1/* SPDX-License-Identifier: GPL-2.0-or-later */
2/*
3 * Serpent Cipher 8-way parallel algorithm (x86_64/SSE2)
4 *
5 * Copyright (C) 2011 Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
6 *
7 * Based on crypto/serpent.c by
8 *  Copyright (C) 2002 Dag Arne Osvik <osvik@ii.uib.no>
9 *                2003 Herbert Valerio Riedel <hvr@gnu.org>
10 */
11
12#include <linux/linkage.h>
13
14.file "serpent-sse2-x86_64-asm_64.S"
15.text
16
17#define CTX %rdi
18
19/**********************************************************************
20  8-way SSE2 serpent
21 **********************************************************************/
22#define RA1 %xmm0
23#define RB1 %xmm1
24#define RC1 %xmm2
25#define RD1 %xmm3
26#define RE1 %xmm4
27
28#define RA2 %xmm5
29#define RB2 %xmm6
30#define RC2 %xmm7
31#define RD2 %xmm8
32#define RE2 %xmm9
33
34#define RNOT %xmm10
35
36#define RK0 %xmm11
37#define RK1 %xmm12
38#define RK2 %xmm13
39#define RK3 %xmm14
40
41#define S0_1(x0, x1, x2, x3, x4) \
42	movdqa x3,		x4; \
43	por x0,			x3; \
44	pxor x4,		x0; \
45	pxor x2,		x4; \
46	pxor RNOT,		x4; \
47	pxor x1,		x3; \
48	pand x0,		x1; \
49	pxor x4,		x1; \
50	pxor x0,		x2;
51#define S0_2(x0, x1, x2, x3, x4) \
52	pxor x3,		x0; \
53	por x0,			x4; \
54	pxor x2,		x0; \
55	pand x1,		x2; \
56	pxor x2,		x3; \
57	pxor RNOT,		x1; \
58	pxor x4,		x2; \
59	pxor x2,		x1;
60
61#define S1_1(x0, x1, x2, x3, x4) \
62	movdqa x1,		x4; \
63	pxor x0,		x1; \
64	pxor x3,		x0; \
65	pxor RNOT,		x3; \
66	pand x1,		x4; \
67	por x1,			x0; \
68	pxor x2,		x3; \
69	pxor x3,		x0; \
70	pxor x3,		x1;
71#define S1_2(x0, x1, x2, x3, x4) \
72	pxor x4,		x3; \
73	por x4,			x1; \
74	pxor x2,		x4; \
75	pand x0,		x2; \
76	pxor x1,		x2; \
77	por x0,			x1; \
78	pxor RNOT,		x0; \
79	pxor x2,		x0; \
80	pxor x1,		x4;
81
82#define S2_1(x0, x1, x2, x3, x4) \
83	pxor RNOT,		x3; \
84	pxor x0,		x1; \
85	movdqa x0,		x4; \
86	pand x2,		x0; \
87	pxor x3,		x0; \
88	por x4,			x3; \
89	pxor x1,		x2; \
90	pxor x1,		x3; \
91	pand x0,		x1;
92#define S2_2(x0, x1, x2, x3, x4) \
93	pxor x2,		x0; \
94	pand x3,		x2; \
95	por x1,			x3; \
96	pxor RNOT,		x0; \
97	pxor x0,		x3; \
98	pxor x0,		x4; \
99	pxor x2,		x0; \
100	por x2,			x1;
101
102#define S3_1(x0, x1, x2, x3, x4) \
103	movdqa x1,		x4; \
104	pxor x3,		x1; \
105	por x0,			x3; \
106	pand x0,		x4; \
107	pxor x2,		x0; \
108	pxor x1,		x2; \
109	pand x3,		x1; \
110	pxor x3,		x2; \
111	por x4,			x0; \
112	pxor x3,		x4;
113#define S3_2(x0, x1, x2, x3, x4) \
114	pxor x0,		x1; \
115	pand x3,		x0; \
116	pand x4,		x3; \
117	pxor x2,		x3; \
118	por x1,			x4; \
119	pand x1,		x2; \
120	pxor x3,		x4; \
