor x2, x4, x4; \
83 vpand x0, x2, x2; \
84 vpxor x1, x2, x2; \
85 vpor x0, x1, x1; \
86 vpxor RNOT, x0, x0; \
87 vpxor x2, x0, x0; \
88 vpxor x1, x4, x4;
89
90 #define S2_1(x0, x1, x2, x3, x4) \
91 vpxor RNOT, x3, x3; \
92 vpxor x0, x1, x1; \
93 vpand x2, x0, tp; \
94 vpxor x3, tp, tp; \
95 vpor x0, x3, x3; \
96 vpxor x1, x2, x2; \
97 vpxor x1, x3, x3; \
98 vpand tp, x1, x1;
99 #define S2_2(x0, x1, x2, x3, x4) \
100 vpxor x2, tp, tp; \
101 vpand x3, x2, x2; \
102 vpor x1, x3, x3; \
103 vpxor RNOT, tp, tp; \
104 vpxor tp, x3, x3; \
105 vpxor tp, x0, x4; \
106 vpxor x2, tp, x0; \
107 vpor x2, x1, x1;
108
109 #define S3_1(x0, x1, x2, x3, x4) \
110 vpxor x3, x1, tp; \
111 vpor x0, x3, x3; \
112 vpand x0, x1, x4; \
113 vpxor x2, x0, x0; \
114 vpxor tp, x2, x2; \
115 vpand x3, tp, x1; \
116 vpxor x3, x2, x2; \
117 vpor x4, x0, x0; \
118 vpxor x3, x4, x4;
119 #define S3_2(x0, x1, x2, x3, x4) \
120 vpxor x0, x1, x1; \
121 vpand x3, x0, x0; \
122 vpand x4, x3, x3; \
123 vpxor x2, x3, x3; \
124 vpor x1, x4, x4; \
125 vpand x1, x2, x2; \
126 vpxor x3, x4, x4; \
127 vpxor x3, x0, x0; \
128 vpxor x2, x3, x3;
129
130 #define S4_1(x0, x1, x2, x3, x4) \
131 vpand x0, x3, tp; \
132 vpxor x3, x0, x0; \
133 vpxor x2, tp, tp; \
134 vpor x3, x2, x2; \
135 vpxor x1, x0, x0; \
136 vpxor tp, x3, x4; \
137 vpor x0, x2, x2; \
138 vpxor x1, x2, x2;
139 #define S4_2(x0, x1, x2, x3, x4) \
140 vpand x0, x1, x1; \
141 vpxor x4, x1, x1; \
142 vpand x2, x4, x4; \
143 vpxor tp, x2, x2; \
144 vpxor x0, x4, x4; \
145 vpor x1, tp, x3; \
146 vpxor RNOT, x1, x1; \
147 vpxor x0, x3, x3;
148
149 #define S5_1(x0, x1, x2, x3, x4) \
150 vpor x0, x1, tp; \
151 vpxor tp, x2, x2; \
152 vpxor RNOT, x3, x3; \
153 vpxor x0, x1, x4; \
154 vpxor x2, x0, x0; \
155 vpand x4, tp, x1; \
156 vpor x3, x4, x4; \
157 vpxor x0, x4, x4;
158 #define S5_2(x0, x1, x2, x3, x4) \
159 vpand x3, x0, x0; \
160 vpxor x3, x1, x1; \
161 vpxor x2, x3, x3; \
162 vpxor x1, x0, x0; \
163 vpand x4, x2, x2; \
164 vpxor x2, x1, x1; \
165 vpand x0, x2, x2; \
166 vpxor x2, x3, x3;
167
168 #define S6_1(x0, x1, x2, x3, x4) \
169 vpxor x0, x3, x3; \
170 vpxor x2, x1, tp; \
171 vpxor x0, x2, x2; \
172 vpand x3, x0, x0; \
173 vpor x3, tp, tp; \
174 vpxor RNOT, x1, x4; \
175 vpxor tp, x0, x0; \
176 vpxor x2, tp, x1;
