Fix undefined behaviour in shifts.
[openssl.git] / crypto / aes / aes_core.c
1 /* crypto/aes/aes_core.c -*- mode:C; c-file-style: "eay" -*- */
2 /**
3  * rijndael-alg-fst.c
4  *
5  * @version 3.0 (December 2000)
6  *
7  * Optimised ANSI C code for the Rijndael cipher (now AES)
8  *
9  * @author Vincent Rijmen <vincent.rijmen@esat.kuleuven.ac.be>
10  * @author Antoon Bosselaers <antoon.bosselaers@esat.kuleuven.ac.be>
11  * @author Paulo Barreto <paulo.barreto@terra.com.br>
12  *
13  * This code is hereby placed in the public domain.
14  *
15  * THIS SOFTWARE IS PROVIDED BY THE AUTHORS ''AS IS'' AND ANY EXPRESS
16  * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
17  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE
19  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
20  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
21  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
22  * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
23  * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE
24  * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
25  * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26  */
27
28 /* Note: rewritten a little bit to provide error control and an OpenSSL-
29    compatible API */
30
31 #ifndef AES_DEBUG
32 # ifndef NDEBUG
33 #  define NDEBUG
34 # endif
35 #endif
36 #include <assert.h>
37
38 #include <stdlib.h>
39 #include <openssl/crypto.h>
40 #include <openssl/aes.h>
41 #include "aes_locl.h"
42
43 #ifndef AES_ASM
44 /*-
45 Te0[x] = S [x].[02, 01, 01, 03];
46 Te1[x] = S [x].[03, 02, 01, 01];
47 Te2[x] = S [x].[01, 03, 02, 01];
48 Te3[x] = S [x].[01, 01, 03, 02];
49
50 Td0[x] = Si[x].[0e, 09, 0d, 0b];
51 Td1[x] = Si[x].[0b, 0e, 09, 0d];
52 Td2[x] = Si[x].[0d, 0b, 0e, 09];
53 Td3[x] = Si[x].[09, 0d, 0b, 0e];
54 Td4[x] = Si[x].[01];
55 */
56
57 static const u32 Te0[256] = {
58     0xc66363a5U, 0xf87c7c84U, 0xee777799U, 0xf67b7b8dU,
59     0xfff2f20dU, 0xd66b6bbdU, 0xde6f6fb1U, 0x91c5c554U,
60     0x60303050U, 0x02010103U, 0xce6767a9U, 0x562b2b7dU,
61     0xe7fefe19U, 0xb5d7d762U, 0x4dababe6U, 0xec76769aU,
62     0x8fcaca45U, 0x1f82829dU, 0x89c9c940U, 0xfa7d7d87U,
63     0xeffafa15U, 0xb25959ebU, 0x8e4747c9U, 0xfbf0f00bU,
64     0x41adadecU, 0xb3d4d467U, 0x5fa2a2fdU, 0x45afafeaU,
65     0x239c9cbfU, 0x53a4a4f7U, 0xe4727296U, 0x9bc0c05bU,
66     0x75b7b7c2U, 0xe1fdfd1cU, 0x3d9393aeU, 0x4c26266aU,
67     0x6c36365aU, 0x7e3f3f41U, 0xf5f7f702U, 0x83cccc4fU,
68     0x6834345cU, 0x51a5a5f4U, 0xd1e5e534U, 0xf9f1f108U,
69     0xe2717193U, 0xabd8d873U, 0x62313153U, 0x2a15153fU,
70     0x0804040cU, 0x95c7c752U, 0x46232365U, 0x9dc3c35eU,
71     0x30181828U, 0x379696a1U, 0x0a05050fU, 0x2f9a9ab5U,
72     0x0e070709U, 0x24121236U, 0x1b80809bU, 0xdfe2e23dU,
73     0xcdebeb26U, 0x4e272769U, 0x7fb2b2cdU, 0xea75759fU,
74     0x1209091bU, 0x1d83839eU, 0x582c2c74U, 0x341a1a2eU,