121	pxor x3,		x0; \
122	pxor x2,		x3;
123
124#define S4_1(x0, x1, x2, x3, x4) \
125	movdqa x3,		x4; \
126	pand x0,		x3; \
127	pxor x4,		x0; \
128	pxor x2,		x3; \
129	por x4,			x2; \
130	pxor x1,		x0; \
131	pxor x3,		x4; \
132	por x0,			x2; \
133	pxor x1,		x2;
134#define S4_2(x0, x1, x2, x3, x4) \
135	pand x0,		x1; \
136	pxor x4,		x1; \
137	pand x2,		x4; \
138	pxor x3,		x2; \
139	pxor x0,		x4; \
140	por x1,			x3; \
141	pxor RNOT,		x1; \
142	pxor x0,		x3;
143
144#define S5_1(x0, x1, x2, x3, x4) \
145	movdqa x1,		x4; \
146	por x0,			x1; \
147	pxor x1,		x2; \
148	pxor RNOT,		x3; \
149	pxor x0,		x4; \
150	pxor x2,		x0; \
151	pand x4,		x1; \
152	por x3,			x4; \
153	pxor x0,		x4;
154#define S5_2(x0, x1, x2, x3, x4) \
155	pand x3,		x0; \
156	pxor x3,		x1; \
157	pxor x2,		x3; \
158	pxor x1,		x0; \
159	pand x4,		x2; \
160	pxor x2,		x1; \
161	pand x0,		x2; \
162	pxor x2,		x3;
163
164#define S6_1(x0, x1, x2, x3, x4) \
165	movdqa x1,		x4; \
166	pxor x0,		x3; \
167	pxor x2,		x1; \
168	pxor x0,		x2; \
169	pand x3,		x0; \
170	por x3,			x1; \
171	pxor RNOT,		x4; \
172	pxor x1,		x0; \
173	pxor x2,		x1;
174#define S6_2(x0, x1, x2, x3, x4) \
175	pxor x4,		x3; \
176	pxor x0,		x4; \
177	pand x0,		x2; \
178	pxor x1,		x4; \
179	pxor x3,		x2; \
180	pand x1,		x3; \
181	pxor x0,		x3; \
182	pxor x2,		x1;
183
184#define S7_1(x0, x1, x2, x3, x4) \
185	pxor RNOT,		x1; \
186	movdqa x1,		x4; \
187	pxor RNOT,		x0; \
188	pand x2,		x1; \
189	pxor x3,		x1; \
190	por x4,			x3; \
191	pxor x2,		x4; \
192	pxor x3,		x2; \
193	pxor x0,		x3; \
194	por x1,			x0;
195#define S7_2(x0, x1, x2, x3, x4) \
196	pand x0,		x2; \
197	pxor x4,		x0; \
198	pxor x3,		x4; \
199	pand x0,		x3; \
200	pxor x1,		x4; \
201	pxor x4,		x2; \
202	pxor x1,		x3; \
203	por x0,			x4; \
204	pxor x1,		x4;
205
206#define SI0_1(x0, x1, x2, x3, x4) \
207	movdqa x3,		x4; \
208	pxor x0,		x1; \
209	por x1,			x3; \
210	pxor x1,		x4; \
211	pxor RNOT,		x0; \
212	pxor x3,		x2; \
213	pxor x0,		x3; \
214	pand x1,		x0; \
215	pxor x2,		x0;
216#define SI0_2(x0, x1, x2, x3, x4) \
217	pand x3,		x2; \
218	pxor x4,		x3; \
219	pxor x3,		x2; \
220	pxor x3,		x1; \
221	pand x0,		x3; \
222	pxor x0,		x1; \
223	pxor x2,		x0; \
224	pxor x3,		x4;
225
226#define SI1_1(x0, x1, x2, x3, x4) \
227	pxor x3,		x1; \
228	movdqa x0,		x4; \
229	pxor x2,		x0; \
230	pxor RNOT,		x2; \
231	por x1,			x4; \
232	pxor x3,		x4; \
233	pand x1,		x3; \
234	pxor x2,		x1; \
235	pand x4,		x2;
236#define SI1_2(x0, x1, x2, x3, x4) \
237	pxor x1,		x4; \
238	por x3,			x1; \
239	pxor x0,		x3; \
240	pxor x0,		x2; \
241	por x4,			x0; \
242	pxor x4,		x2; \
