MIPS: Switch FPU emulator trap to BREAK instruction.
[linux-2.6] / crypto / camellia.c
1 /*
2  * Copyright (C) 2006
3  * NTT (Nippon Telegraph and Telephone Corporation).
4  *
5  * This program is free software; you can redistribute it and/or
6  * modify it under the terms of the GNU General Public License
7  * as published by the Free Software Foundation; either version 2
8  * of the License, or (at your option) any later version.
9  *
10  * This program is distributed in the hope that it will be useful,
11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
13  * GNU General Public License for more details.
14  *
15  * You should have received a copy of the GNU General Public License
16  * along with this program; if not, write to the Free Software
17  * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
18  */
19
20 /*
21  * Algorithm Specification
22  *  http://info.isl.ntt.co.jp/crypt/eng/camellia/specifications.html
23  */
24
25 /*
26  *
27  * NOTE --- NOTE --- NOTE --- NOTE
28  * This implementation assumes that all memory addresses passed
29  * as parameters are four-byte aligned.
30  *
31  */
32
33 #include <linux/crypto.h>
34 #include <linux/errno.h>
35 #include <linux/init.h>
36 #include <linux/kernel.h>
37 #include <linux/module.h>
38
39 static const u32 camellia_sp1110[256] = {
40         0x70707000,0x82828200,0x2c2c2c00,0xececec00,
41         0xb3b3b300,0x27272700,0xc0c0c000,0xe5e5e500,
42         0xe4e4e400,0x85858500,0x57575700,0x35353500,
43         0xeaeaea00,0x0c0c0c00,0xaeaeae00,0x41414100,
44         0x23232300,0xefefef00,0x6b6b6b00,0x93939300,
45         0x45454500,0x19191900,0xa5a5a500,0x21212100,
46         0xededed00,0x0e0e0e00,0x4f4f4f00,0x4e4e4e00,
47         0x1d1d1d00,0x65656500,0x92929200,0xbdbdbd00,
48         0x86868600,0xb8b8b800,0xafafaf00,0x8f8f8f00,
49         0x7c7c7c00,0xebebeb00,0x1f1f1f00,0xcecece00,
50         0x3e3e3e00,0x30303000,0xdcdcdc00,0x5f5f5f00,
51         0x5e5e5e00,0xc5c5c500,0x0b0b0b00,0x1a1a1a00,
52         0xa6a6a600,0xe1e1e100,0x39393900,0xcacaca00,
53         0xd5d5d500,0x47474700,0x5d5d5d00,0x3d3d3d00,
54         0xd9d9d900,0x01010100,0x5a5a5a00,0xd6d6d600,
55         0x51515100,0x56565600,0x6c6c6c00,0x4d4d4d00,
56         0x8b8b8b00,0x0d0d0d00,0x9a9a9a00,0x66666600,
57         0xfbfbfb00,0xcccccc00,0xb0b0b000,0x2d2d2d00,
58         0x74747400,0x12121200,0x2b2b2b00,0x20202000,
59         0xf0f0f000,0xb1b1b100,0x84848400,0x99999900,
60         0xdfdfdf00,0x4c4c4c00,0xcbcbcb00,0xc2c2c200,
61         0x34343400,0x7e7e7e00,0x76767600,0x05050500,
62         0x6d6d6d00,0xb7b7b700,0xa9a9a900,0x31313100,
63         0xd1d1d100,0x17171700,0x04040400,0xd7d7d700,
64         0x14141400,0x58585800,0x3a3a3a00,0x61616100,
65         0xdedede00,0x1b1b1b00,0x11111100,0x1c1c1c00,
66         0x32323200,0x0f0f0f00,0x9c9c9c00,0x16161600,
67         0x53535300,0x18181800,0xf2f2f200,0x22222200,
68         0xfefefe00,0x44444400,0xcfcfcf00,0xb2b2b200,
69         0xc3c3c300,0xb5b5b500,0x7a7a7a00,0x91919100,
70         0x24242400,0x08080800,0xe8e8e800,0xa8a8a800,
71         0x60606000,0xfcfcfc00,0x69696900,0x50505000,
72         0xaaaaaa00,0xd0d0d000,0xa0a0a000,0x7d7d7d00,
73         0xa1a1a100,0x89898900,0x62626200,0x97979700,
74         0x54545400,0x5b5b5b00,0x1e1e1e00,0x95959500,
75         0xe0e0e000,0xffffff00,0x64646400,0xd2d2d200,
76         0x10101000,0xc4c4c400,0x00000000,0x48484800,
77         0xa3a3a300,0xf7f7f700,0x75757500,0xdbdbdb00,
78         0x8a8a8a00,0x03030300,0xe6e6e600,0xdadada00,
79         0x09090900,0x3f3f3f00,0xdddddd00,0x94949400,
80         0x87878700,0x5c5c5c00,0x83838300,0x02020200,
81         0xcdcdcd00,0x4a4a4a00,0x90909000,0x33333300,
82         0x73737300,0x67676700,0xf6f6f600,0xf3f3f300,
83         0x9d9d9d00,0x7f7f7f00,0xbfbfbf00,0xe2e2e200,
84         0x52525200,0x9b9b9b00,0xd8d8d800,0x26262600,
85         0xc8c8c800,0x37373700,0xc6c6c600,0x3b3b3b00,
86         0x81818100,0x96969600,0x6f6f6f00,0x4b4b4b00,
87         0x13131300,0xbebebe00,0x63636300,0x2e2e2e00,
88         0xe9e9e900,0x79797900,0xa7a7a700,0x8c8c8c00,
89         0x9f9f9f00,0x6e6e6e00,0xbcbcbc00,0x8e8e8e00,
90         0x29292900,0xf5f5f500,0xf9f9f900,0xb6b6b600,
91         0x2f2f2f00,0xfdfdfd00,0xb4b4b400,0x59595900,
92         0x78787800,0x98989800,0x06060600,0x6a6a6a00,
93         0xe7e7e700,0x46464600,0x71717100,0xbababa00,
94         0xd4d4d400,0x25252500,0xababab00,0x42424200,
95         0x88888800,0xa2a2a200,0x8d8d8d00,0xfafafa00,
96         0x72727200,0x07070700,0xb9b9b900,0x55555500,
97         0xf8f8f800,0xeeeeee00,0xacacac00,0x0a0a0a00,
98         0x36363600,0x49494900,0x2a2a2a00,0x68686800,
99         0x3c3c3c00,0x38383800,0xf1f1f100,0xa4a4a400,
100         0x40404000,0x28282800,0xd3d3d300,0x7b7b7b00,
101         0xbbbbbb00,0xc9c9c900,0x43434300,0xc1c1c100,
102         0x15151500,0xe3e3e300,0xadadad00,0xf4f4f400,
103         0x77777700,0xc7c7c700,0x80808000,0x9e9e9e00,
104 };
105
106 static const u32 camellia_sp0222[256] = {
107         0x00e0e0e0,0x00050505,0x00585858,0x00d9d9d9,
108         0x00676767,0x004e4e4e,0x00818181,0x00cbcbcb,
109         0x00c9c9c9,0x000b0b0b,0x00aeaeae,0x006a6a6a,
110         0x00d5d5d5,0x00181818,0x005d5d5d,0x00828282,
111         0x00464646,0x00dfdfdf,0x00d6d6d6,0x00272727,
112         0x008a8a8a,0x00323232,0x004b4b4b,0x00424242,
113         0x00dbdbdb,0x001c1c1c,0x009e9e9e,0x009c9c9c,
114         0x003a3a3a,0x00cacaca,0x00252525,0x007b7b7b,
115         0x000d0d0d,0x00717171,0x005f5f5f,0x001f1f1f,