75     0x361b1b2dU, 0xdc6e6eb2U, 0xb45a5aeeU, 0x5ba0a0fbU,
76     0xa45252f6U, 0x763b3b4dU, 0xb7d6d661U, 0x7db3b3ceU,
77     0x5229297bU, 0xdde3e33eU, 0x5e2f2f71U, 0x13848497U,
78     0xa65353f5U, 0xb9d1d168U, 0x00000000U, 0xc1eded2cU,
79     0x40202060U, 0xe3fcfc1fU, 0x79b1b1c8U, 0xb65b5bedU,
80     0xd46a6abeU, 0x8dcbcb46U, 0x67bebed9U, 0x7239394bU,
81     0x944a4adeU, 0x984c4cd4U, 0xb05858e8U, 0x85cfcf4aU,
82     0xbbd0d06bU, 0xc5efef2aU, 0x4faaaae5U, 0xedfbfb16U,
83     0x864343c5U, 0x9a4d4dd7U, 0x66333355U, 0x11858594U,
84     0x8a4545cfU, 0xe9f9f910U, 0x04020206U, 0xfe7f7f81U,
85     0xa05050f0U, 0x783c3c44U, 0x259f9fbaU, 0x4ba8a8e3U,
86     0xa25151f3U, 0x5da3a3feU, 0x804040c0U, 0x058f8f8aU,
87     0x3f9292adU, 0x219d9dbcU, 0x70383848U, 0xf1f5f504U,
88     0x63bcbcdfU, 0x77b6b6c1U, 0xafdada75U, 0x42212163U,
89     0x20101030U, 0xe5ffff1aU, 0xfdf3f30eU, 0xbfd2d26dU,
90     0x81cdcd4cU, 0x180c0c14U, 0x26131335U, 0xc3ecec2fU,
91     0xbe5f5fe1U, 0x359797a2U, 0x884444ccU, 0x2e171739U,
92     0x93c4c457U, 0x55a7a7f2U, 0xfc7e7e82U, 0x7a3d3d47U,
93     0xc86464acU, 0xba5d5de7U, 0x3219192bU, 0xe6737395U,
94     0xc06060a0U, 0x19818198U, 0x9e4f4fd1U, 0xa3dcdc7fU,
95     0x44222266U, 0x542a2a7eU, 0x3b9090abU, 0x0b888883U,
96     0x8c4646caU, 0xc7eeee29U, 0x6bb8b8d3U, 0x2814143cU,
97     0xa7dede79U, 0xbc5e5ee2U, 0x160b0b1dU, 0xaddbdb76U,
98     0xdbe0e03bU, 0x64323256U, 0x743a3a4eU, 0x140a0a1eU,
99     0x924949dbU, 0x0c06060aU, 0x4824246cU, 0xb85c5ce4U,
100     0x9fc2c25dU, 0xbdd3d36eU, 0x43acacefU, 0xc46262a6U,
101     0x399191a8U, 0x319595a4U, 0xd3e4e437U, 0xf279798bU,
102     0xd5e7e732U, 0x8bc8c843U, 0x6e373759U, 0xda6d6db7U,
103     0x018d8d8cU, 0xb1d5d564U, 0x9c4e4ed2U, 0x49a9a9e0U,
104     0xd86c6cb4U, 0xac5656faU, 0xf3f4f407U, 0xcfeaea25U,
105     0xca6565afU, 0xf47a7a8eU, 0x47aeaee9U, 0x10080818U,
106     0x6fbabad5U, 0xf0787888U, 0x4a25256fU, 0x5c2e2e72U,
107     0x381c1c24U, 0x57a6a6f1U, 0x73b4b4c7U, 0x97c6c651U,
108     0xcbe8e823U, 0xa1dddd7cU, 0xe874749cU, 0x3e1f1f21U,
109     0x964b4bddU, 0x61bdbddcU, 0x0d8b8b86U, 0x0f8a8a85U,
110     0xe0707090U, 0x7c3e3e42U, 0x71b5b5c4U, 0xcc6666aaU,
111     0x904848d8U, 0x06030305U, 0xf7f6f601U, 0x1c0e0e12U,
112     0xc26161a3U, 0x6a35355fU, 0xae5757f9U, 0x69b9b9d0U,
113     0x17868691U, 0x99c1c158U, 0x3a1d1d27U, 0x279e9eb9U,
114     0xd9e1e138U, 0xebf8f813U, 0x2b9898b3U, 0x22111133U,
115     0xd26969bbU, 0xa9d9d970U, 0x078e8e89U, 0x339494a7U,
116     0x2d9b9bb6U, 0x3c1e1e22U, 0x15878792U, 0xc9e9e920U,
117     0x87cece49U, 0xaa5555ffU, 0x50282878U, 0xa5dfdf7aU,
118     0x038c8c8fU, 0x59a1a1f8U, 0x09898980U, 0x1a0d0d17U,
119     0x65bfbfdaU, 0xd7e6e631U, 0x844242c6U, 0xd06868b8U,
120     0x824141c3U, 0x299999b0U, 0x5a2d2d77U, 0x1e0f0f11U,
121     0x7bb0b0cbU, 0xa85454fcU, 0x6dbbbbd6U, 0x2c16163aU,
122 };