243	pxor x0,		x1; \
244	pxor x1,		x4;
245
246#define SI2_1(x0, x1, x2, x3, x4) \
247	pxor x1,		x2; \
248	movdqa x3,		x4; \
249	pxor RNOT,		x3; \
250	por x2,			x3; \
251	pxor x4,		x2; \
252	pxor x0,		x4; \
253	pxor x1,		x3; \
254	por x2,			x1; \
255	pxor x0,		x2;
256#define SI2_2(x0, x1, x2, x3, x4) \
257	pxor x4,		x1; \
258	por x3,			x4; \
259	pxor x3,		x2; \
260	pxor x2,		x4; \
261	pand x1,		x2; \
262	pxor x3,		x2; \
263	pxor x4,		x3; \
264	pxor x0,		x4;
265
266#define SI3_1(x0, x1, x2, x3, x4) \
267	pxor x1,		x2; \
268	movdqa x1,		x4; \
269	pand x2,		x1; \
270	pxor x0,		x1; \
271	por x4,			x0; \
272	pxor x3,		x4; \
273	pxor x3,		x0; \
274	por x1,			x3; \
275	pxor x2,		x1;
276#define SI3_2(x0, x1, x2, x3, x4) \
277	pxor x3,		x1; \
278	pxor x2,		x0; \
279	pxor x3,		x2; \
280	pand x1,		x3; \
281	pxor x0,		x1; \
282	pand x2,		x0; \
283	pxor x3,		x4; \
284	pxor x0,		x3; \
285	pxor x1,		x0;
286
287#define SI4_1(x0, x1, x2, x3, x4) \
288	pxor x3,		x2; \
289	movdqa x0,		x4; \
290	pand x1,		x0; \
291	pxor x2,		x0; \
292	por x3,			x2; \
293	pxor RNOT,		x4; \
294	pxor x0,		x1; \
295	pxor x2,		x0; \
296	pand x4,		x2;
297#define SI4_2(x0, x1, x2, x3, x4) \
298	pxor x0,		x2; \
299	por x4,			x0; \
300	pxor x3,		x0; \
301	pand x2,		x3; \
302	pxor x3,		x4; \
303	pxor x1,		x3; \
304	pand x0,		x1; \
305	pxor x1,		x4; \
306	pxor x3,		x0;
307
308#define SI5_1(x0, x1, x2, x3, x4) \
309	movdqa x1,		x4; \
310	por x2,			x1; \
311	pxor x4,		x2; \
312	pxor x3,		x1; \
313	pand x4,		x3; \
314	pxor x3,		x2; \
315	por x0,			x3; \
316	pxor RNOT,		x0; \
317	pxor x2,		x3; \
318	por x0,			x2;
319#define SI5_2(x0, x1, x2, x3, x4) \
320	pxor x1,		x4; \
321	pxor x4,		x2; \
322	pand x0,		x4; \
323	pxor x1,		x0; \
324	pxor x3,		x1; \
325	pand x2,		x0; \
326	pxor x3,		x2; \
327	pxor x2,		x0; \
328	pxor x4,		x2; \
329	pxor x3,		x4;
330
331#define SI6_1(x0, x1, x2, x3, x4) \
332	pxor x2,		x0; \
333	movdqa x0,		x4; \
334	pand x3,		x0; \
335	pxor x3,		x2; \
336	pxor x2,		x0; \
337	pxor x1,		x3; \
338	por x4,			x2; \
339	pxor x3,		x2; \
340	pand x0,		x3;
341#define SI6_2(x0, x1, x2, x3, x4) \
342	pxor RNOT,		x0; \
343	pxor x1,		x3; \
344	pand x2,		x1; \
345	pxor x0,		x4; \
346	pxor x4,		x3; \
347	pxor x2,		x4; \
348	pxor x1,		x0; \
349	pxor x0,		x2;
350
351#define SI7_1(x0, x1, x2, x3, x4) \
352	movdqa x3,		x4; \
353	pand x0,		x3; \
354	pxor x2,		x0; \
355	por x4,			x2; \
356	pxor x1,		x4; \
357	pxor RNOT,		x0; \
358	por x3,			x1; \
359	pxor x0,		x4; \
360	pand x2,		x0; \
361	pxor x1,		x0;
362#define SI7_2(x0, x1, x2, x3, x4) \
363	pand x2,		x1; \
364	pxor x2,		x3; \
365	pxor x3,		x4; \