116         0x00f8f8f8,0x00d7d7d7,0x003e3e3e,0x009d9d9d,
117         0x007c7c7c,0x00606060,0x00b9b9b9,0x00bebebe,
118         0x00bcbcbc,0x008b8b8b,0x00161616,0x00343434,
119         0x004d4d4d,0x00c3c3c3,0x00727272,0x00959595,
120         0x00ababab,0x008e8e8e,0x00bababa,0x007a7a7a,
121         0x00b3b3b3,0x00020202,0x00b4b4b4,0x00adadad,
122         0x00a2a2a2,0x00acacac,0x00d8d8d8,0x009a9a9a,
123         0x00171717,0x001a1a1a,0x00353535,0x00cccccc,
124         0x00f7f7f7,0x00999999,0x00616161,0x005a5a5a,
125         0x00e8e8e8,0x00242424,0x00565656,0x00404040,
126         0x00e1e1e1,0x00636363,0x00090909,0x00333333,
127         0x00bfbfbf,0x00989898,0x00979797,0x00858585,
128         0x00686868,0x00fcfcfc,0x00ececec,0x000a0a0a,
129         0x00dadada,0x006f6f6f,0x00535353,0x00626262,
130         0x00a3a3a3,0x002e2e2e,0x00080808,0x00afafaf,
131         0x00282828,0x00b0b0b0,0x00747474,0x00c2c2c2,
132         0x00bdbdbd,0x00363636,0x00222222,0x00383838,
133         0x00646464,0x001e1e1e,0x00393939,0x002c2c2c,
134         0x00a6a6a6,0x00303030,0x00e5e5e5,0x00444444,
135         0x00fdfdfd,0x00888888,0x009f9f9f,0x00656565,
136         0x00878787,0x006b6b6b,0x00f4f4f4,0x00232323,
137         0x00484848,0x00101010,0x00d1d1d1,0x00515151,
138         0x00c0c0c0,0x00f9f9f9,0x00d2d2d2,0x00a0a0a0,
139         0x00555555,0x00a1a1a1,0x00414141,0x00fafafa,
140         0x00434343,0x00131313,0x00c4c4c4,0x002f2f2f,
141         0x00a8a8a8,0x00b6b6b6,0x003c3c3c,0x002b2b2b,
142         0x00c1c1c1,0x00ffffff,0x00c8c8c8,0x00a5a5a5,
143         0x00202020,0x00898989,0x00000000,0x00909090,
144         0x00474747,0x00efefef,0x00eaeaea,0x00b7b7b7,
145         0x00151515,0x00060606,0x00cdcdcd,0x00b5b5b5,
146         0x00121212,0x007e7e7e,0x00bbbbbb,0x00292929,
147         0x000f0f0f,0x00b8b8b8,0x00070707,0x00040404,
148         0x009b9b9b,0x00949494,0x00212121,0x00666666,
149         0x00e6e6e6,0x00cecece,0x00ededed,0x00e7e7e7,
150         0x003b3b3b,0x00fefefe,0x007f7f7f,0x00c5c5c5,
151         0x00a4a4a4,0x00373737,0x00b1b1b1,0x004c4c4c,
152         0x00919191,0x006e6e6e,0x008d8d8d,0x00767676,
153         0x00030303,0x002d2d2d,0x00dedede,0x00969696,
154         0x00262626,0x007d7d7d,0x00c6c6c6,0x005c5c5c,
155         0x00d3d3d3,0x00f2f2f2,0x004f4f4f,0x00191919,
156         0x003f3f3f,0x00dcdcdc,0x00797979,0x001d1d1d,
157         0x00525252,0x00ebebeb,0x00f3f3f3,0x006d6d6d,
158         0x005e5e5e,0x00fbfbfb,0x00696969,0x00b2b2b2,
159         0x00f0f0f0,0x00313131,0x000c0c0c,0x00d4d4d4,
160         0x00cfcfcf,0x008c8c8c,0x00e2e2e2,0x00757575,
161         0x00a9a9a9,0x004a4a4a,0x00575757,0x00848484,
162         0x00111111,0x00454545,0x001b1b1b,0x00f5f5f5,
163         0x00e4e4e4,0x000e0e0e,0x00737373,0x00aaaaaa,
164         0x00f1f1f1,0x00dddddd,0x00595959,0x00141414,
165         0x006c6c6c,0x00929292,0x00545454,0x00d0d0d0,
166         0x00787878,0x00707070,0x00e3e3e3,0x00494949,
167         0x00808080,0x00505050,0x00a7a7a7,0x00f6f6f6,
168         0x00777777,0x00939393,0x00868686,0x00838383,
169         0x002a2a2a,0x00c7c7c7,0x005b5b5b,0x00e9e9e9,
170         0x00eeeeee,0x008f8f8f,0x00010101,0x003d3d3d,
171 };
172
173 static const u32 camellia_sp3033[256] = {
174         0x38003838,0x41004141,0x16001616,0x76007676,
175         0xd900d9d9,0x93009393,0x60006060,0xf200f2f2,
176         0x72007272,0xc200c2c2,0xab00abab,0x9a009a9a,
177         0x75007575,0x06000606,0x57005757,0xa000a0a0,
178         0x91009191,0xf700f7f7,0xb500b5b5,0xc900c9c9,
179         0xa200a2a2,0x8c008c8c,0xd200d2d2,0x90009090,
180         0xf600f6f6,0x07000707,0xa700a7a7,0x27002727,
181         0x8e008e8e,0xb200b2b2,0x49004949,0xde00dede,
182         0x43004343,0x5c005c5c,0xd700d7d7,0xc700c7c7,
183         0x3e003e3e,0xf500f5f5,0x8f008f8f,0x67006767,
184         0x1f001f1f,0x18001818,0x6e006e6e,0xaf00afaf,
185         0x2f002f2f,0xe200e2e2,0x85008585,0x0d000d0d,
186         0x53005353,0xf000f0f0,0x9c009c9c,0x65006565,
187         0xea00eaea,0xa300a3a3,0xae00aeae,0x9e009e9e,
188         0xec00ecec,0x80008080,0x2d002d2d,0x6b006b6b,
189         0xa800a8a8,0x2b002b2b,0x36003636,0xa600a6a6,
190         0xc500c5c5,0x86008686,0x4d004d4d,0x33003333,
191         0xfd00fdfd,0x66006666,0x58005858,0x96009696,
192         0x3a003a3a,0x09000909,0x95009595,0x10001010,
193         0x78007878,0xd800d8d8,0x42004242,0xcc00cccc,
194         0xef00efef,0x26002626,0xe500e5e5,0x61006161,
195         0x1a001a1a,0x3f003f3f,0x3b003b3b,0x82008282,
196         0xb600b6b6,0xdb00dbdb,0xd400d4d4,0x98009898,
197         0xe800e8e8,0x8b008b8b,0x02000202,0xeb00ebeb,
198         0x0a000a0a,0x2c002c2c,0x1d001d1d,0xb000b0b0,
199         0x6f006f6f,0x8d008d8d,0x88008888,0x0e000e0e,
200         0x19001919,0x87008787,0x4e004e4e,0x0b000b0b,
201         0xa900a9a9,0x0c000c0c,0x79007979,0x11001111,
202         0x7f007f7f,0x22002222,0xe700e7e7,0x59005959,
203         0xe100e1e1,0xda00dada,0x3d003d3d,0xc800c8c8,
204         0x12001212,0x04000404,0x74007474,0x54005454,
205         0x30003030,0x7e007e7e,0xb400b4b4,0x28002828,
206         0x55005555,0x68006868,0x50005050,0xbe00bebe,
207         0xd000d0d0,0xc400c4c4,0x31003131,0xcb00cbcb,
208         0x2a002a2a,0xad00adad,0x0f000f0f,0xca00caca,
209         0x70007070,0xff00ffff,0x32003232,0x69006969,
210         0x08000808,0x62006262,0x00000000,0x24002424,
211         0xd100d1d1,0xfb00fbfb,0xba00baba,0xed00eded,
212         0x45004545,0x81008181,0x73007373,0x6d006d6d,
213         0x84008484,0x9f009f9f,0xee00eeee,0x4a004a4a,