123 static const u32 Te1[256] = {
124     0xa5c66363U, 0x84f87c7cU, 0x99ee7777U, 0x8df67b7bU,
125     0x0dfff2f2U, 0xbdd66b6bU, 0xb1de6f6fU, 0x5491c5c5U,
126     0x50603030U, 0x03020101U, 0xa9ce6767U, 0x7d562b2bU,
127     0x19e7fefeU, 0x62b5d7d7U, 0xe64dababU, 0x9aec7676U,
128     0x458fcacaU, 0x9d1f8282U, 0x4089c9c9U, 0x87fa7d7dU,
129     0x15effafaU, 0xebb25959U, 0xc98e4747U, 0x0bfbf0f0U,
130     0xec41adadU, 0x67b3d4d4U, 0xfd5fa2a2U, 0xea45afafU,
131     0xbf239c9cU, 0xf753a4a4U, 0x96e47272U, 0x5b9bc0c0U,
132     0xc275b7b7U, 0x1ce1fdfdU, 0xae3d9393U, 0x6a4c2626U,
133     0x5a6c3636U, 0x417e3f3fU, 0x02f5f7f7U, 0x4f83ccccU,
134     0x5c683434U, 0xf451a5a5U, 0x34d1e5e5U, 0x08f9f1f1U,
135     0x93e27171U, 0x73abd8d8U, 0x53623131U, 0x3f2a1515U,
136     0x0c080404U, 0x5295c7c7U, 0x65462323U, 0x5e9dc3c3U,
137     0x28301818U, 0xa1379696U, 0x0f0a0505U, 0xb52f9a9aU,
138     0x090e0707U, 0x36241212U, 0x9b1b8080U, 0x3ddfe2e2U,
139     0x26cdebebU, 0x694e2727U, 0xcd7fb2b2U, 0x9fea7575U,
140     0x1b120909U, 0x9e1d8383U, 0x74582c2cU, 0x2e341a1aU,
141     0x2d361b1bU, 0xb2dc6e6eU, 0xeeb45a5aU, 0xfb5ba0a0U,
142     0xf6a45252U, 0x4d763b3bU, 0x61b7d6d6U, 0xce7db3b3U,
143     0x7b522929U, 0x3edde3e3U, 0x715e2f2fU, 0x97138484U,
144     0xf5a65353U, 0x68b9d1d1U, 0x00000000U, 0x2cc1ededU,
145     0x60402020U, 0x1fe3fcfcU, 0xc879b1b1U, 0xedb65b5bU,
146     0xbed46a6aU, 0x468dcbcbU, 0xd967bebeU, 0x4b723939U,
147     0xde944a4aU, 0xd4984c4cU, 0xe8b05858U, 0x4a85cfcfU,
148     0x6bbbd0d0U, 0x2ac5efefU, 0xe54faaaaU, 0x16edfbfbU,
149     0xc5864343U, 0xd79a4d4dU, 0x55663333U, 0x94118585U,
150     0xcf8a4545U, 0x10e9f9f9U, 0x06040202U, 0x81fe7f7fU,
151     0xf0a05050U, 0x44783c3cU, 0xba259f9fU, 0xe34ba8a8U,
152     0xf3a25151U, 0xfe5da3a3U, 0xc0804040U, 0x8a058f8fU,
153     0xad3f9292U, 0xbc219d9dU, 0x48703838U, 0x04f1f5f5U,
154     0xdf63bcbcU, 0xc177b6b6U, 0x75afdadaU, 0x63422121U,
155     0x30201010U, 0x1ae5ffffU, 0x0efdf3f3U, 0x6dbfd2d2U,
156     0x4c81cdcdU, 0x14180c0cU, 0x35261313U, 0x2fc3ececU,
157     0xe1be5f5fU, 0xa2359797U, 0xcc884444U, 0x392e1717U,
158     0x5793c4c4U, 0xf255a7a7U, 0x82fc7e7eU, 0x477a3d3dU,
159     0xacc86464U, 0xe7ba5d5dU, 0x2b321919U, 0x95e67373U,
160     0xa0c06060U, 0x98198181U, 0xd19e4f4fU, 0x7fa3dcdcU,
161     0x66442222U, 0x7e542a2aU, 0xab3b9090U, 0x830b8888U,
162     0xca8c4646U, 0x29c7eeeeU, 0xd36bb8b8U, 0x3c281414U,
163     0x79a7dedeU, 0xe2bc5e5eU, 0x1d160b0bU, 0x76addbdbU,
164     0x3bdbe0e0U, 0x56643232U, 0x4e743a3aU, 0x1e140a0aU,
165     0xdb924949U, 0x0a0c0606U, 0x6c482424U, 0xe4b85c5cU,
166     0x5d9fc2c2U, 0x6ebdd3d3U, 0xef43acacU, 0xa6c46262U,
167     0xa8399191U, 0xa4319595U, 0x37d3e4e4U, 0x8bf27979U,
168     0x32d5e7e7U, 0x438bc8c8U, 0x596e3737U, 0xb7da6d6dU,
169     0x8c018d8dU, 0x64b1d5d5U, 0xd29c4e4eU, 0xe049a9a9U,
170     0xb4d86c6cU, 0xfaac5656U, 0x07f3f4f4U, 0x25cfeaeaU,