177 #define S6_2(x0, x1, x2, x3, x4) \
178 vpxor x4, x3, x3; \
179 vpxor x0, x4, x4; \
180 vpand x0, x2, x2; \
181 vpxor x1, x4, x4; \
182 vpxor x3, x2, x2; \
183 vpand x1, x3, x3; \
184 vpxor x0, x3, x3; \
185 vpxor x2, x1, x1;
186
187 #define S7_1(x0, x1, x2, x3, x4) \
188 vpxor RNOT, x1, tp; \
189 vpxor RNOT, x0, x0; \
190 vpand x2, tp, x1; \
191 vpxor x3, x1, x1; \
192 vpor tp, x3, x3; \
193 vpxor x2, tp, x4; \
194 vpxor x3, x2, x2; \
195 vpxor x0, x3, x3; \
196 vpor x1, x0, x0;
197 #define S7_2(x0, x1, x2, x3, x4) \
198 vpand x0, x2, x2; \
199 vpxor x4, x0, x0; \
200 vpxor x3, x4, x4; \
201 vpand x0, x3, x3; \
202 vpxor x1, x4, x4; \
203 vpxor x4, x2, x2; \
204 vpxor x1, x3, x3; \
205 vpor x0, x4, x4; \
206 vpxor x1, x4, x4;
207
208 #define SI0_1(x0, x1, x2, x3, x4) \
209 vpxor x0, x1, x1; \
210 vpor x1, x3, tp; \
211 vpxor x1, x3, x4; \
212 vpxor RNOT, x0, x0; \
213 vpxor tp, x2, x2; \
214 vpxor x0, tp, x3; \
215 vpand x1, x0, x0; \
216 vpxor x2, x0, x0;
217 #define SI0_2(x0, x1, x2, x3, x4) \
218 vpand x3, x2, x2; \
219 vpxor x4, x3, x3; \
220 vpxor x3, x2, x2; \
221 vpxor x3, x1, x1; \
222 vpand x0, x3, x3; \
223 vpxor x0, x1, x1; \
224 vpxor x2, x0, x0; \
225 vpxor x3, x4, x4;
226
227 #define SI1_1(x0, x1, x2, x3, x4) \
228 vpxor x3, x1, x1; \
229 vpxor x2, x0, tp; \
230 vpxor RNOT, x2, x2; \
231 vpor x1, x0, x4; \
232 vpxor x3, x4, x4; \
233 vpand x1, x3, x3; \
234 vpxor x2, x1, x1; \
235 vpand x4, x2, x2;
236 #define SI1_2(x0, x1, x2, x3, x4) \
237 vpxor x1, x4, x4; \
238 vpor x3, x1, x1; \
239 vpxor tp, x3, x3; \
240 vpxor tp, x2, x2; \
241 vpor x4, tp, x0; \
242 vpxor x4, x2, x2; \
243 vpxor x0, x1, x1; \
244 vpxor x1, x4, x4;
245
246 #define SI2_1(x0, x1, x2, x3, x4) \
247 vpxor x1, x2, x2; \
248 vpxor RNOT, x3, tp; \
249 vpor x2, tp, tp; \
250 vpxor x3, x2, x2; \
251 vpxor x0, x3, x4; \
252 vpxor x1, tp, x3; \
253 vpor x2, x1, x1; \
254 vpxor x0, x2, x2;
255 #define SI2_2(x0, x1, x2, x3, x4) \
256 vpxor x4, x1, x1; \
257 vpor x3, x4, x4; \
258 vpxor x3, x2, x2; \
259 vpxor x2, x4, x4; \
260 vpand x1, x2, x2; \
261 vpxor x3, x2, x2; \
262 vpxor x4, x3, x3; \
263 vpxor x0, x4, x4;
264
265 #define SI3_1(x0, x1, x2, x3, x4) \
266 vpxor x1, x2, x2; \
267 vpand x2, x1, tp; \
268 vpxor x0, tp, tp; \
269 vpor x1, x0, x0; \