214         0xc300c3c3,0x2e002e2e,0xc100c1c1,0x01000101,
215         0xe600e6e6,0x25002525,0x48004848,0x99009999,
216         0xb900b9b9,0xb300b3b3,0x7b007b7b,0xf900f9f9,
217         0xce00cece,0xbf00bfbf,0xdf00dfdf,0x71007171,
218         0x29002929,0xcd00cdcd,0x6c006c6c,0x13001313,
219         0x64006464,0x9b009b9b,0x63006363,0x9d009d9d,
220         0xc000c0c0,0x4b004b4b,0xb700b7b7,0xa500a5a5,
221         0x89008989,0x5f005f5f,0xb100b1b1,0x17001717,
222         0xf400f4f4,0xbc00bcbc,0xd300d3d3,0x46004646,
223         0xcf00cfcf,0x37003737,0x5e005e5e,0x47004747,
224         0x94009494,0xfa00fafa,0xfc00fcfc,0x5b005b5b,
225         0x97009797,0xfe00fefe,0x5a005a5a,0xac00acac,
226         0x3c003c3c,0x4c004c4c,0x03000303,0x35003535,
227         0xf300f3f3,0x23002323,0xb800b8b8,0x5d005d5d,
228         0x6a006a6a,0x92009292,0xd500d5d5,0x21002121,
229         0x44004444,0x51005151,0xc600c6c6,0x7d007d7d,
230         0x39003939,0x83008383,0xdc00dcdc,0xaa00aaaa,
231         0x7c007c7c,0x77007777,0x56005656,0x05000505,
232         0x1b001b1b,0xa400a4a4,0x15001515,0x34003434,
233         0x1e001e1e,0x1c001c1c,0xf800f8f8,0x52005252,
234         0x20002020,0x14001414,0xe900e9e9,0xbd00bdbd,
235         0xdd00dddd,0xe400e4e4,0xa100a1a1,0xe000e0e0,
236         0x8a008a8a,0xf100f1f1,0xd600d6d6,0x7a007a7a,
237         0xbb00bbbb,0xe300e3e3,0x40004040,0x4f004f4f,
238 };
239
240 static const u32 camellia_sp4404[256] = {
241         0x70700070,0x2c2c002c,0xb3b300b3,0xc0c000c0,
242         0xe4e400e4,0x57570057,0xeaea00ea,0xaeae00ae,
243         0x23230023,0x6b6b006b,0x45450045,0xa5a500a5,
244         0xeded00ed,0x4f4f004f,0x1d1d001d,0x92920092,
245         0x86860086,0xafaf00af,0x7c7c007c,0x1f1f001f,
246         0x3e3e003e,0xdcdc00dc,0x5e5e005e,0x0b0b000b,
247         0xa6a600a6,0x39390039,0xd5d500d5,0x5d5d005d,
248         0xd9d900d9,0x5a5a005a,0x51510051,0x6c6c006c,
249         0x8b8b008b,0x9a9a009a,0xfbfb00fb,0xb0b000b0,
250         0x74740074,0x2b2b002b,0xf0f000f0,0x84840084,
251         0xdfdf00df,0xcbcb00cb,0x34340034,0x76760076,
252         0x6d6d006d,0xa9a900a9,0xd1d100d1,0x04040004,
253         0x14140014,0x3a3a003a,0xdede00de,0x11110011,
254         0x32320032,0x9c9c009c,0x53530053,0xf2f200f2,
255         0xfefe00fe,0xcfcf00cf,0xc3c300c3,0x7a7a007a,
256         0x24240024,0xe8e800e8,0x60600060,0x69690069,
257         0xaaaa00aa,0xa0a000a0,0xa1a100a1,0x62620062,
258         0x54540054,0x1e1e001e,0xe0e000e0,0x64640064,
259         0x10100010,0x00000000,0xa3a300a3,0x75750075,
260         0x8a8a008a,0xe6e600e6,0x09090009,0xdddd00dd,
261         0x87870087,0x83830083,0xcdcd00cd,0x90900090,
262         0x73730073,0xf6f600f6,0x9d9d009d,0xbfbf00bf,
263         0x52520052,0xd8d800d8,0xc8c800c8,0xc6c600c6,
264         0x81810081,0x6f6f006f,0x13130013,0x63630063,
265         0xe9e900e9,0xa7a700a7,0x9f9f009f,0xbcbc00bc,
266         0x29290029,0xf9f900f9,0x2f2f002f,0xb4b400b4,
267         0x78780078,0x06060006,0xe7e700e7,0x71710071,
268         0xd4d400d4,0xabab00ab,0x88880088,0x8d8d008d,
269         0x72720072,0xb9b900b9,0xf8f800f8,0xacac00ac,
270         0x36360036,0x2a2a002a,0x3c3c003c,0xf1f100f1,
271         0x40400040,0xd3d300d3,0xbbbb00bb,0x43430043,
272         0x15150015,0xadad00ad,0x77770077,0x80800080,
273         0x82820082,0xecec00ec,0x27270027,0xe5e500e5,
274         0x85850085,0x35350035,0x0c0c000c,0x41410041,
275         0xefef00ef,0x93930093,0x19190019,0x21210021,
276         0x0e0e000e,0x4e4e004e,0x65650065,0xbdbd00bd,
277         0xb8b800b8,0x8f8f008f,0xebeb00eb,0xcece00ce,
278         0x30300030,0x5f5f005f,0xc5c500c5,0x1a1a001a,
279         0xe1e100e1,0xcaca00ca,0x47470047,0x3d3d003d,
280         0x01010001,0xd6d600d6,0x56560056,0x4d4d004d,
281         0x0d0d000d,0x66660066,0xcccc00cc,0x2d2d002d,
282         0x12120012,0x20200020,0xb1b100b1,0x99990099,
283         0x4c4c004c,0xc2c200c2,0x7e7e007e,0x05050005,
284         0xb7b700b7,0x31310031,0x17170017,0xd7d700d7,
285         0x58580058,0x61610061,0x1b1b001b,0x1c1c001c,
286         0x0f0f000f,0x16160016,0x18180018,0x22220022,
287         0x44440044,0xb2b200b2,0xb5b500b5,0x91910091,
288         0x08080008,0xa8a800a8,0xfcfc00fc,0x50500050,
289         0xd0d000d0,0x7d7d007d,0x89890089,0x97970097,
290         0x5b5b005b,0x95950095,0xffff00ff,0xd2d200d2,
291         0xc4c400c4,0x48480048,0xf7f700f7,0xdbdb00db,
292         0x03030003,0xdada00da,0x3f3f003f,0x94940094,
293         0x5c5c005c,0x02020002,0x4a4a004a,0x33330033,
294         0x67670067,0xf3f300f3,0x7f7f007f,0xe2e200e2,
295         0x9b9b009b,0x26260026,0x37370037,0x3b3b003b,
296         0x96960096,0x4b4b004b,0xbebe00be,0x2e2e002e,
297         0x79790079,0x8c8c008c,0x6e6e006e,0x8e8e008e,
298         0xf5f500f5,0xb6b600b6,0xfdfd00fd,0x59590059,
299         0x98980098,0x6a6a006a,0x46460046,0xbaba00ba,
300         0x25250025,0x42420042,0xa2a200a2,0xfafa00fa,
301         0x07070007,0x55550055,0xeeee00ee,0x0a0a000a,
302         0x49490049,0x68680068,0x38380038,0xa4a400a4,
303         0x28280028,0x7b7b007b,0xc9c900c9,0xc1c100c1,
304         0xe3e300e3,0xf4f400f4,0xc7c700c7,0x9e9e009e,
305 };
306
307
308 #define CAMELLIA_MIN_KEY_SIZE        16
309 #define CAMELLIA_MAX_KEY_SIZE        32
310 #define CAMELLIA_BLOCK_SIZE          16
311 #define CAMELLIA_TABLE_BYTE_LEN     272
312
313 /*
314  * NB: L and R below stand for 'left' and 'right' as in written numbers.
315  * That is, in (xxxL,xxxR) pair xxxL holds most significant digits,
316  * _not_ least significant ones!