171     0xafca6565U, 0x8ef47a7aU, 0xe947aeaeU, 0x18100808U,
172     0xd56fbabaU, 0x88f07878U, 0x6f4a2525U, 0x725c2e2eU,
173     0x24381c1cU, 0xf157a6a6U, 0xc773b4b4U, 0x5197c6c6U,
174     0x23cbe8e8U, 0x7ca1ddddU, 0x9ce87474U, 0x213e1f1fU,
175     0xdd964b4bU, 0xdc61bdbdU, 0x860d8b8bU, 0x850f8a8aU,
176     0x90e07070U, 0x427c3e3eU, 0xc471b5b5U, 0xaacc6666U,
177     0xd8904848U, 0x05060303U, 0x01f7f6f6U, 0x121c0e0eU,
178     0xa3c26161U, 0x5f6a3535U, 0xf9ae5757U, 0xd069b9b9U,
179     0x91178686U, 0x5899c1c1U, 0x273a1d1dU, 0xb9279e9eU,
180     0x38d9e1e1U, 0x13ebf8f8U, 0xb32b9898U, 0x33221111U,
181     0xbbd26969U, 0x70a9d9d9U, 0x89078e8eU, 0xa7339494U,
182     0xb62d9b9bU, 0x223c1e1eU, 0x92158787U, 0x20c9e9e9U,
183     0x4987ceceU, 0xffaa5555U, 0x78502828U, 0x7aa5dfdfU,
184     0x8f038c8cU, 0xf859a1a1U, 0x80098989U, 0x171a0d0dU,
185     0xda65bfbfU, 0x31d7e6e6U, 0xc6844242U, 0xb8d06868U,
186     0xc3824141U, 0xb0299999U, 0x775a2d2dU, 0x111e0f0fU,
187     0xcb7bb0b0U, 0xfca85454U, 0xd66dbbbbU, 0x3a2c1616U,
188 };
189 static const u32 Te2[256] = {
190     0x63a5c663U, 0x7c84f87cU, 0x7799ee77U, 0x7b8df67bU,
191     0xf20dfff2U, 0x6bbdd66bU, 0x6fb1de6fU, 0xc55491c5U,
192     0x30506030U, 0x01030201U, 0x67a9ce67U, 0x2b7d562bU,
193     0xfe19e7feU, 0xd762b5d7U, 0xabe64dabU, 0x769aec76U,
194     0xca458fcaU, 0x829d1f82U, 0xc94089c9U, 0x7d87fa7dU,
195     0xfa15effaU, 0x59ebb259U, 0x47c98e47U, 0xf00bfbf0U,
196     0xadec41adU, 0xd467b3d4U, 0xa2fd5fa2U, 0xafea45afU,
197     0x9cbf239cU, 0xa4f753a4U, 0x7296e472U, 0xc05b9bc0U,
198     0xb7c275b7U, 0xfd1ce1fdU, 0x93ae3d93U, 0x266a4c26U,
199     0x365a6c36U, 0x3f417e3fU, 0xf702f5f7U, 0xcc4f83ccU,
200     0x345c6834U, 0xa5f451a5U, 0xe534d1e5U, 0xf108f9f1U,
201     0x7193e271U, 0xd873abd8U, 0x31536231U, 0x153f2a15U,
202     0x040c0804U, 0xc75295c7U, 0x23654623U, 0xc35e9dc3U,
203     0x18283018U, 0x96a13796U, 0x050f0a05U, 0x9ab52f9aU,
204     0x07090e07U, 0x12362412U, 0x809b1b80U, 0xe23ddfe2U,
205     0xeb26cdebU, 0x27694e27U, 0xb2cd7fb2U, 0x759fea75U,
206     0x091b1209U, 0x839e1d83U, 0x2c74582cU, 0x1a2e341aU,
207     0x1b2d361bU, 0x6eb2dc6eU, 0x5aeeb45aU, 0xa0fb5ba0U,
208     0x52f6a452U, 0x3b4d763bU, 0xd661b7d6U, 0xb3ce7db3U,
209     0x297b5229U, 0xe33edde3U, 0x2f715e2fU, 0x84971384U,
210     0x53f5a653U, 0xd168b9d1U, 0x00000000U, 0xed2cc1edU,
211     0x20604020U, 0xfc1fe3fcU, 0xb1c879b1U, 0x5bedb65bU,
212     0x6abed46aU, 0xcb468dcbU, 0xbed967beU, 0x394b7239U,
213     0x4ade944aU, 0x4cd4984cU, 0x58e8b058U, 0xcf4a85cfU,
214     0xd06bbbd0U, 0xef2ac5efU, 0xaae54faaU, 0xfb16edfbU,
215     0x43c58643U, 0x4dd79a4dU, 0x33556633U, 0x85941185U,
216     0x45cf8a45U, 0xf910e9f9U, 0x02060402U, 0x7f81fe7fU,
217     0x50f0a050U, 0x3c44783cU, 0x9fba259fU, 0xa8e34ba8U,
218     0x51f3a251U, 0xa3fe5da3U, 0x40c08040U, 0x8f8a058fU,