366	pand x3,		x2; \
367	por x0,			x3; \
368	pxor x4,		x1; \
369	pxor x4,		x3; \
370	pand x0,		x4; \
371	pxor x2,		x4;
372
373#define get_key(i, j, t) \
374	movd (4*(i)+(j))*4(CTX), t; \
375	pshufd $0, t, t;
376
377#define K2(x0, x1, x2, x3, x4, i) \
378	get_key(i, 0, RK0); \
379	get_key(i, 1, RK1); \
380	get_key(i, 2, RK2); \
381	get_key(i, 3, RK3); \
382	pxor RK0,		x0 ## 1; \
383	pxor RK1,		x1 ## 1; \
384	pxor RK2,		x2 ## 1; \
385	pxor RK3,		x3 ## 1; \
386		pxor RK0,		x0 ## 2; \
387		pxor RK1,		x1 ## 2; \
388		pxor RK2,		x2 ## 2; \
389		pxor RK3,		x3 ## 2;
390
391#define LK2(x0, x1, x2, x3, x4, i) \
392	movdqa x0 ## 1,		x4 ## 1; \
393	pslld $13,		x0 ## 1; \
394	psrld $(32 - 13),	x4 ## 1; \
395	por x4 ## 1,		x0 ## 1; \
396	pxor x0 ## 1,		x1 ## 1; \
397	movdqa x2 ## 1,		x4 ## 1; \
398	pslld $3,		x2 ## 1; \
399	psrld $(32 - 3),	x4 ## 1; \
400	por x4 ## 1,		x2 ## 1; \
401	pxor x2 ## 1,		x1 ## 1; \
402		movdqa x0 ## 2,		x4 ## 2; \
403		pslld $13,		x0 ## 2; \
404		psrld $(32 - 13),	x4 ## 2; \
405		por x4 ## 2,		x0 ## 2; \
406		pxor x0 ## 2,		x1 ## 2; \
407		movdqa x2 ## 2,		x4 ## 2; \
408		pslld $3,		x2 ## 2; \
409		psrld $(32 - 3),	x4 ## 2; \
410		por x4 ## 2,		x2 ## 2; \
411		pxor x2 ## 2,		x1 ## 2; \
412	movdqa x1 ## 1,		x4 ## 1; \
413	pslld $1,		x1 ## 1; \
414	psrld $(32 - 1),	x4 ## 1; \
415	por x4 ## 1,		x1 ## 1; \
416	movdqa x0 ## 1,		x4 ## 1; \
417	pslld $3,		x4 ## 1; \
418	pxor x2 ## 1,		x3 ## 1; \
419	pxor x4 ## 1,		x3 ## 1; \
420	movdqa x3 ## 1,		x4 ## 1; \
421	get_key(i, 1, RK1); \
422		movdqa x1 ## 2,		x4 ## 2; \
423		pslld $1,		x1 ## 2; \
424		psrld $(32 - 1),	x4 ## 2; \
425		por x4 ## 2,		x1 ## 2; \
426		movdqa x0 ## 2,		x4 ## 2; \
427		pslld $3,		x4 ## 2; \
428		pxor x2 ## 2,		x3 ## 2; \
429		pxor x4 ## 2,		x3 ## 2; \
430		movdqa x3 ## 2,		x4 ## 2; \
431		get_key(i, 3, RK3); \
432	pslld $7,		x3 ## 1; \
433	psrld $(32 - 7),	x4 ## 1; \
434	por x4 ## 1,		x3 ## 1; \
435	movdqa x1 ## 1,		x4 ## 1; \
436	pslld $7,		x4 ## 1; \
437	pxor x1 ## 1,		x0 ## 1; \
438	pxor x3 ## 1,		x0 ## 1; \
439	pxor x3 ## 1,		x2 ## 1; \
440	pxor x4 ## 1,		x2 ## 1; \
441	get_key(i, 0, RK0); \
442		pslld $7,		x3 ## 2; \
443		psrld $(32 - 7),	x4 ## 2; \
444		por x4 ## 2,		x3 ## 2; \
445		movdqa x1 ## 2,		x4 ## 2; \
446		pslld $7,		x4 ## 2; \
447		pxor x1 ## 2,		x0 ## 2; \
448		pxor x3 ## 2,		x0 ## 2; \
449		pxor x3 ## 2,		x2 ## 2; \
450		pxor x4 ## 2,		x2 ## 2; \
451		get_key(i, 2, RK2); \
452	pxor RK1,		x1 ## 1; \
453	pxor RK3,		x3 ## 1; \
454	movdqa x0 ## 1,		x4 ## 1; \
455	pslld $5,		x0 ## 1; \
456	psrld $(32 - 5),	x4 ## 1; \
457	por x4 ## 1,		x0 ## 1; \