317  */
318
319
320 /* key constants */
321
322 #define CAMELLIA_SIGMA1L (0xA09E667FL)
323 #define CAMELLIA_SIGMA1R (0x3BCC908BL)
324 #define CAMELLIA_SIGMA2L (0xB67AE858L)
325 #define CAMELLIA_SIGMA2R (0x4CAA73B2L)
326 #define CAMELLIA_SIGMA3L (0xC6EF372FL)
327 #define CAMELLIA_SIGMA3R (0xE94F82BEL)
328 #define CAMELLIA_SIGMA4L (0x54FF53A5L)
329 #define CAMELLIA_SIGMA4R (0xF1D36F1CL)
330 #define CAMELLIA_SIGMA5L (0x10E527FAL)
331 #define CAMELLIA_SIGMA5R (0xDE682D1DL)
332 #define CAMELLIA_SIGMA6L (0xB05688C2L)
333 #define CAMELLIA_SIGMA6R (0xB3E6C1FDL)
334
335 /*
336  *  macros
337  */
338 #define GETU32(v, pt) \
339     do { \
340         /* latest breed of gcc is clever enough to use move */ \
341         memcpy(&(v), (pt), 4); \
342         (v) = be32_to_cpu(v); \
343     } while(0)
344
345 /* rotation right shift 1byte */
346 #define ROR8(x) (((x) >> 8) + ((x) << 24))
347 /* rotation left shift 1bit */
348 #define ROL1(x) (((x) << 1) + ((x) >> 31))
349 /* rotation left shift 1byte */
350 #define ROL8(x) (((x) << 8) + ((x) >> 24))
351
352 #define ROLDQ(ll, lr, rl, rr, w0, w1, bits)             \
353     do {                                                \
354         w0 = ll;                                        \
355         ll = (ll << bits) + (lr >> (32 - bits));        \
356         lr = (lr << bits) + (rl >> (32 - bits));        \
357         rl = (rl << bits) + (rr >> (32 - bits));        \
358         rr = (rr << bits) + (w0 >> (32 - bits));        \
359     } while(0)
360
361 #define ROLDQo32(ll, lr, rl, rr, w0, w1, bits)          \
362     do {                                                \
363         w0 = ll;                                        \
364         w1 = lr;                                        \
365         ll = (lr << (bits - 32)) + (rl >> (64 - bits)); \
366         lr = (rl << (bits - 32)) + (rr >> (64 - bits)); \
367         rl = (rr << (bits - 32)) + (w0 >> (64 - bits)); \
368         rr = (w0 << (bits - 32)) + (w1 >> (64 - bits)); \
369     } while(0)
370
371 #define CAMELLIA_F(xl, xr, kl, kr, yl, yr, il, ir, t0, t1)      \
372     do {                                                        \
373         il = xl ^ kl;                                           \
374         ir = xr ^ kr;                                           \
375         t0 = il >> 16;                                          \
376         t1 = ir >> 16;                                          \
377         yl = camellia_sp1110[(u8)(ir     )]                     \
378            ^ camellia_sp0222[    (t1 >> 8)]                     \
379            ^ camellia_sp3033[(u8)(t1     )]                     \
380            ^ camellia_sp4404[(u8)(ir >> 8)];                    \
381         yr = camellia_sp1110[    (t0 >> 8)]                     \
382            ^ camellia_sp0222[(u8)(t0     )]                     \
383            ^ camellia_sp3033[(u8)(il >> 8)]                     \
384            ^ camellia_sp4404[(u8)(il     )];                    \
385         yl ^= yr;                                               \
386         yr = ROR8(yr);                                          \
387         yr ^= yl;                                               \
388     } while(0)
389
390 #define SUBKEY_L(INDEX) (subkey[(INDEX)*2])
391 #define SUBKEY_R(INDEX) (subkey[(INDEX)*2 + 1])
392
393 static void camellia_setup_tail(u32 *subkey, u32 *subL, u32 *subR, int max)
394 {
395         u32 dw, tl, tr;
396         u32 kw4l, kw4r;
397         int i;
398
399         /* absorb kw2 to other subkeys */
400         /* round 2 */
401         subL[3] ^= subL[1]; subR[3] ^= subR[1];
402         /* round 4 */
403         subL[5] ^= subL[1]; subR[5] ^= subR[1];
404         /* round 6 */
405         subL[7] ^= subL[1]; subR[7] ^= subR[1];
406         subL[1] ^= subR[1] & ~subR[9];
407         dw = subL[1] & subL[9],
408                 subR[1] ^= ROL1(dw); /* modified for FLinv(kl2) */
409         /* round 8 */
410         subL[11] ^= subL[1]; subR[11] ^= subR[1];
411         /* round 10 */
412         subL[13] ^= subL[1]; subR[13] ^= subR[1];
413         /* round 12 */
414         subL[15] ^= subL[1]; subR[15] ^= subR[1];
415         subL[1] ^= subR[1] & ~subR[17];
416         dw = subL[1] & subL[17],
417                 subR[1] ^= ROL1(dw); /* modified for FLinv(kl4) */
418         /* round 14 */
419         subL[19] ^= subL[1]; subR[19] ^= subR[1];
420         /* round 16 */
421         subL[21] ^= subL[1]; subR[21] ^= subR[1];
422         /* round 18 */
423         subL[23] ^= subL[1]; subR[23] ^= subR[1];
424         if (max == 24) {
425                 /* kw3 */
426                 subL[24] ^= subL[1]; subR[24] ^= subR[1];
427
428         /* absorb kw4 to other subkeys */
429                 kw4l = subL[25]; kw4r = subR[25];
430         } else {
431                 subL[1] ^= subR[1] & ~subR[25];
432                 dw = subL[1] & subL[25],
433                         subR[1] ^= ROL1(dw); /* modified for FLinv(kl6) */
434                 /* round 20 */
435                 subL[27] ^= subL[1]; subR[27] ^= subR[1];
436                 /* round 22 */
437                 subL[29] ^= subL[1]; subR[29] ^= subR[1];
438                 /* round 24 */
439                 subL[31] ^= subL[1]; subR[31] ^= subR[1];
440                 /* kw3 */
441                 subL[32] ^= subL[1]; subR[32] ^= subR[1];
442
443         /* absorb kw4 to other subkeys */
444                 kw4l = subL[33]; kw4r = subR[33];
445                 /* round 23 */
446                 subL[30] ^= kw4l; subR[30] ^= kw4r;
447                 /* round 21 */
448                 subL[28] ^= kw4l; subR[28] ^= kw4r;
449                 /* round 19 */
450                 subL[26] ^= kw4l; subR[26] ^= kw4r;
451                 kw4l ^= kw4r & ~subR[24];
452                 dw = kw4l & subL[24],
453                         kw4r ^= ROL1(dw); /* modified for FL(kl5) */
454         }
455         /* round 17 */
456         subL[22] ^= kw4l; subR[22] ^= kw4r;
457         /* round 15 */
458         subL[20] ^= kw4l; subR[20] ^= kw4r;
459         /* round 13 */
460         subL[18] ^= kw4l; subR[18] ^= kw4r;
461         kw4l ^= kw4r & ~subR[16];
462         dw = kw4l & subL[16],
463                 kw4r ^= ROL1(dw); /* modified for FL(kl3) */
464         /* round 11 */
465         subL[14] ^= kw4l; subR[14] ^= kw4r;
466         /* round 9 */
467         subL[12] ^= kw4l; subR[12] ^= kw4r;
468         /* round 7 */
469         subL[10] ^= kw4l; subR[10] ^= kw4r;
470         kw4l ^= kw4r & ~subR[8];
471         dw = kw4l & subL[8],
472                 kw4r ^= ROL1(dw); /* modified for FL(kl1) */
473         /* round 5 */
474         subL[6] ^= kw4l; subR[6] ^= kw4r;
475         /* round 3 */