219     0x92ad3f92U, 0x9dbc219dU, 0x38487038U, 0xf504f1f5U,
220     0xbcdf63bcU, 0xb6c177b6U, 0xda75afdaU, 0x21634221U,
221     0x10302010U, 0xff1ae5ffU, 0xf30efdf3U, 0xd26dbfd2U,
222     0xcd4c81cdU, 0x0c14180cU, 0x13352613U, 0xec2fc3ecU,
223     0x5fe1be5fU, 0x97a23597U, 0x44cc8844U, 0x17392e17U,
224     0xc45793c4U, 0xa7f255a7U, 0x7e82fc7eU, 0x3d477a3dU,
225     0x64acc864U, 0x5de7ba5dU, 0x192b3219U, 0x7395e673U,
226     0x60a0c060U, 0x81981981U, 0x4fd19e4fU, 0xdc7fa3dcU,
227     0x22664422U, 0x2a7e542aU, 0x90ab3b90U, 0x88830b88U,
228     0x46ca8c46U, 0xee29c7eeU, 0xb8d36bb8U, 0x143c2814U,
229     0xde79a7deU, 0x5ee2bc5eU, 0x0b1d160bU, 0xdb76addbU,
230     0xe03bdbe0U, 0x32566432U, 0x3a4e743aU, 0x0a1e140aU,
231     0x49db9249U, 0x060a0c06U, 0x246c4824U, 0x5ce4b85cU,
232     0xc25d9fc2U, 0xd36ebdd3U, 0xacef43acU, 0x62a6c462U,
233     0x91a83991U, 0x95a43195U, 0xe437d3e4U, 0x798bf279U,
234     0xe732d5e7U, 0xc8438bc8U, 0x37596e37U, 0x6db7da6dU,
235     0x8d8c018dU, 0xd564b1d5U, 0x4ed29c4eU, 0xa9e049a9U,
236     0x6cb4d86cU, 0x56faac56U, 0xf407f3f4U, 0xea25cfeaU,
237     0x65afca65U, 0x7a8ef47aU, 0xaee947aeU, 0x08181008U,
238     0xbad56fbaU, 0x7888f078U, 0x256f4a25U, 0x2e725c2eU,
239     0x1c24381cU, 0xa6f157a6U, 0xb4c773b4U, 0xc65197c6U,
240     0xe823cbe8U, 0xdd7ca1ddU, 0x749ce874U, 0x1f213e1fU,
241     0x4bdd964bU, 0xbddc61bdU, 0x8b860d8bU, 0x8a850f8aU,
242     0x7090e070U, 0x3e427c3eU, 0xb5c471b5U, 0x66aacc66U,
243     0x48d89048U, 0x03050603U, 0xf601f7f6U, 0x0e121c0eU,
244     0x61a3c261U, 0x355f6a35U, 0x57f9ae57U, 0xb9d069b9U,
245     0x86911786U, 0xc15899c1U, 0x1d273a1dU, 0x9eb9279eU,
246     0xe138d9e1U, 0xf813ebf8U, 0x98b32b98U, 0x11332211U,
247     0x69bbd269U, 0xd970a9d9U, 0x8e89078eU, 0x94a73394U,
248     0x9bb62d9bU, 0x1e223c1eU, 0x87921587U, 0xe920c9e9U,
249     0xce4987ceU, 0x55ffaa55U, 0x28785028U, 0xdf7aa5dfU,
250     0x8c8f038cU, 0xa1f859a1U, 0x89800989U, 0x0d171a0dU,
251     0xbfda65bfU, 0xe631d7e6U, 0x42c68442U, 0x68b8d068U,
252     0x41c38241U, 0x99b02999U, 0x2d775a2dU, 0x0f111e0fU,
253     0xb0cb7bb0U, 0x54fca854U, 0xbbd66dbbU, 0x163a2c16U,
254 };
255 static const u32 Te3[256] = {
256     0x6363a5c6U, 0x7c7c84f8U, 0x777799eeU, 0x7b7b8df6U,
257     0xf2f20dffU, 0x6b6bbdd6U, 0x6f6fb1deU, 0xc5c55491U,
258     0x30305060U, 0x01010302U, 0x6767a9ceU, 0x2b2b7d56U,
259     0xfefe19e7U, 0xd7d762b5U, 0xababe64dU, 0x76769aecU,
260     0xcaca458fU, 0x82829d1fU, 0xc9c94089U, 0x7d7d87faU,
261     0xfafa15efU, 0x5959ebb2U, 0x4747c98eU, 0xf0f00bfbU,
262     0xadadec41U, 0xd4d467b3U, 0xa2a2fd5fU, 0xafafea45U,
263     0x9c9cbf23U, 0xa4a4f753U, 0x727296e4U, 0xc0c05b9bU,
264     0xb7b7c275U, 0xfdfd1ce1U, 0x9393ae3dU, 0x26266a4cU,
265     0x36365a6cU, 0x3f3f417eU, 0xf7f702f5U, 0xcccc4f83U,
266     0x34345c68U, 0xa5a5f451U, 0xe5e534d1U, 0xf1f108f9U,
267     0x717193e2U, 0xd8d873abU, 0x31315362U, 0x15153f2aU,