270 vpxor x3, x1, x4; \
271 vpxor x3, x0, x0; \
272 vpor tp, x3, x3; \
273 vpxor x2, tp, x1;
274 #define SI3_2(x0, x1, x2, x3, x4) \
275 vpxor x3, x1, x1; \
276 vpxor x2, x0, x0; \
277 vpxor x3, x2, x2; \
278 vpand x1, x3, x3; \
279 vpxor x0, x1, x1; \
280 vpand x2, x0, x0; \
281 vpxor x3, x4, x4; \
282 vpxor x0, x3, x3; \
283 vpxor x1, x0, x0;
284
285 #define SI4_1(x0, x1, x2, x3, x4) \
286 vpxor x3, x2, x2; \
287 vpand x1, x0, tp; \
288 vpxor x2, tp, tp; \
289 vpor x3, x2, x2; \
290 vpxor RNOT, x0, x4; \
291 vpxor tp, x1, x1; \
292 vpxor x2, tp, x0; \
293 vpand x4, x2, x2;
294 #define SI4_2(x0, x1, x2, x3, x4) \
295 vpxor x0, x2, x2; \
296 vpor x4, x0, x0; \
297 vpxor x3, x0, x0; \
298 vpand x2, x3, x3; \
299 vpxor x3, x4, x4; \
300 vpxor x1, x3, x3; \
301 vpand x0, x1, x1; \
302 vpxor x1, x4, x4; \
303 vpxor x3, x0, x0;
304
305 #define SI5_1(x0, x1, x2, x3, x4) \
306 vpor x2, x1, tp; \
307 vpxor x1, x2, x2; \
308 vpxor x3, tp, tp; \
309 vpand x1, x3, x3; \
310 vpxor x3, x2, x2; \
311 vpor x0, x3, x3; \
312 vpxor RNOT, x0, x0; \
313 vpxor x2, x3, x3; \
314 vpor x0, x2, x2;
315 #define SI5_2(x0, x1, x2, x3, x4) \
316 vpxor tp, x1, x4; \
317 vpxor x4, x2, x2; \
318 vpand x0, x4, x4; \
319 vpxor tp, x0, x0; \
320 vpxor x3, tp, x1; \
321 vpand x2, x0, x0; \
322 vpxor x3, x2, x2; \
323 vpxor x2, x0, x0; \
324 vpxor x4, x2, x2; \
325 vpxor x3, x4, x4;
326
327 #define SI6_1(x0, x1, x2, x3, x4) \
328 vpxor x2, x0, x0; \
329 vpand x3, x0, tp; \
330 vpxor x3, x2, x2; \
331 vpxor x2, tp, tp; \
332 vpxor x1, x3, x3; \
333 vpor x0, x2, x2; \
334 vpxor x3, x2, x2; \
335 vpand tp, x3, x3;
336 #define SI6_2(x0, x1, x2, x3, x4) \
337 vpxor RNOT, tp, tp; \
338 vpxor x1, x3, x3; \
339 vpand x2, x1, x1; \
340 vpxor tp, x0, x4; \
341 vpxor x4, x3, x3; \
342 vpxor x2, x4, x4; \
343 vpxor x1, tp, x0; \
344 vpxor x0, x2, x2;
345
346 #define SI7_1(x0, x1, x2, x3, x4) \
347 vpand x0, x3, tp; \
348 vpxor x2, x0, x0; \
349 vpor x3, x2, x2; \
350 vpxor x1, x3, x4; \
351 vpxor RNOT, x0, x0; \
352 vpor tp, x1, x1; \
353 vpxor x0, x4, x4; \
354 vpand x2, x0, x0; \
355 vpxor x1, x0, x0;
356 #define SI7_2(x0, x1, x2, x3, x4) \
357 vpand x2, x1, x1; \
358 vpxor x2, tp, x3; \
359 vpxor x3, x4, x4; \
360 vpand x3, x2, x2; \
361 vpor x0, x3, x3; \
362 vpxor x4, x1, x1; \