476         subL[4] ^= kw4l; subR[4] ^= kw4r;
477         /* round 1 */
478         subL[2] ^= kw4l; subR[2] ^= kw4r;
479         /* kw1 */
480         subL[0] ^= kw4l; subR[0] ^= kw4r;
481
482         /* key XOR is end of F-function */
483         SUBKEY_L(0) = subL[0] ^ subL[2];/* kw1 */
484         SUBKEY_R(0) = subR[0] ^ subR[2];
485         SUBKEY_L(2) = subL[3];       /* round 1 */
486         SUBKEY_R(2) = subR[3];
487         SUBKEY_L(3) = subL[2] ^ subL[4]; /* round 2 */
488         SUBKEY_R(3) = subR[2] ^ subR[4];
489         SUBKEY_L(4) = subL[3] ^ subL[5]; /* round 3 */
490         SUBKEY_R(4) = subR[3] ^ subR[5];
491         SUBKEY_L(5) = subL[4] ^ subL[6]; /* round 4 */
492         SUBKEY_R(5) = subR[4] ^ subR[6];
493         SUBKEY_L(6) = subL[5] ^ subL[7]; /* round 5 */
494         SUBKEY_R(6) = subR[5] ^ subR[7];
495         tl = subL[10] ^ (subR[10] & ~subR[8]);
496         dw = tl & subL[8],  /* FL(kl1) */
497                 tr = subR[10] ^ ROL1(dw);
498         SUBKEY_L(7) = subL[6] ^ tl; /* round 6 */
499         SUBKEY_R(7) = subR[6] ^ tr;
500         SUBKEY_L(8) = subL[8];       /* FL(kl1) */
501         SUBKEY_R(8) = subR[8];
502         SUBKEY_L(9) = subL[9];       /* FLinv(kl2) */
503         SUBKEY_R(9) = subR[9];
504         tl = subL[7] ^ (subR[7] & ~subR[9]);
505         dw = tl & subL[9],  /* FLinv(kl2) */
506                 tr = subR[7] ^ ROL1(dw);
507         SUBKEY_L(10) = tl ^ subL[11]; /* round 7 */
508         SUBKEY_R(10) = tr ^ subR[11];
509         SUBKEY_L(11) = subL[10] ^ subL[12]; /* round 8 */
510         SUBKEY_R(11) = subR[10] ^ subR[12];
511         SUBKEY_L(12) = subL[11] ^ subL[13]; /* round 9 */
512         SUBKEY_R(12) = subR[11] ^ subR[13];
513         SUBKEY_L(13) = subL[12] ^ subL[14]; /* round 10 */
514         SUBKEY_R(13) = subR[12] ^ subR[14];
515         SUBKEY_L(14) = subL[13] ^ subL[15]; /* round 11 */
516         SUBKEY_R(14) = subR[13] ^ subR[15];
517         tl = subL[18] ^ (subR[18] & ~subR[16]);
518         dw = tl & subL[16], /* FL(kl3) */
519                 tr = subR[18] ^ ROL1(dw);
520         SUBKEY_L(15) = subL[14] ^ tl; /* round 12 */
521         SUBKEY_R(15) = subR[14] ^ tr;
522         SUBKEY_L(16) = subL[16];     /* FL(kl3) */
523         SUBKEY_R(16) = subR[16];
524         SUBKEY_L(17) = subL[17];     /* FLinv(kl4) */
525         SUBKEY_R(17) = subR[17];
526         tl = subL[15] ^ (subR[15] & ~subR[17]);
527         dw = tl & subL[17], /* FLinv(kl4) */
528                 tr = subR[15] ^ ROL1(dw);
529         SUBKEY_L(18) = tl ^ subL[19]; /* round 13 */
530         SUBKEY_R(18) = tr ^ subR[19];
531         SUBKEY_L(19) = subL[18] ^ subL[20]; /* round 14 */
532         SUBKEY_R(19) = subR[18] ^ subR[20];
533         SUBKEY_L(20) = subL[19] ^ subL[21]; /* round 15 */
534         SUBKEY_R(20) = subR[19] ^ subR[21];
535         SUBKEY_L(21) = subL[20] ^ subL[22]; /* round 16 */
536         SUBKEY_R(21) = subR[20] ^ subR[22];
537         SUBKEY_L(22) = subL[21] ^ subL[23]; /* round 17 */
538         SUBKEY_R(22) = subR[21] ^ subR[23];
539         if (max == 24) {
540                 SUBKEY_L(23) = subL[22];     /* round 18 */
541                 SUBKEY_R(23) = subR[22];
542                 SUBKEY_L(24) = subL[24] ^ subL[23]; /* kw3 */
543                 SUBKEY_R(24) = subR[24] ^ subR[23];
544         } else {
545                 tl = subL[26] ^ (subR[26] & ~subR[24]);
546                 dw = tl & subL[24], /* FL(kl5) */
547                         tr = subR[26] ^ ROL1(dw);
548                 SUBKEY_L(23) = subL[22] ^ tl; /* round 18 */
549                 SUBKEY_R(23) = subR[22] ^ tr;
550                 SUBKEY_L(24) = subL[24];     /* FL(kl5) */
551                 SUBKEY_R(24) = subR[24];
552                 SUBKEY_L(25) = subL[25];     /* FLinv(kl6) */
553                 SUBKEY_R(25) = subR[25];
554                 tl = subL[23] ^ (subR[23] & ~subR[25]);
555                 dw = tl & subL[25], /* FLinv(kl6) */
556                         tr = subR[23] ^ ROL1(dw);
557                 SUBKEY_L(26) = tl ^ subL[27]; /* round 19 */
558                 SUBKEY_R(26) = tr ^ subR[27];
559                 SUBKEY_L(27) = subL[26] ^ subL[28]; /* round 20 */
560                 SUBKEY_R(27) = subR[26] ^ subR[28];
561                 SUBKEY_L(28) = subL[27] ^ subL[29]; /* round 21 */
562                 SUBKEY_R(28) = subR[27] ^ subR[29];
563                 SUBKEY_L(29) = subL[28] ^ subL[30]; /* round 22 */
564                 SUBKEY_R(29) = subR[28] ^ subR[30];
565                 SUBKEY_L(30) = subL[29] ^ subL[31]; /* round 23 */
566                 SUBKEY_R(30) = subR[29] ^ subR[31];
567                 SUBKEY_L(31) = subL[30];     /* round 24 */
568                 SUBKEY_R(31) = subR[30];
569                 SUBKEY_L(32) = subL[32] ^ subL[31]; /* kw3 */
570                 SUBKEY_R(32) = subR[32] ^ subR[31];
571         }
572
573         /* apply the inverse of the last half of P-function */
574         i = 2;
575         do {
576                 dw = SUBKEY_L(i + 0) ^ SUBKEY_R(i + 0); dw = ROL8(dw);/* round 1 */
577                 SUBKEY_R(i + 0) = SUBKEY_L(i + 0) ^ dw; SUBKEY_L(i + 0) = dw;
578                 dw = SUBKEY_L(i + 1) ^ SUBKEY_R(i + 1); dw = ROL8(dw);/* round 2 */
579                 SUBKEY_R(i + 1) = SUBKEY_L(i + 1) ^ dw; SUBKEY_L(i + 1) = dw;
580                 dw = SUBKEY_L(i + 2) ^ SUBKEY_R(i + 2); dw = ROL8(dw);/* round 3 */
581                 SUBKEY_R(i + 2) = SUBKEY_L(i + 2) ^ dw; SUBKEY_L(i + 2) = dw;
582                 dw = SUBKEY_L(i + 3) ^ SUBKEY_R(i + 3); dw = ROL8(dw);/* round 4 */
583                 SUBKEY_R(i + 3) = SUBKEY_L(i + 3) ^ dw; SUBKEY_L(i + 3) = dw;
584                 dw = SUBKEY_L(i + 4) ^ SUBKEY_R(i + 4); dw = ROL8(dw);/* round 5 */
585                 SUBKEY_R(i + 4) = SUBKEY_L(i + 4) ^ dw; SUBKEY_L(i + 4) = dw;
586                 dw = SUBKEY_L(i + 5) ^ SUBKEY_R(i + 5); dw = ROL8(dw);/* round 6 */
587                 SUBKEY_R(i + 5) = SUBKEY_L(i + 5) ^ dw; SUBKEY_L(i + 5) = dw;
588                 i += 8;
589         } while (i < max);
590 }
591
592 static void camellia_setup128(const unsigned char *key, u32 *subkey)
593 {
594         u32 kll, klr, krl, krr;
595         u32 il, ir, t0, t1, w0, w1;
596         u32 subL[26];
597         u32 subR[26];
598
599         /**
600          *  k == kll || klr || krl || krr (|| is concatenation)
601          */
602         GETU32(kll, key     );
603         GETU32(klr, key +  4);
604         GETU32(krl, key +  8);
605         GETU32(krr, key + 12);
606
607         /* generate KL dependent subkeys */
608         /* kw1 */
609         subL[0] = kll; subR[0] = klr;
610         /* kw2 */
611         subL[1] = krl; subR[1] = krr;
612         /* rotation left shift 15bit */