268     0x04040c08U, 0xc7c75295U, 0x23236546U, 0xc3c35e9dU,
269     0x18182830U, 0x9696a137U, 0x05050f0aU, 0x9a9ab52fU,
270     0x0707090eU, 0x12123624U, 0x80809b1bU, 0xe2e23ddfU,
271     0xebeb26cdU, 0x2727694eU, 0xb2b2cd7fU, 0x75759feaU,
272     0x09091b12U, 0x83839e1dU, 0x2c2c7458U, 0x1a1a2e34U,
273     0x1b1b2d36U, 0x6e6eb2dcU, 0x5a5aeeb4U, 0xa0a0fb5bU,
274     0x5252f6a4U, 0x3b3b4d76U, 0xd6d661b7U, 0xb3b3ce7dU,
275     0x29297b52U, 0xe3e33eddU, 0x2f2f715eU, 0x84849713U,
276     0x5353f5a6U, 0xd1d168b9U, 0x00000000U, 0xeded2cc1U,
277     0x20206040U, 0xfcfc1fe3U, 0xb1b1c879U, 0x5b5bedb6U,
278     0x6a6abed4U, 0xcbcb468dU, 0xbebed967U, 0x39394b72U,
279     0x4a4ade94U, 0x4c4cd498U, 0x5858e8b0U, 0xcfcf4a85U,
280     0xd0d06bbbU, 0xefef2ac5U, 0xaaaae54fU, 0xfbfb16edU,
281     0x4343c586U, 0x4d4dd79aU, 0x33335566U, 0x85859411U,
282     0x4545cf8aU, 0xf9f910e9U, 0x02020604U, 0x7f7f81feU,
283     0x5050f0a0U, 0x3c3c4478U, 0x9f9fba25U, 0xa8a8e34bU,
284     0x5151f3a2U, 0xa3a3fe5dU, 0x4040c080U, 0x8f8f8a05U,
285     0x9292ad3fU, 0x9d9dbc21U, 0x38384870U, 0xf5f504f1U,
286     0xbcbcdf63U, 0xb6b6c177U, 0xdada75afU, 0x21216342U,
287     0x10103020U, 0xffff1ae5U, 0xf3f30efdU, 0xd2d26dbfU,
288     0xcdcd4c81U, 0x0c0c1418U, 0x13133526U, 0xecec2fc3U,
289     0x5f5fe1beU, 0x9797a235U, 0x4444cc88U, 0x1717392eU,
290     0xc4c45793U, 0xa7a7f255U, 0x7e7e82fcU, 0x3d3d477aU,
291     0x6464acc8U, 0x5d5de7baU, 0x19192b32U, 0x737395e6U,
292     0x6060a0c0U, 0x81819819U, 0x4f4fd19eU, 0xdcdc7fa3U,
293     0x22226644U, 0x2a2a7e54U, 0x9090ab3bU, 0x8888830bU,
294     0x4646ca8cU, 0xeeee29c7U, 0xb8b8d36bU, 0x14143c28U,
295     0xdede79a7U, 0x5e5ee2bcU, 0x0b0b1d16U, 0xdbdb76adU,
296     0xe0e03bdbU, 0x32325664U, 0x3a3a4e74U, 0x0a0a1e14U,
297     0x4949db92U, 0x06060a0cU, 0x24246c48U, 0x5c5ce4b8U,
298     0xc2c25d9fU, 0xd3d36ebdU, 0xacacef43U, 0x6262a6c4U,
299     0x9191a839U, 0x9595a431U, 0xe4e437d3U, 0x79798bf2U,
300     0xe7e732d5U, 0xc8c8438bU, 0x3737596eU, 0x6d6db7daU,
301     0x8d8d8c01U, 0xd5d564b1U, 0x4e4ed29cU, 0xa9a9e049U,
302     0x6c6cb4d8U, 0x5656faacU, 0xf4f407f3U, 0xeaea25cfU,
303     0x6565afcaU, 0x7a7a8ef4U, 0xaeaee947U, 0x08081810U,
304     0xbabad56fU, 0x787888f0U, 0x25256f4aU, 0x2e2e725cU,
305     0x1c1c2438U, 0xa6a6f157U, 0xb4b4c773U, 0xc6c65197U,
306     0xe8e823cbU, 0xdddd7ca1U, 0x74749ce8U, 0x1f1f213eU,
307     0x4b4bdd96U, 0xbdbddc61U, 0x8b8b860dU, 0x8a8a850fU,
308     0x707090e0U, 0x3e3e427cU, 0xb5b5c471U, 0x6666aaccU,
309     0x4848d890U, 0x03030506U, 0xf6f601f7U, 0x0e0e121cU,
310     0x6161a3c2U, 0x35355f6aU, 0x5757f9aeU, 0xb9b9d069U,
311     0x86869117U, 0xc1c15899U, 0x1d1d273aU, 0x9e9eb927U,
312     0xe1e138d9U, 0xf8f813ebU, 0x9898b32bU, 0x11113322U,
313     0x6969bbd2U, 0xd9d970a9U, 0x8e8e8907U, 0x9494a733U,
314     0x9b9bb62dU, 0x1e1e223cU, 0x87879215U, 0xe9e920c9U,