458	movdqa x2 ## 1,		x4 ## 1; \
459	pslld $22,		x2 ## 1; \
460	psrld $(32 - 22),	x4 ## 1; \
461	por x4 ## 1,		x2 ## 1; \
462	pxor RK0,		x0 ## 1; \
463	pxor RK2,		x2 ## 1; \
464		pxor RK1,		x1 ## 2; \
465		pxor RK3,		x3 ## 2; \
466		movdqa x0 ## 2,		x4 ## 2; \
467		pslld $5,		x0 ## 2; \
468		psrld $(32 - 5),	x4 ## 2; \
469		por x4 ## 2,		x0 ## 2; \
470		movdqa x2 ## 2,		x4 ## 2; \
471		pslld $22,		x2 ## 2; \
472		psrld $(32 - 22),	x4 ## 2; \
473		por x4 ## 2,		x2 ## 2; \
474		pxor RK0,		x0 ## 2; \
475		pxor RK2,		x2 ## 2;
476
477#define KL2(x0, x1, x2, x3, x4, i) \
478	pxor RK0,		x0 ## 1; \
479	pxor RK2,		x2 ## 1; \
480	movdqa x0 ## 1,		x4 ## 1; \
481	psrld $5,		x0 ## 1; \
482	pslld $(32 - 5),	x4 ## 1; \
483	por x4 ## 1,		x0 ## 1; \
484	pxor RK3,		x3 ## 1; \
485	pxor RK1,		x1 ## 1; \
486	movdqa x2 ## 1,		x4 ## 1; \
487	psrld $22,		x2 ## 1; \
488	pslld $(32 - 22),	x4 ## 1; \
489	por x4 ## 1,		x2 ## 1; \
490	pxor x3 ## 1,		x2 ## 1; \
491		pxor RK0,		x0 ## 2; \
492		pxor RK2,		x2 ## 2; \
493		movdqa x0 ## 2,		x4 ## 2; \
494		psrld $5,		x0 ## 2; \
495		pslld $(32 - 5),	x4 ## 2; \
496		por x4 ## 2,		x0 ## 2; \
497		pxor RK3,		x3 ## 2; \
498		pxor RK1,		x1 ## 2; \
499		movdqa x2 ## 2,		x4 ## 2; \
500		psrld $22,		x2 ## 2; \
501		pslld $(32 - 22),	x4 ## 2; \
502		por x4 ## 2,		x2 ## 2; \
503		pxor x3 ## 2,		x2 ## 2; \
504	pxor x3 ## 1,		x0 ## 1; \
505	movdqa x1 ## 1,		x4 ## 1; \
506	pslld $7,		x4 ## 1; \
507	pxor x1 ## 1,		x0 ## 1; \
508	pxor x4 ## 1,		x2 ## 1; \
509	movdqa x1 ## 1,		x4 ## 1; \
510	psrld $1,		x1 ## 1; \
511	pslld $(32 - 1),	x4 ## 1; \
512	por x4 ## 1,		x1 ## 1; \
513		pxor x3 ## 2,		x0 ## 2; \
514		movdqa x1 ## 2,		x4 ## 2; \
515		pslld $7,		x4 ## 2; \
516		pxor x1 ## 2,		x0 ## 2; \
517		pxor x4 ## 2,		x2 ## 2; \
518		movdqa x1 ## 2,		x4 ## 2; \
519		psrld $1,		x1 ## 2; \
520		pslld $(32 - 1),	x4 ## 2; \
521		por x4 ## 2,		x1 ## 2; \
522	movdqa x3 ## 1,		x4 ## 1; \
523	psrld $7,		x3 ## 1; \
524	pslld $(32 - 7),	x4 ## 1; \
525	por x4 ## 1,		x3 ## 1; \
526	pxor x0 ## 1,		x1 ## 1; \
527	movdqa x0 ## 1,		x4 ## 1; \
528	pslld $3,		x4 ## 1; \
529	pxor x4 ## 1,		x3 ## 1; \
530	movdqa x0 ## 1,		x4 ## 1; \
531		movdqa x3 ## 2,		x4 ## 2; \
532		psrld $7,		x3 ## 2; \
533		pslld $(32 - 7),	x4 ## 2; \
534		por x4 ## 2,		x3 ## 2; \
535		pxor x0 ## 2,		x1 ## 2; \
536		movdqa x0 ## 2,		x4 ## 2; \
537		pslld $3,		x4 ## 2; \
538		pxor x4 ## 2,		x3 ## 2; \
539		movdqa x0 ## 2,		x4 ## 2; \
540	psrld $13,		x0 ## 1; \
541	pslld $(32 - 13),	x4 ## 1; \
542	por x4 ## 1,		x0 ## 1; \
543	pxor x2 ## 1,		x1 ## 1; \
544	pxor x2 ## 1,		x3 ## 1; \