613         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
614         /* k3 */
615         subL[4] = kll; subR[4] = klr;
616         /* k4 */
617         subL[5] = krl; subR[5] = krr;
618         /* rotation left shift 15+30bit */
619         ROLDQ(kll, klr, krl, krr, w0, w1, 30);
620         /* k7 */
621         subL[10] = kll; subR[10] = klr;
622         /* k8 */
623         subL[11] = krl; subR[11] = krr;
624         /* rotation left shift 15+30+15bit */
625         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
626         /* k10 */
627         subL[13] = krl; subR[13] = krr;
628         /* rotation left shift 15+30+15+17 bit */
629         ROLDQ(kll, klr, krl, krr, w0, w1, 17);
630         /* kl3 */
631         subL[16] = kll; subR[16] = klr;
632         /* kl4 */
633         subL[17] = krl; subR[17] = krr;
634         /* rotation left shift 15+30+15+17+17 bit */
635         ROLDQ(kll, klr, krl, krr, w0, w1, 17);
636         /* k13 */
637         subL[18] = kll; subR[18] = klr;
638         /* k14 */
639         subL[19] = krl; subR[19] = krr;
640         /* rotation left shift 15+30+15+17+17+17 bit */
641         ROLDQ(kll, klr, krl, krr, w0, w1, 17);
642         /* k17 */
643         subL[22] = kll; subR[22] = klr;
644         /* k18 */
645         subL[23] = krl; subR[23] = krr;
646
647         /* generate KA */
648         kll = subL[0]; klr = subR[0];
649         krl = subL[1]; krr = subR[1];
650         CAMELLIA_F(kll, klr,
651                    CAMELLIA_SIGMA1L, CAMELLIA_SIGMA1R,
652                    w0, w1, il, ir, t0, t1);
653         krl ^= w0; krr ^= w1;
654         CAMELLIA_F(krl, krr,
655                    CAMELLIA_SIGMA2L, CAMELLIA_SIGMA2R,
656                    kll, klr, il, ir, t0, t1);
657         /* current status == (kll, klr, w0, w1) */
658         CAMELLIA_F(kll, klr,
659                    CAMELLIA_SIGMA3L, CAMELLIA_SIGMA3R,
660                    krl, krr, il, ir, t0, t1);
661         krl ^= w0; krr ^= w1;
662         CAMELLIA_F(krl, krr,
663                    CAMELLIA_SIGMA4L, CAMELLIA_SIGMA4R,
664                    w0, w1, il, ir, t0, t1);
665         kll ^= w0; klr ^= w1;
666
667         /* generate KA dependent subkeys */
668         /* k1, k2 */
669         subL[2] = kll; subR[2] = klr;
670         subL[3] = krl; subR[3] = krr;
671         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
672         /* k5,k6 */
673         subL[6] = kll; subR[6] = klr;
674         subL[7] = krl; subR[7] = krr;
675         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
676         /* kl1, kl2 */
677         subL[8] = kll; subR[8] = klr;
678         subL[9] = krl; subR[9] = krr;
679         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
680         /* k9 */
681         subL[12] = kll; subR[12] = klr;
682         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
683         /* k11, k12 */
684         subL[14] = kll; subR[14] = klr;
685         subL[15] = krl; subR[15] = krr;
686         ROLDQo32(kll, klr, krl, krr, w0, w1, 34);
687         /* k15, k16 */
688         subL[20] = kll; subR[20] = klr;
689         subL[21] = krl; subR[21] = krr;
690         ROLDQ(kll, klr, krl, krr, w0, w1, 17);
691         /* kw3, kw4 */
692         subL[24] = kll; subR[24] = klr;
693         subL[25] = krl; subR[25] = krr;
694
695         camellia_setup_tail(subkey, subL, subR, 24);
696 }
697
698 static void camellia_setup256(const unsigned char *key, u32 *subkey)
699 {
700         u32 kll, klr, krl, krr;        /* left half of key */
701         u32 krll, krlr, krrl, krrr;    /* right half of key */
702         u32 il, ir, t0, t1, w0, w1;    /* temporary variables */
703         u32 subL[34];
704         u32 subR[34];
705
706         /**
707          *  key = (kll || klr || krl || krr || krll || krlr || krrl || krrr)
708          *  (|| is concatenation)
709          */
710         GETU32(kll,  key     );
711         GETU32(klr,  key +  4);
712         GETU32(krl,  key +  8);
713         GETU32(krr,  key + 12);
714         GETU32(krll, key + 16);
715         GETU32(krlr, key + 20);
716         GETU32(krrl, key + 24);
717         GETU32(krrr, key + 28);
718
719         /* generate KL dependent subkeys */
720         /* kw1 */
721         subL[0] = kll; subR[0] = klr;
722         /* kw2 */
723         subL[1] = krl; subR[1] = krr;
724         ROLDQo32(kll, klr, krl, krr, w0, w1, 45);
725         /* k9 */
726         subL[12] = kll; subR[12] = klr;
727         /* k10 */
728         subL[13] = krl; subR[13] = krr;
729         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
730         /* kl3 */
731         subL[16] = kll; subR[16] = klr;
732         /* kl4 */
733         subL[17] = krl; subR[17] = krr;
734         ROLDQ(kll, klr, krl, krr, w0, w1, 17);
735         /* k17 */
736         subL[22] = kll; subR[22] = klr;
737         /* k18 */
738         subL[23] = krl; subR[23] = krr;
739         ROLDQo32(kll, klr, krl, krr, w0, w1, 34);
740         /* k23 */
741         subL[30] = kll; subR[30] = klr;
742         /* k24 */
743         subL[31] = krl; subR[31] = krr;
744
745         /* generate KR dependent subkeys */
746         ROLDQ(krll, krlr, krrl, krrr, w0, w1, 15);
747         /* k3 */
748         subL[4] = krll; subR[4] = krlr;
749         /* k4 */
750         subL[5] = krrl; subR[5] = krrr;
751         ROLDQ(krll, krlr, krrl, krrr, w0, w1, 15);
752         /* kl1 */
753         subL[8] = krll; subR[8] = krlr;
754         /* kl2 */
755         subL[9] = krrl; subR[9] = krrr;
756         ROLDQ(krll, krlr, krrl, krrr, w0, w1, 30);
757         /* k13 */
758         subL[18] = krll; subR[18] = krlr;
759         /* k14 */
760         subL[19] = krrl; subR[19] = krrr;
761         ROLDQo32(krll, krlr, krrl, krrr, w0, w1, 34);
762         /* k19 */
763         subL[26] = krll; subR[26] = krlr;
764         /* k20 */
765         subL[27] = krrl; subR[27] = krrr;
766         ROLDQo32(krll, krlr, krrl, krrr, w0, w1, 34);
767
768         /* generate KA */
769         kll = subL[0] ^ krll; klr = subR[0] ^ krlr;
770         krl = subL[1] ^ krrl; krr = subR[1] ^ krrr;
771         CAMELLIA_F(kll, klr,
772                    CAMELLIA_SIGMA1L, CAMELLIA_SIGMA1R,
773                    w0, w1, il, ir, t0, t1);
774         krl ^= w0; krr ^= w1;
775         CAMELLIA_F(krl, krr,
776                    CAMELLIA_SIGMA2L, CAMELLIA_SIGMA2R,
777                    kll, klr, il, ir, t0, t1);
778         kll ^= krll; klr ^= krlr;
779         CAMELLIA_F(kll, klr,
780                    CAMELLIA_SIGMA3L, CAMELLIA_SIGMA3R,
781                    krl, krr, il, ir, t0, t1);
782         krl ^= w0 ^ krrl; krr ^= w1 ^ krrr;
783         CAMELLIA_F(krl, krr,
784                    CAMELLIA_SIGMA4L, CAMELLIA_SIGMA4R,