315     0xcece4987U, 0x5555ffaaU, 0x28287850U, 0xdfdf7aa5U,
316     0x8c8c8f03U, 0xa1a1f859U, 0x89898009U, 0x0d0d171aU,
317     0xbfbfda65U, 0xe6e631d7U, 0x4242c684U, 0x6868b8d0U,
318     0x4141c382U, 0x9999b029U, 0x2d2d775aU, 0x0f0f111eU,
319     0xb0b0cb7bU, 0x5454fca8U, 0xbbbbd66dU, 0x16163a2cU,
320 };
321
322 static const u32 Td0[256] = {
323     0x51f4a750U, 0x7e416553U, 0x1a17a4c3U, 0x3a275e96U,
324     0x3bab6bcbU, 0x1f9d45f1U, 0xacfa58abU, 0x4be30393U,
325     0x2030fa55U, 0xad766df6U, 0x88cc7691U, 0xf5024c25U,
326     0x4fe5d7fcU, 0xc52acbd7U, 0x26354480U, 0xb562a38fU,
327     0xdeb15a49U, 0x25ba1b67U, 0x45ea0e98U, 0x5dfec0e1U,
328     0xc32f7502U, 0x814cf012U, 0x8d4697a3U, 0x6bd3f9c6U,
329     0x038f5fe7U, 0x15929c95U, 0xbf6d7aebU, 0x955259daU,
330     0xd4be832dU, 0x587421d3U, 0x49e06929U, 0x8ec9c844U,
331     0x75c2896aU, 0xf48e7978U, 0x99583e6bU, 0x27b971ddU,
332     0xbee14fb6U, 0xf088ad17U, 0xc920ac66U, 0x7dce3ab4U,
333     0x63df4a18U, 0xe51a3182U, 0x97513360U, 0x62537f45U,
334     0xb16477e0U, 0xbb6bae84U, 0xfe81a01cU, 0xf9082b94U,
335     0x70486858U, 0x8f45fd19U, 0x94de6c87U, 0x527bf8b7U,
336     0xab73d323U, 0x724b02e2U, 0xe31f8f57U, 0x6655ab2aU,
337     0xb2eb2807U, 0x2fb5c203U, 0x86c57b9aU, 0xd33708a5U,
338     0x302887f2U, 0x23bfa5b2U, 0x02036abaU, 0xed16825cU,
339     0x8acf1c2bU, 0xa779b492U, 0xf307f2f0U, 0x4e69e2a1U,
340     0x65daf4cdU, 0x0605bed5U, 0xd134621fU, 0xc4a6fe8aU,
341     0x342e539dU, 0xa2f355a0U, 0x058ae132U, 0xa4f6eb75U,
342     0x0b83ec39U, 0x4060efaaU, 0x5e719f06U, 0xbd6e1051U,
343     0x3e218af9U, 0x96dd063dU, 0xdd3e05aeU, 0x4de6bd46U,
344     0x91548db5U, 0x71c45d05U, 0x0406d46fU, 0x605015ffU,
345     0x1998fb24U, 0xd6bde997U, 0x894043ccU, 0x67d99e77U,
346     0xb0e842bdU, 0x07898b88U, 0xe7195b38U, 0x79c8eedbU,
347     0xa17c0a47U, 0x7c420fe9U, 0xf8841ec9U, 0x00000000U,
348     0x09808683U, 0x322bed48U, 0x1e1170acU, 0x6c5a724eU,
349     0xfd0efffbU, 0x0f853856U, 0x3daed51eU, 0x362d3927U,
350     0x0a0fd964U, 0x685ca621U, 0x9b5b54d1U, 0x24362e3aU,
351     0x0c0a67b1U, 0x9357e70fU, 0xb4ee96d2U, 0x1b9b919eU,
352     0x80c0c54fU, 0x61dc20a2U, 0x5a774b69U, 0x1c121a16U,
353     0xe293ba0aU, 0xc0a02ae5U, 0x3c22e043U, 0x121b171dU,
354     0x0e090d0bU, 0xf28bc7adU, 0x2db6a8b9U, 0x141ea9c8U,
355     0x57f11985U, 0xaf75074cU, 0xee99ddbbU, 0xa37f60fdU,
356     0xf701269fU, 0x5c72f5bcU, 0x44663bc5U, 0x5bfb7e34U,
357     0x8b432976U, 0xcb23c6dcU, 0xb6edfc68U, 0xb8e4f163U,
358     0xd731dccaU, 0x42638510U, 0x13972240U, 0x84c61120U,
359     0x854a247dU, 0xd2bb3df8U, 0xaef93211U, 0xc729a16dU,
360     0x1d9e2f4bU, 0xdcb230f3U, 0x0d8652ecU, 0x77c1e3d0U,
361     0x2bb3166cU, 0xa970b999U, 0x119448faU, 0x47e96422U,
362     0xa8fc8cc4U, 0xa0f03f1aU, 0x567d2cd8U, 0x223390efU,
363     0x87494ec7U, 0xd938d1c1U, 0x8ccaa2feU, 0x98d40b36U,
364     0xa6f581cfU, 0xa57ade28U, 0xdab78e26U, 0x3fadbfa4U,