363 vpxor x4, x3, x3; \
364 vpand x0, x4, x4; \
365 vpxor x2, x4, x4;
366
367 #define get_key(i,j,t) \
368 vpbroadcastd (4*(i)+(j))*4(CTX), t;
369
370 #define K2(x0, x1, x2, x3, x4, i) \
371 get_key(i, 0, RK0); \
372 get_key(i, 1, RK1); \
373 get_key(i, 2, RK2); \
374 get_key(i, 3, RK3); \
375 vpxor RK0, x0 ## 1, x0 ## 1; \
376 vpxor RK1, x1 ## 1, x1 ## 1; \
377 vpxor RK2, x2 ## 1, x2 ## 1; \
378 vpxor RK3, x3 ## 1, x3 ## 1; \
379 vpxor RK0, x0 ## 2, x0 ##
380 vpxor RK1, x1 ## 2, x1 ##
381 vpxor RK2, x2 ## 2, x2 ##
382 vpxor RK3, x3 ## 2, x3 ##
383
384 #define LK2(x0, x1, x2, x3, x4, i) \
385 vpslld $13, x0 ## 1, x4 ##
386 vpsrld $(32 - 13), x0 ## 1, x0 ##
387 vpor x4 ## 1, x0 ##
388 vpxor x0 ## 1, x1 ##
389 vpslld $3, x2 ## 1, x4 ##
390 vpsrld $(32 - 3), x2 ## 1, x2 ##
391 vpor x4 ## 1, x2 ##
392 vpxor x2 ## 1, x1 ##
393 vpslld $13, x0 ##
394 vpsrld $(32 - 13), x0 ##
395 vpor x4 ##
396 vpxor x0 ##
397 vpslld $3, x2 ##
398 vpsrld $(32 - 3), x2 ##
399 vpor x4 ##
400 vpxor x2 ##
401 vpslld $1, x1 ## 1, x4 ##
402 vpsrld $(32 - 1), x1 ## 1, x1 ##
403 vpor x4 ## 1, x1 ##
404 vpslld $3, x0 ## 1, x4 ##
405 vpxor x2 ## 1, x3 ##
406 vpxor x4 ## 1, x3 ##
407 get_key(i, 1, RK1); \
408 vpslld $1, x1 ##
409 vpsrld $(32 - 1), x1 ##
410 vpor x4 ##
411 vpslld $3, x0 ##
412 vpxor x2 ##
413 vpxor x4 ##
414 get_key(i, 3, RK3); \
415 vpslld $7, x3 ## 1, x4 ##
416 vpsrld $(32 - 7), x3 ## 1, x3 ##
417 vpor x4 ## 1, x3 ##
418 vpslld $7, x1 ## 1, x4 ##
419 vpxor x1 ## 1, x0 ##
420 vpxor x3 ## 1, x0 ##
421 vpxor x3 ## 1, x2 ##
422 vpxor x4 ## 1, x2 ##
423 get_key(i, 0, RK0); \
424 vpslld $7, x3 ##
425 vpsrld $(32 - 7), x3 ##
426 vpor x4 ##
427 vpslld $7, x1 ##
428 vpxor x1 ##
429 vpxor x3 ##
430 vpxor x3 ##
431 vpxor x4 ##
432 get_key(i, 2, RK2); \
433 vpxor RK1, x1 ## 1,
434 vpxor RK3, x3 ## 1,
435 vpslld $5, x0 ## 1, x4 ##
436 vpsrld $(32 - 5), x0 ## 1, x0 ##
437 vpor x4 ## 1, x0 ##
438 vpslld $22, x2 ## 1, x4 ##
439 vpsrld $(32 - 22), x2 ## 1, x2 ##
440 vpor x4 ## 1, x2 ##
441 vpxor RK0, x0 ## 1,
442 vpxor RK2, x2 ## 1,
443 vpxor RK1, x
444 vpxor RK3, x
445 vpslld $5, x0 ##
446 vpsrld $(32 - 5), x0 ##
447 vpor x4 ##
448 vpslld $22, x2 ##
449 vpsrld $(32 - 22), x2 ##
450 vpor x4 ##
451 vpxor RK0, x
452 vpxor RK2, x
453
454 #define KL2(x0, x1, x2, x3, x4, i) \