785                    w0, w1, il, ir, t0, t1);
786         kll ^= w0; klr ^= w1;
787
788         /* generate KB */
789         krll ^= kll; krlr ^= klr;
790         krrl ^= krl; krrr ^= krr;
791         CAMELLIA_F(krll, krlr,
792                    CAMELLIA_SIGMA5L, CAMELLIA_SIGMA5R,
793                    w0, w1, il, ir, t0, t1);
794         krrl ^= w0; krrr ^= w1;
795         CAMELLIA_F(krrl, krrr,
796                    CAMELLIA_SIGMA6L, CAMELLIA_SIGMA6R,
797                    w0, w1, il, ir, t0, t1);
798         krll ^= w0; krlr ^= w1;
799
800         /* generate KA dependent subkeys */
801         ROLDQ(kll, klr, krl, krr, w0, w1, 15);
802         /* k5 */
803         subL[6] = kll; subR[6] = klr;
804         /* k6 */
805         subL[7] = krl; subR[7] = krr;
806         ROLDQ(kll, klr, krl, krr, w0, w1, 30);
807         /* k11 */
808         subL[14] = kll; subR[14] = klr;
809         /* k12 */
810         subL[15] = krl; subR[15] = krr;
811         /* rotation left shift 32bit */
812         /* kl5 */
813         subL[24] = klr; subR[24] = krl;
814         /* kl6 */
815         subL[25] = krr; subR[25] = kll;
816         /* rotation left shift 49 from k11,k12 -> k21,k22 */
817         ROLDQo32(kll, klr, krl, krr, w0, w1, 49);
818         /* k21 */
819         subL[28] = kll; subR[28] = klr;
820         /* k22 */
821         subL[29] = krl; subR[29] = krr;
822
823         /* generate KB dependent subkeys */
824         /* k1 */
825         subL[2] = krll; subR[2] = krlr;
826         /* k2 */
827         subL[3] = krrl; subR[3] = krrr;
828         ROLDQ(krll, krlr, krrl, krrr, w0, w1, 30);
829         /* k7 */
830         subL[10] = krll; subR[10] = krlr;
831         /* k8 */
832         subL[11] = krrl; subR[11] = krrr;
833         ROLDQ(krll, krlr, krrl, krrr, w0, w1, 30);
834         /* k15 */
835         subL[20] = krll; subR[20] = krlr;
836         /* k16 */
837         subL[21] = krrl; subR[21] = krrr;
838         ROLDQo32(krll, krlr, krrl, krrr, w0, w1, 51);
839         /* kw3 */
840         subL[32] = krll; subR[32] = krlr;
841         /* kw4 */
842         subL[33] = krrl; subR[33] = krrr;
843
844         camellia_setup_tail(subkey, subL, subR, 32);
845 }
846
847 static void camellia_setup192(const unsigned char *key, u32 *subkey)
848 {
849         unsigned char kk[32];
850         u32 krll, krlr, krrl,krrr;
851
852         memcpy(kk, key, 24);
853         memcpy((unsigned char *)&krll, key+16, 4);
854         memcpy((unsigned char *)&krlr, key+20, 4);
855         krrl = ~krll;
856         krrr = ~krlr;
857         memcpy(kk+24, (unsigned char *)&krrl, 4);
858         memcpy(kk+28, (unsigned char *)&krrr, 4);
859         camellia_setup256(kk, subkey);
860 }
861
862
863 /*
864  * Encrypt/decrypt
865  */
866 #define CAMELLIA_FLS(ll, lr, rl, rr, kll, klr, krl, krr, t0, t1, t2, t3) \
867     do {                                                                \
868         t0 = kll;                                                       \
869         t2 = krr;                                                       \
870         t0 &= ll;                                                       \
871         t2 |= rr;                                                       \
872         rl ^= t2;                                                       \
873         lr ^= ROL1(t0);                                                 \
874         t3 = krl;                                                       \
875         t1 = klr;                                                       \
876         t3 &= rl;                                                       \
877         t1 |= lr;                                                       \
878         ll ^= t1;                                                       \
879         rr ^= ROL1(t3);                                                 \
880     } while(0)
881
882 #define CAMELLIA_ROUNDSM(xl, xr, kl, kr, yl, yr, il, ir)                \
883     do {                                                                \
884         ir =  camellia_sp1110[(u8)xr];                                  \
885         il =  camellia_sp1110[    (xl >> 24)];                          \
886         ir ^= camellia_sp0222[    (xr >> 24)];                          \
887         il ^= camellia_sp0222[(u8)(xl >> 16)];                          \
888         ir ^= camellia_sp3033[(u8)(xr >> 16)];                          \
889         il ^= camellia_sp3033[(u8)(xl >> 8)];                           \
890         ir ^= camellia_sp4404[(u8)(xr >> 8)];                           \
891         il ^= camellia_sp4404[(u8)xl];                                  \
892         il ^= kl;                                                       \
893         ir ^= il ^ kr;                                                  \
894         yl ^= ir;                                                       \
895         yr ^= ROR8(il) ^ ir;                                            \
896     } while(0)
897
898 /* max = 24: 128bit encrypt, max = 32: 256bit encrypt */
899 static void camellia_do_encrypt(const u32 *subkey, u32 *io, unsigned max)
900 {
901         u32 il,ir,t0,t1;               /* temporary variables */
902
903         /* pre whitening but absorb kw2 */
904         io[0] ^= SUBKEY_L(0);
905         io[1] ^= SUBKEY_R(0);
906
907         /* main iteration */
908 #define ROUNDS(i) do { \
909         CAMELLIA_ROUNDSM(io[0],io[1], \
910                          SUBKEY_L(i + 2),SUBKEY_R(i + 2), \
911                          io[2],io[3],il,ir); \
912         CAMELLIA_ROUNDSM(io[2],io[3], \
913                          SUBKEY_L(i + 3),SUBKEY_R(i + 3), \
914                          io[0],io[1],il,ir); \
915         CAMELLIA_ROUNDSM(io[0],io[1], \
916                          SUBKEY_L(i + 4),SUBKEY_R(i + 4), \
917                          io[2],io[3],il,ir); \
918         CAMELLIA_ROUNDSM(io[2],io[3], \
919                          SUBKEY_L(i + 5),SUBKEY_R(i + 5), \
920                          io[0],io[1],il,ir); \
921         CAMELLIA_ROUNDSM(io[0],io[1], \
922                          SUBKEY_L(i + 6),SUBKEY_R(i + 6), \
923                          io[2],io[3],il,ir); \
924         CAMELLIA_ROUNDSM(io[2],io[3], \
925                          SUBKEY_L(i + 7),SUBKEY_R(i + 7), \
926                          io[0],io[1],il,ir); \
927 } while (0)
928 #define FLS(i) do { \
929         CAMELLIA_FLS(io[0],io[1],io[2],io[3], \
930                      SUBKEY_L(i + 0),SUBKEY_R(i + 0), \
931                      SUBKEY_L(i + 1),SUBKEY_R(i + 1), \
932                      t0,t1,il,ir); \
933 } while (0)
934
935         ROUNDS(0);
936         FLS(8);
937         ROUNDS(8);
938         FLS(16);
939         ROUNDS(16);
