1/*-
2 * Copyright 2013 John-Mark Gurney <jmg@FreeBSD.org>
3 * Copyright 2015 Netflix, Inc.
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 * 1. Redistributions of source code must retain the above copyright
10 *    notice, this list of conditions and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 *    notice, this list of conditions and the following disclaimer in the
13 *    documentation and/or other materials provided with the distribution.
14 *
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
19 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25 * SUCH DAMAGE.
26 *
27 * $FreeBSD$
28 *
29 */
30
31#ifndef _AESENCDEC_H_
32#define _AESENCDEC_H_
33
34#include <crypto/aesni/aesni_os.h>
35
36#include <wmmintrin.h>
37
38static inline void
39aesni_enc8(int rounds, const __m128i *keysched, __m128i a,
40    __m128i b, __m128i c, __m128i d, __m128i e, __m128i f, __m128i g,
41    __m128i h, __m128i out[8])
42{
43	int i;
44
45	a ^= keysched[0];
46	b ^= keysched[0];
47	c ^= keysched[0];
48	d ^= keysched[0];
49	e ^= keysched[0];
50	f ^= keysched[0];
51	g ^= keysched[0];
52	h ^= keysched[0];
53
54	for (i = 0; i < rounds; i++) {
55		a = _mm_aesenc_si128(a, keysched[i + 1]);
56		b = _mm_aesenc_si128(b, keysched[i + 1]);
57		c = _mm_aesenc_si128(c, keysched[i + 1]);
58		d = _mm_aesenc_si128(d, keysched[i + 1]);
59		e = _mm_aesenc_si128(e, keysched[i + 1]);
60		f = _mm_aesenc_si128(f, keysched[i + 1]);
61		g = _mm_aesenc_si128(g, keysched[i + 1]);
62		h = _mm_aesenc_si128(h, keysched[i + 1]);
63	}
64
65	out[0] = _mm_aesenclast_si128(a, keysched[i + 1]);
66	out[1] = _mm_aesenclast_si128(b, keysched[i + 1]);
67	out[2] = _mm_aesenclast_si128(c, keysched[i + 1]);
68	out[3] = _mm_aesenclast_si128(d, keysched[i + 1]);
69	out[4] = _mm_aesenclast_si128(e, keysched[i + 1]);
70	out[5] = _mm_aesenclast_si128(f, keysched[i + 1]);
71	out[6] = _mm_aesenclast_si128(g, keysched[i + 1]);
72	out[7] = _mm_aesenclast_si128(h, keysched[i + 1]);
73}
74
75static inline void
76aesni_dec8(int rounds, const __m128i *keysched, __m128i a,
77    __m128i b, __m128i c, __m128i d, __m128i e, __m128i f, __m128i g,
78    __m128i h, __m128i out[8])
79{
80	int i;
81
82	a ^= keysched[0];
83	b ^= keysched[0];
84	c ^= keysched[0];
85	d ^= keysched[0];
86	e ^= keysched[0];
87	f ^= keysched[0];
88	g ^= keysched[0];
89	h ^= keysched[0];
90
91	for (i = 0; i < rounds; i++) {
92		a = _mm_aesdec_si128(a, keysched[i + 1]);
93		b = _mm_aesdec_si128(b, keysched[i + 1]);
94		c = _mm_aesdec_si128(c, keysched[i + 1]);
95		d = _mm_aesdec_si128(d, keysched[i + 1]);
96		e = _mm_aesdec_si128(e, keysched[i + 1]);
97		f = _mm_aesdec_si128(f, keysched[i + 1]);
98		g = _mm_aesdec_si128(g, keysched[i + 1]);
99		h = _mm_aesdec_si128(h, keysched[i + 1]);
100	}
101
102	out[0] = _mm_aesdeclast_si128(a, keysched[i + 1]);
103	out[1] = _mm_aesdeclast_si128(b, keysched[i + 1]);
104	out[2] = _mm_aesdeclast_si128(c, keysched[i + 1]);
105	out[3] = _mm_aesdeclast_si128(d, keysched[i + 1]);
106	out[4] = _mm_aesdeclast_si128(e, keysched[i + 1]);
107	out[5] = _mm_aesdeclast_si128(f, keysched[i + 1]);
108	out[6] = _mm_aesdeclast_si128(g, keysched[i + 1]);
109	out[7] = _mm_aesdeclast_si128(h, keysched[i + 1]);
110}
111
112/* rounds is passed in as rounds - 1 */
113static inline __m128i
114aesni_enc(int rounds, const __m128i *keysched, const __m128i from)
115{
116	__m128i tmp;
117	int i;
118
119	tmp = from ^ keysched[0];
120	for (i = 1; i < rounds; i += 2) {
121		tmp = _mm_aesenc_si128(tmp, keysched[i]);
122		tmp = _mm_aesenc_si128(tmp, keysched[i + 1]);
123	}
124
125	tmp = _mm_aesenc_si128(tmp, keysched[rounds]);
126	return _mm_aesenclast_si128(tmp, keysched[rounds + 1]);
127}
128
129static inline __m128i
130aesni_dec(int rounds, const __m128i *keysched, const __m128i from)
131{
132	__m128i tmp;
133	int i;
134
135	tmp = from ^ keysched[0];
136
137	for (i = 1; i < rounds; i += 2) {
138		tmp = _mm_aesdec_si128(tmp, keysched[i]);
139		tmp = _mm_aesdec_si128(tmp, keysched[i + 1]);
140	}
141
142	tmp = _mm_aesdec_si128(tmp, keysched[rounds]);
143	return _mm_aesdeclast_si128(tmp, keysched[rounds + 1]);
144}
145
146#endif /* _AESENCDEC_H_ */
147