455 vpxor RK0, x0 ## 1,
456 vpxor RK2, x2 ## 1,
457 vpsrld $5, x0 ## 1, x4 ##
458 vpslld $(32 - 5), x0 ## 1, x0 ##
459 vpor x4 ## 1, x0 ##
460 vpxor RK3, x3 ## 1,
461 vpxor RK1, x1 ## 1,
462 vpsrld $22, x2 ## 1, x4 ##
463 vpslld $(32 - 22), x2 ## 1, x2 ##
464 vpor x4 ## 1, x2 ##
465 vpxor x3 ## 1, x2 ##
466 vpxor RK0, x
467 vpxor RK2, x
468 vpsrld $5, x0 ##
469 vpslld $(32 - 5), x0 ##
470 vpor x4 ##
471 vpxor RK3, x
472 vpxor RK1, x
473 vpsrld $22, x2 ##
474 vpslld $(32 - 22), x2 ##
475 vpor x4 ##
476 vpxor x3 ##
477 vpxor x3 ## 1, x0 ##
478 vpslld $7, x1 ## 1, x4 ##
479 vpxor x1 ## 1, x0 ##
480 vpxor x4 ## 1, x2 ##
481 vpsrld $1, x1 ## 1, x4 ##
482 vpslld $(32 - 1), x1 ## 1, x1 ##
483 vpor x4 ## 1, x1 ##
484 vpxor x3 ##
485 vpslld $7, x1 ##
486 vpxor x1 ##
487 vpxor x4 ##
488 vpsrld $1, x1 ##
489 vpslld $(32 - 1), x1 ##
490 vpor x4 ##
491 vpsrld $7, x3 ## 1, x4 ##
492 vpslld $(32 - 7), x3 ## 1, x3 ##
493 vpor x4 ## 1, x3 ##
494 vpxor x0 ## 1, x1 ##
495 vpslld $3, x0 ## 1, x4 ##
496 vpxor x4 ## 1, x3 ##
497 vpsrld $7, x3 ##
498 vpslld $(32 - 7), x3 ##
499 vpor x4 ##
500 vpxor x0 ##
501 vpslld $3, x0 ##
502 vpxor x4 ##
503 vpsrld $13, x0 ## 1, x4 ##
504 vpslld $(32 - 13), x0 ## 1, x0 ##
505 vpor x4 ## 1, x0 ##
506 vpxor x2 ## 1, x1 ##
507 vpxor x2 ## 1, x3 ##
508 vpsrld $3, x2 ## 1, x4 ##
509 vpslld $(32 - 3), x2 ## 1, x2 ##
510 vpor x4 ## 1, x2 ##
511 vpsrld $13, x0 ##
512 vpslld $(32 - 13), x0 ##
513 vpor x4 ##
514 vpxor x2 ##
515 vpxor x2 ##
516 vpsrld $3, x2 ##
517 vpslld $(32 - 3), x2 ##
518 vpor x4 ##
519
520 #define S(SBOX, x0, x1, x2, x3, x4) \
521 SBOX ## _1(x0 ## 1, x1 ## 1, x2 ## 1,
522 SBOX ## _2(x0 ## 1, x1 ## 1, x2 ## 1,
523 SBOX ## _1(x0 ## 2, x1 ## 2, x2 ## 2,
524 SBOX ## _2(x0 ## 2, x1 ## 2, x2 ## 2,
525
526 #define SP(SBOX, x0, x1, x2, x3, x4, i) \
527 get_key(i, 0, RK0); \
528 SBOX ## _1(x0 ## 1, x1 ## 1, x2 ## 1,
529 get_key(i, 2, RK2); \
530 SBOX ## _2(x0 ## 1, x1 ## 1, x2 ## 1,
531 get_key(i, 3, RK3); \
532 SBOX ## _1(x0 ## 2, x1 ## 2, x2 ## 2,
533 get_key(i, 1, RK1); \
534 SBOX ## _2(x0 ## 2, x1 ## 2, x2 ## 2,
535
536 #define transpose_4x4(x0, x1, x2, x3, t0, t1,
537 vpunpckldq x1, x0, t0; \
538 vpunpckhdq x1, x0, t2; \
539 vpunpckldq x3, x2, t1; \
540 vpunpckhdq x3, x2, x3; \
541 \