545	movdqa x2 ## 1,		x4 ## 1; \
546	psrld $3,		x2 ## 1; \
547	pslld $(32 - 3),	x4 ## 1; \
548	por x4 ## 1,		x2 ## 1; \
549		psrld $13,		x0 ## 2; \
550		pslld $(32 - 13),	x4 ## 2; \
551		por x4 ## 2,		x0 ## 2; \
552		pxor x2 ## 2,		x1 ## 2; \
553		pxor x2 ## 2,		x3 ## 2; \
554		movdqa x2 ## 2,		x4 ## 2; \
555		psrld $3,		x2 ## 2; \
556		pslld $(32 - 3),	x4 ## 2; \
557		por x4 ## 2,		x2 ## 2;
558
559#define S(SBOX, x0, x1, x2, x3, x4) \
560	SBOX ## _1(x0 ## 1, x1 ## 1, x2 ## 1, x3 ## 1, x4 ## 1); \
561	SBOX ## _2(x0 ## 1, x1 ## 1, x2 ## 1, x3 ## 1, x4 ## 1); \
562	SBOX ## _1(x0 ## 2, x1 ## 2, x2 ## 2, x3 ## 2, x4 ## 2); \
563	SBOX ## _2(x0 ## 2, x1 ## 2, x2 ## 2, x3 ## 2, x4 ## 2);
564
565#define SP(SBOX, x0, x1, x2, x3, x4, i) \
566	get_key(i, 0, RK0); \
567	SBOX ## _1(x0 ## 1, x1 ## 1, x2 ## 1, x3 ## 1, x4 ## 1); \
568	get_key(i, 2, RK2); \
569	SBOX ## _1(x0 ## 2, x1 ## 2, x2 ## 2, x3 ## 2, x4 ## 2); \
570	get_key(i, 3, RK3); \
571	SBOX ## _2(x0 ## 1, x1 ## 1, x2 ## 1, x3 ## 1, x4 ## 1); \
572	get_key(i, 1, RK1); \
573	SBOX ## _2(x0 ## 2, x1 ## 2, x2 ## 2, x3 ## 2, x4 ## 2); \
574
575#define transpose_4x4(x0, x1, x2, x3, t0, t1, t2) \
576	movdqa x0,		t2; \
577	punpckldq x1,		x0; \
578	punpckhdq x1,		t2; \
579	movdqa x2,		t1; \
580	punpckhdq x3,		x2; \
581	punpckldq x3,		t1; \
582	movdqa x0,		x1; \
583	punpcklqdq t1,		x0; \
584	punpckhqdq t1,		x1; \
585	movdqa t2,		x3; \
586	punpcklqdq x2,		t2; \
587	punpckhqdq x2,		x3; \
588	movdqa t2,		x2;
589
590#define read_blocks(in, x0, x1, x2, x3, t0, t1, t2) \
591	movdqu (0*4*4)(in),	x0; \
592	movdqu (1*4*4)(in),	x1; \
593	movdqu (2*4*4)(in),	x2; \
594	movdqu (3*4*4)(in),	x3; \
595	\
596	transpose_4x4(x0, x1, x2, x3, t0, t1, t2)
597
598#define write_blocks(out, x0, x1, x2, x3, t0, t1, t2) \
599	transpose_4x4(x0, x1, x2, x3, t0, t1, t2) \
600	\
601	movdqu x0,		(0*4*4)(out); \
602	movdqu x1,		(1*4*4)(out); \
603	movdqu x2,		(2*4*4)(out); \
604	movdqu x3,		(3*4*4)(out);
605
606#define xor_blocks(out, x0, x1, x2, x3, t0, t1, t2) \
607	transpose_4x4(x0, x1, x2, x3, t0, t1, t2) \
608	\
609	movdqu (0*4*4)(out),	t0; \
610	pxor t0,		x0; \
611	movdqu x0,		(0*4*4)(out); \
612	movdqu (1*4*4)(out),	t0; \
613	pxor t0,		x1; \
614	movdqu x1,		(1*4*4)(out); \
615	movdqu (2*4*4)(out),	t0; \
616	pxor t0,		x2; \
617	movdqu x2,		(2*4*4)(out); \
618	movdqu (3*4*4)(out),	t0; \
619	pxor t0,		x3; \
620	movdqu x3,		(3*4*4)(out);
621
622ENTRY(__serpent_enc_blk_8way)
623	/* input:
624	 *	%rdi: ctx, CTX
625	 *	%rsi: dst
626	 *	%rdx: src
627	 *	%rcx: bool, if true: xor output
628	 */
629
630	pcmpeqd RNOT, RNOT;
631
632	leaq (4*4*4)(%rdx), %rax;