940         if (max == 32) {
941                 FLS(24);
942                 ROUNDS(24);
943         }
944
945 #undef ROUNDS
946 #undef FLS
947
948         /* post whitening but kw4 */
949         io[2] ^= SUBKEY_L(max);
950         io[3] ^= SUBKEY_R(max);
951         /* NB: io[0],[1] should be swapped with [2],[3] by caller! */
952 }
953
954 static void camellia_do_decrypt(const u32 *subkey, u32 *io, unsigned i)
955 {
956         u32 il,ir,t0,t1;               /* temporary variables */
957
958         /* pre whitening but absorb kw2 */
959         io[0] ^= SUBKEY_L(i);
960         io[1] ^= SUBKEY_R(i);
961
962         /* main iteration */
963 #define ROUNDS(i) do { \
964         CAMELLIA_ROUNDSM(io[0],io[1], \
965                          SUBKEY_L(i + 7),SUBKEY_R(i + 7), \
966                          io[2],io[3],il,ir); \
967         CAMELLIA_ROUNDSM(io[2],io[3], \
968                          SUBKEY_L(i + 6),SUBKEY_R(i + 6), \
969                          io[0],io[1],il,ir); \
970         CAMELLIA_ROUNDSM(io[0],io[1], \
971                          SUBKEY_L(i + 5),SUBKEY_R(i + 5), \
972                          io[2],io[3],il,ir); \
973         CAMELLIA_ROUNDSM(io[2],io[3], \
974                          SUBKEY_L(i + 4),SUBKEY_R(i + 4), \
975                          io[0],io[1],il,ir); \
976         CAMELLIA_ROUNDSM(io[0],io[1], \
977                          SUBKEY_L(i + 3),SUBKEY_R(i + 3), \
978                          io[2],io[3],il,ir); \
979         CAMELLIA_ROUNDSM(io[2],io[3], \
980                          SUBKEY_L(i + 2),SUBKEY_R(i + 2), \
981                          io[0],io[1],il,ir); \
982 } while (0)
983 #define FLS(i) do { \
984         CAMELLIA_FLS(io[0],io[1],io[2],io[3], \
985                      SUBKEY_L(i + 1),SUBKEY_R(i + 1), \
986                      SUBKEY_L(i + 0),SUBKEY_R(i + 0), \
987                      t0,t1,il,ir); \
988 } while (0)
989
990         if (i == 32) {
991                 ROUNDS(24);
992                 FLS(24);
993         }
994         ROUNDS(16);
995         FLS(16);
996         ROUNDS(8);
997         FLS(8);
998         ROUNDS(0);
999
1000 #undef ROUNDS
1001 #undef FLS
1002
1003         /* post whitening but kw4 */
1004         io[2] ^= SUBKEY_L(0);
1005         io[3] ^= SUBKEY_R(0);
1006         /* NB: 0,1 should be swapped with 2,3 by caller! */
1007 }
1008
1009
1010 struct camellia_ctx {
1011         int key_length;
1012         u32 key_table[CAMELLIA_TABLE_BYTE_LEN / sizeof(u32)];
1013 };
1014
1015 static int
1016 camellia_set_key(struct crypto_tfm *tfm, const u8 *in_key,
1017                  unsigned int key_len)
1018 {
1019         struct camellia_ctx *cctx = crypto_tfm_ctx(tfm);
1020         const unsigned char *key = (const unsigned char *)in_key;
1021         u32 *flags = &tfm->crt_flags;
1022
1023         if (key_len != 16 && key_len != 24 && key_len != 32) {
1024                 *flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
1025                 return -EINVAL;
1026         }
1027
1028         cctx->key_length = key_len;
1029
1030         switch (key_len) {
1031         case 16:
1032                 camellia_setup128(key, cctx->key_table);
1033                 break;
1034         case 24:
1035                 camellia_setup192(key, cctx->key_table);
1036                 break;
1037         case 32:
1038                 camellia_setup256(key, cctx->key_table);
1039                 break;
1040         }
1041
1042         return 0;
1043 }
1044
1045 static void camellia_encrypt(struct crypto_tfm *tfm, u8 *out, const u8 *in)
1046 {
1047         const struct camellia_ctx *cctx = crypto_tfm_ctx(tfm);
1048         const __be32 *src = (const __be32 *)in;
1049         __be32 *dst = (__be32 *)out;
1050
1051         u32 tmp[4];
1052
1053         tmp[0] = be32_to_cpu(src[0]);
1054         tmp[1] = be32_to_cpu(src[1]);
1055         tmp[2] = be32_to_cpu(src[2]);
1056         tmp[3] = be32_to_cpu(src[3]);
1057
1058         camellia_do_encrypt(cctx->key_table, tmp,
1059                 cctx->key_length == 16 ? 24 : 32 /* for key lengths of 24 and 32 */
1060         );
1061
1062         /* do_encrypt returns 0,1 swapped with 2,3 */
1063         dst[0] = cpu_to_be32(tmp[2]);
1064         dst[1] = cpu_to_be32(tmp[3]);
1065         dst[2] = cpu_to_be32(tmp[0]);
1066         dst[3] = cpu_to_be32(tmp[1]);
1067 }
1068
1069 static void camellia_decrypt(struct crypto_tfm *tfm, u8 *out, const u8 *in)
1070 {
1071         const struct camellia_ctx *cctx = crypto_tfm_ctx(tfm);
1072         const __be32 *src = (const __be32 *)in;
1073         __be32 *dst = (__be32 *)out;
1074
1075         u32 tmp[4];
1076
1077         tmp[0] = be32_to_cpu(src[0]);
1078         tmp[1] = be32_to_cpu(src[1]);
1079         tmp[2] = be32_to_cpu(src[2]);
1080         tmp[3] = be32_to_cpu(src[3]);
1081
1082         camellia_do_decrypt(cctx->key_table, tmp,
1083                 cctx->key_length == 16 ? 24 : 32 /* for key lengths of 24 and 32 */
1084         );
1085
1086         /* do_decrypt returns 0,1 swapped with 2,3 */
1087         dst[0] = cpu_to_be32(tmp[2]);
1088         dst[1] = cpu_to_be32(tmp[3]);
1089         dst[2] = cpu_to_be32(tmp[0]);
1090         dst[3] = cpu_to_be32(tmp[1]);
1091 }
1092
1093 static struct crypto_alg camellia_alg = {
1094         .cra_name               =       "camellia",
1095         .cra_driver_name        =       "camellia-generic",
1096         .cra_priority           =       100,
1097         .cra_flags              =       CRYPTO_ALG_TYPE_CIPHER,
1098         .cra_blocksize          =       CAMELLIA_BLOCK_SIZE,
1099         .cra_ctxsize            =       sizeof(struct camellia_ctx),
1100         .cra_alignmask          =       3,
1101         .cra_module             =       THIS_MODULE,
1102         .cra_list               =       LIST_HEAD_INIT(camellia_alg.cra_list),
1103         .cra_u                  =       {
1104                 .cipher = {
1105                         .cia_min_keysize        =       CAMELLIA_MIN_KEY_SIZE,
1106                         .cia_max_keysize        =       CAMELLIA_MAX_KEY_SIZE,
1107                         .cia_setkey             =       camellia_set_key,
1108                         .cia_encrypt            =       camellia_encrypt,
1109                         .cia_decrypt            =       camellia_decrypt
1110                 }
1111         }
1112 };
1113
1114 static int __init camellia_init(void)
1115 {
1116         return crypto_register_alg(&camellia_alg);
1117 }
1118
1119 static void __exit camellia_fini(void)
1120 {
1121         crypto_unregister_alg(&camellia_alg);
1122 }
1123
1124 module_init(camellia_init);
1125 module_exit(camellia_fini);
1126
1127 MODULE_DESCRIPTION("Camellia Cipher Algorithm");
1128 MODULE_LICENSE("GPL");