542 vpunpcklqdq t1, t0, x0; \
543 vpunpckhqdq t1, t0, x1; \
544 vpunpcklqdq x3, t2, x2; \
545 vpunpckhqdq x3, t2, x3;
546
547 #define read_blocks(x0, x1, x2, x3, t0, t1, t2
548 transpose_4x4(x0, x1, x2, x3, t0, t1,
549
550 #define write_blocks(x0, x1, x2, x3, t0, t1, t
551 transpose_4x4(x0, x1, x2, x3, t0, t1,
552
553 SYM_FUNC_START_LOCAL(__serpent_enc_blk16)
554 /* input:
555 * %rdi: ctx, CTX
556 * RA1, RB1, RC1, RD1, RA2, RB2,
557 * output:
558 * RA1, RB1, RC1, RD1, RA2, RB2,
559 */
560
561 vpcmpeqd RNOT, RNOT, RNOT;
562
563 read_blocks(RA1, RB1, RC1, RD1, RK0, R
564 read_blocks(RA2, RB2, RC2, RD2, RK0, R
565
566
567 S(S0, RA, RB, RC, RD, RE);
568 S(S1, RC, RB, RD, RA, RE);
569 S(S2, RE, RD, RA, RC, RB);
570 S(S3, RB, RD, RE, RC, RA);
571 S(S4, RC, RA, RD, RB, RE);
572 S(S5, RA, RD, RB, RE, RC);
573 S(S6, RC, RA, RD, RE, RB);
574 S(S7, RD, RB, RA, RE, RC);
575 S(S0, RC, RA, RE, RD, RB);
576 S(S1, RE, RA, RD, RC, RB);
577 S(S2, RB, RD, RC, RE, RA);
578 S(S3, RA, RD, RB, RE, RC);
579 S(S4, RE, RC, RD, RA, RB);
580 S(S5, RC, RD, RA, RB, RE);
581 S(S6, RE, RC, RD, RB, RA);
582 S(S7, RD, RA, RC, RB, RE);
583 S(S0, RE, RC, RB, RD, RA);
584 S(S1, RB, RC, RD, RE, RA);
585 S(S2, RA, RD, RE, RB, RC);
586 S(S3, RC, RD, RA, RB, RE);
587 S(S4, RB, RE, RD, RC, RA);
588 S(S5, RE, RD, RC, RA, RB);
589 S(S6, RB, RE, RD, RA, RC);
590 S(S7, RD, RC, RE, RA, RB);
591 S(S0, RB, RE, RA, RD, RC);
592 S(S1, RA, RE, RD, RB, RC);
593 S(S2, RC, RD, RB, RA, RE);
594 S(S3, RE, RD, RC, RA, RB);
595 S(S4, RA, RB, RD, RE, RC);
596 S(S5, RB, RD, RE, RC, RA);
597 S(S6, RA, RB, RD, RC, RE);
598 S(S7, RD, RE, RB, RC, RA);
599
600 write_blocks(RA1, RB1, RC1, RD1, RK0,
601 write_blocks(RA2, RB2, RC2, RD2, RK0,
602
603 RET;
604 SYM_FUNC_END(__serpent_enc_blk16)
605
606 SYM_FUNC_START_LOCAL(__serpent_dec_blk16)
607 /* input:
608 * %rdi: ctx, CTX
609 * RA1, RB1, RC1, RD1, RA2, RB2,
610 * output:
611 * RC1, RD1, RB1, RE1, RC2, RD2,
612 */
613
614 vpcmpeqd RNOT, RNOT, RNOT;
615
616 read_blocks(RA1, RB1, RC1, RD1, RK0, R
617 read_blocks(RA2, RB2, RC2, RD2, RK0, R
618
619
620 SP(SI7, RA, RB, RC, RD, RE, 31);
621 SP(SI6, RB, RD, RA, RE, RC, 30);
622 SP(SI5, RA, RC, RE, RB, RD, 29);
623 SP(SI4, RC, RD, RA, RE, RB, 28);
624 SP(SI3, RC, RA, RB, RE, RD, 27);
625 SP(SI2, RB, RC, RD, RE, RA, 26);