633	read_blocks(%rdx, RA1, RB1, RC1, RD1, RK0, RK1, RK2);
634	read_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2);
635
636						 K2(RA, RB, RC, RD, RE, 0);
637	S(S0, RA, RB, RC, RD, RE);		LK2(RC, RB, RD, RA, RE, 1);
638	S(S1, RC, RB, RD, RA, RE);		LK2(RE, RD, RA, RC, RB, 2);
639	S(S2, RE, RD, RA, RC, RB);		LK2(RB, RD, RE, RC, RA, 3);
640	S(S3, RB, RD, RE, RC, RA);		LK2(RC, RA, RD, RB, RE, 4);
641	S(S4, RC, RA, RD, RB, RE);		LK2(RA, RD, RB, RE, RC, 5);
642	S(S5, RA, RD, RB, RE, RC);		LK2(RC, RA, RD, RE, RB, 6);
643	S(S6, RC, RA, RD, RE, RB);		LK2(RD, RB, RA, RE, RC, 7);
644	S(S7, RD, RB, RA, RE, RC);		LK2(RC, RA, RE, RD, RB, 8);
645	S(S0, RC, RA, RE, RD, RB);		LK2(RE, RA, RD, RC, RB, 9);
646	S(S1, RE, RA, RD, RC, RB);		LK2(RB, RD, RC, RE, RA, 10);
647	S(S2, RB, RD, RC, RE, RA);		LK2(RA, RD, RB, RE, RC, 11);
648	S(S3, RA, RD, RB, RE, RC);		LK2(RE, RC, RD, RA, RB, 12);
649	S(S4, RE, RC, RD, RA, RB);		LK2(RC, RD, RA, RB, RE, 13);
650	S(S5, RC, RD, RA, RB, RE);		LK2(RE, RC, RD, RB, RA, 14);
651	S(S6, RE, RC, RD, RB, RA);		LK2(RD, RA, RC, RB, RE, 15);
652	S(S7, RD, RA, RC, RB, RE);		LK2(RE, RC, RB, RD, RA, 16);
653	S(S0, RE, RC, RB, RD, RA);		LK2(RB, RC, RD, RE, RA, 17);
654	S(S1, RB, RC, RD, RE, RA);		LK2(RA, RD, RE, RB, RC, 18);
655	S(S2, RA, RD, RE, RB, RC);		LK2(RC, RD, RA, RB, RE, 19);
656	S(S3, RC, RD, RA, RB, RE);		LK2(RB, RE, RD, RC, RA, 20);
657	S(S4, RB, RE, RD, RC, RA);		LK2(RE, RD, RC, RA, RB, 21);
658	S(S5, RE, RD, RC, RA, RB);		LK2(RB, RE, RD, RA, RC, 22);
659	S(S6, RB, RE, RD, RA, RC);		LK2(RD, RC, RE, RA, RB, 23);
660	S(S7, RD, RC, RE, RA, RB);		LK2(RB, RE, RA, RD, RC, 24);
661	S(S0, RB, RE, RA, RD, RC);		LK2(RA, RE, RD, RB, RC, 25);
662	S(S1, RA, RE, RD, RB, RC);		LK2(RC, RD, RB, RA, RE, 26);
663	S(S2, RC, RD, RB, RA, RE);		LK2(RE, RD, RC, RA, RB, 27);
664	S(S3, RE, RD, RC, RA, RB);		LK2(RA, RB, RD, RE, RC, 28);
665	S(S4, RA, RB, RD, RE, RC);		LK2(RB, RD, RE, RC, RA, 29);
666	S(S5, RB, RD, RE, RC, RA);		LK2(RA, RB, RD, RC, RE, 30);
667	S(S6, RA, RB, RD, RC, RE);		LK2(RD, RE, RB, RC, RA, 31);
668	S(S7, RD, RE, RB, RC, RA);		 K2(RA, RB, RC, RD, RE, 32);
669
670	leaq (4*4*4)(%rsi), %rax;
671
672	testb %cl, %cl;
673	jnz .L__enc_xor8;
674
675	write_blocks(%rsi, RA1, RB1, RC1, RD1, RK0, RK1, RK2);
676	write_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2);
677
678	ret;
679
680.L__enc_xor8:
681	xor_blocks(%rsi, RA1, RB1, RC1, RD1, RK0, RK1, RK2);
682	xor_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2);
683
684	ret;
685ENDPROC(__serpent_enc_blk_8way)
686
687ENTRY(serpent_dec_blk_8way)
688	/* input:
689	 *	%rdi: ctx, CTX
690	 *	%rsi: dst
691	 *	%rdx: src
692	 */
693
694	pcmpeqd RNOT, RNOT;
695
696	leaq (4*4*4)(%rdx), %rax;
697	read_blocks(%rdx, RA1, RB1, RC1, RD1, RK0, RK1, RK2);
698	read_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2);
699
700						 K2(RA, RB, RC, RD, RE, 32);
701	SP(SI7, RA, RB, RC, RD, RE, 31);	KL2(RB, RD, RA, RE, RC, 31);
702	SP(SI6, RB, RD, RA, RE, RC, 30);	KL2(RA, RC, RE, RB, RD, 30);
703	SP(SI5, RA, RC, RE, RB, RD, 29);	KL2(RC, RD, RA, RE, RB, 29);
704	SP(SI4, RC, RD, RA, RE, RB, 28);	KL2(RC, RA, RB, RE, RD, 28);
705	SP(SI3, RC, RA, RB, RE, RD, 27);	KL2(RB, RC, RD, RE, RA, 27);
706	SP(SI2, RB, RC, RD, RE, RA, 26);	KL2(RC, RA, RE, RD, RB, 26);
707	SP(SI1, RC, RA, RE, RD, RB, 25);	KL2(RB, RA, RE, RD, RC, 25);
708	SP(SI0, RB, RA, RE, RD, RC, 24);	KL2(RE, RC, RA, RB, RD, 24);
709	SP(SI7, RE, RC, RA, RB, RD, 23);	KL2(RC, RB, RE, RD, RA, 23);
710	SP(SI6, RC, RB, RE, RD, RA, 22);	KL2(RE, RA, RD, RC, RB, 22);
711	SP(SI5, RE, RA, RD, RC, RB, 21);	KL2(RA, RB, RE, RD, RC, 21);
712	SP(SI4, RA, RB, RE, RD, RC, 20);	KL2(RA, RE, RC, RD, RB, 20);
713	SP(SI3, RA, RE, RC, RD, RB, 19);	KL2(RC, RA, RB, RD, RE, 19);
714	SP(SI2, RC, RA, RB, RD, RE, 18);	KL2(RA, RE, RD, RB, RC, 18);
715	SP(SI1, RA, RE, RD, RB, RC, 17);	KL2(RC, RE, RD, RB, RA, 17);
716	SP(SI0, RC, RE, RD, RB, RA, 16);	KL2(RD, RA, RE, RC, RB, 16);
717	SP(SI7, RD, RA, RE, RC, RB, 15);	KL2(RA, RC, RD, RB, RE, 15);
718	SP(SI6, RA, RC, RD, RB, RE, 14);	KL2(RD, RE, RB, RA, RC, 14);
719	SP(SI5, RD, RE, RB, RA, RC, 13);	KL2(RE, RC, RD, RB, RA, 13);
720	SP(SI4, RE, RC, RD, RB, RA, 12);	KL2(RE, RD, RA, RB, RC, 12);
721	SP(SI3, RE, RD, RA, RB, RC, 11);	KL2(RA, RE, RC, RB, RD, 11);
722	SP(SI2, RA, RE, RC, RB, RD, 10);	KL2(RE, RD, RB, RC, RA, 10);
723	SP(SI1, RE, RD, RB, RC, RA, 9);		KL2(RA, RD, RB, RC, RE, 9);
724	SP(SI0, RA, RD, RB, RC, RE, 8);		KL2(RB, RE, RD, RA, RC, 8);
725	SP(SI7, RB, RE, RD, RA, RC, 7);		KL2(RE, RA, RB, RC, RD, 7);
726	SP(SI6, RE, RA, RB, RC, RD, 6);		KL2(RB, RD, RC, RE, RA, 6);
727	SP(SI5, RB, RD, RC, RE, RA, 5);		KL2(RD, RA, RB, RC, RE, 5);
728	SP(SI4, RD, RA, RB, RC, RE, 4);		KL2(RD, RB, RE, RC, RA, 4);
729	SP(SI3, RD, RB, RE, RC, RA, 3);		KL2(RE, RD, RA, RC, RB, 3);
730	SP(SI2, RE, RD, RA, RC, RB, 2);		KL2(RD, RB, RC, RA, RE, 2);
731	SP(SI1, RD, RB, RC, RA, RE, 1);		KL2(RE, RB, RC, RA, RD, 1);
732	S(SI0, RE, RB, RC, RA, RD);		 K2(RC, RD, RB, RE, RA, 0);
733
734	leaq (4*4*4)(%rsi), %rax;
735	write_blocks(%rsi, RC1, RD1, RB1, RE1, RK0, RK1, RK2);
736	write_blocks(%rax, RC2, RD2, RB2, RE2, RK0, RK1, RK2);
737
738	ret;
739ENDPROC(serpent_dec_blk_8way)
740