626 SP(SI1, RC, RA, RE, RD, RB, 25);
627 SP(SI0, RB, RA, RE, RD, RC, 24);
628 SP(SI7, RE, RC, RA, RB, RD, 23);
629 SP(SI6, RC, RB, RE, RD, RA, 22);
630 SP(SI5, RE, RA, RD, RC, RB, 21);
631 SP(SI4, RA, RB, RE, RD, RC, 20);
632 SP(SI3, RA, RE, RC, RD, RB, 19);
633 SP(SI2, RC, RA, RB, RD, RE, 18);
634 SP(SI1, RA, RE, RD, RB, RC, 17);
635 SP(SI0, RC, RE, RD, RB, RA, 16);
636 SP(SI7, RD, RA, RE, RC, RB, 15);
637 SP(SI6, RA, RC, RD, RB, RE, 14);
638 SP(SI5, RD, RE, RB, RA, RC, 13);
639 SP(SI4, RE, RC, RD, RB, RA, 12);
640 SP(SI3, RE, RD, RA, RB, RC, 11);
641 SP(SI2, RA, RE, RC, RB, RD, 10);
642 SP(SI1, RE, RD, RB, RC, RA, 9);
643 SP(SI0, RA, RD, RB, RC, RE, 8);
644 SP(SI7, RB, RE, RD, RA, RC, 7);
645 SP(SI6, RE, RA, RB, RC, RD, 6);
646 SP(SI5, RB, RD, RC, RE, RA, 5);
647 SP(SI4, RD, RA, RB, RC, RE, 4);
648 SP(SI3, RD, RB, RE, RC, RA, 3);
649 SP(SI2, RE, RD, RA, RC, RB, 2);
650 SP(SI1, RD, RB, RC, RA, RE, 1);
651 S(SI0, RE, RB, RC, RA, RD);
652
653 write_blocks(RC1, RD1, RB1, RE1, RK0,
654 write_blocks(RC2, RD2, RB2, RE2, RK0,
655
656 RET;
657 SYM_FUNC_END(__serpent_dec_blk16)
658
659 SYM_FUNC_START(serpent_ecb_enc_16way)
660 /* input:
661 * %rdi: ctx, CTX
662 * %rsi: dst
663 * %rdx: src
664 */
665 FRAME_BEGIN
666
667 vzeroupper;
668
669 load_16way(%rdx, RA1, RB1, RC1, RD1, R
670
671 call __serpent_enc_blk16;
672
673 store_16way(%rsi, RA1, RB1, RC1, RD1,
674
675 vzeroupper;
676
677 FRAME_END
678 RET;
679 SYM_FUNC_END(serpent_ecb_enc_16way)
680
681 SYM_FUNC_START(serpent_ecb_dec_16way)
682 /* input:
683 * %rdi: ctx, CTX
684 * %rsi: dst
685 * %rdx: src
686 */
687 FRAME_BEGIN
688
689 vzeroupper;
690
691 load_16way(%rdx, RA1, RB1, RC1, RD1, R
692
693 call __serpent_dec_blk16;
694
695 store_16way(%rsi, RC1, RD1, RB1, RE1,
696
697 vzeroupper;
698
699 FRAME_END
700 RET;
701 SYM_FUNC_END(serpent_ecb_dec_16way)
702
703 SYM_FUNC_START(serpent_cbc_dec_16way)
704 /* input:
705 * %rdi: ctx, CTX
706 * %rsi: dst
707 * %rdx: src
708 */
709 FRAME_BEGIN
710
711 vzeroupper;
712
713 load_16way(%rdx, RA1, RB1, RC1, RD1, R
714
715 call __serpent_dec_blk16;
716
717 store_cbc_16way(%rdx, %rsi, RC1, RD1,
718 RK0);
719
720 vzeroupper;
721
722 FRAME_END
723 RET;
724 SYM_FUNC_END(serpent_cbc_dec_16way)
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.