365     0x2c3a9de4U, 0x5078920dU, 0x6a5fcc9bU, 0x547e4662U,
366     0xf68d13c2U, 0x90d8b8e8U, 0x2e39f75eU, 0x82c3aff5U,
367     0x9f5d80beU, 0x69d0937cU, 0x6fd52da9U, 0xcf2512b3U,
368     0xc8ac993bU, 0x10187da7U, 0xe89c636eU, 0xdb3bbb7bU,
369     0xcd267809U, 0x6e5918f4U, 0xec9ab701U, 0x834f9aa8U,
370     0xe6956e65U, 0xaaffe67eU, 0x21bccf08U, 0xef15e8e6U,
371     0xbae79bd9U, 0x4a6f36ceU, 0xea9f09d4U, 0x29b07cd6U,
372     0x31a4b2afU, 0x2a3f2331U, 0xc6a59430U, 0x35a266c0U,
373     0x744ebc37U, 0xfc82caa6U, 0xe090d0b0U, 0x33a7d815U,
374     0xf104984aU, 0x41ecdaf7U, 0x7fcd500eU, 0x1791f62fU,
375     0x764dd68dU, 0x43efb04dU, 0xccaa4d54U, 0xe49604dfU,
376     0x9ed1b5e3U, 0x4c6a881bU, 0xc12c1fb8U, 0x4665517fU,
377     0x9d5eea04U, 0x018c355dU, 0xfa877473U, 0xfb0b412eU,
378     0xb3671d5aU, 0x92dbd252U, 0xe9105633U, 0x6dd64713U,
379     0x9ad7618cU, 0x37a10c7aU, 0x59f8148eU, 0xeb133c89U,
380     0xcea927eeU, 0xb761c935U, 0xe11ce5edU, 0x7a47b13cU,
381     0x9cd2df59U, 0x55f2733fU, 0x1814ce79U, 0x73c737bfU,
382     0x53f7cdeaU, 0x5ffdaa5bU, 0xdf3d6f14U, 0x7844db86U,
383     0xcaaff381U, 0xb968c43eU, 0x3824342cU, 0xc2a3405fU,
384     0x161dc372U, 0xbce2250cU, 0x283c498bU, 0xff0d9541U,
385     0x39a80171U, 0x080cb3deU, 0xd8b4e49cU, 0x6456c190U,
386     0x7bcb8461U, 0xd532b670U, 0x486c5c74U, 0xd0b85742U,
387 };
388 static const u32 Td1[256] = {
389     0x5051f4a7U, 0x537e4165U, 0xc31a17a4U, 0x963a275eU,
390     0xcb3bab6bU, 0xf11f9d45U, 0xabacfa58U, 0x934be303U,
391     0x552030faU, 0xf6ad766dU, 0x9188cc76U, 0x25f5024cU,
392     0xfc4fe5d7U, 0xd7c52acbU, 0x80263544U, 0x8fb562a3U,
393     0x49deb15aU, 0x6725ba1bU, 0x9845ea0eU, 0xe15dfec0U,
394     0x02c32f75U, 0x12814cf0U, 0xa38d4697U, 0xc66bd3f9U,
395     0xe7038f5fU, 0x9515929cU, 0xebbf6d7aU, 0xda955259U,
396     0x2dd4be83U, 0xd3587421U, 0x2949e069U, 0x448ec9c8U,
397     0x6a75c289U, 0x78f48e79U, 0x6b99583eU, 0xdd27b971U,
398     0xb6bee14fU, 0x17f088adU, 0x66c920acU, 0xb47dce3aU,
399     0x1863df4aU, 0x82e51a31U, 0x60975133U, 0x4562537fU,
400     0xe0b16477U, 0x84bb6baeU, 0x1cfe81a0U, 0x94f9082bU,
401     0x58704868U, 0x198f45fdU, 0x8794de6cU, 0xb7527bf8U,
402     0x23ab73d3U, 0xe2724b02U, 0x57e31f8fU, 0x2a6655abU,
403     0x07b2eb28U, 0x032fb5c2U, 0x9a86c57bU, 0xa5d33708U,
404     0xf2302887U, 0xb223bfa5U, 0xba02036aU, 0x5ced1682U,
405     0x2b8acf1cU, 0x92a779b4U, 0xf0f307f2U, 0xa14e69e2U,
406     0xcd65daf4U, 0xd50605beU, 0x1fd13462U, 0x8ac4a6feU,
407     0x9d342e53U, 0xa0a2f355U, 0x32058ae1U, 0x75a4f6ebU,
408     0x390b83ecU, 0xaa4060efU, 0x065e719fU, 0x51bd6e10U,
409     0xf93e218aU, 0x3d96dd06U, 0xaedd3e05U, 0x464de6bdU,
410     0xb591548dU, 0x0571c45dU, 0x6f0406d4U, 0xff605015U,
411     0x241998fbU, 0x97d6bde9U, 0xcc894043U, 0x7767d99eU,
412     0xbdb0e842U, 0x8807898bU, 0x38e7195bU, 0xdb79c8eeU,
413     0x47a17c0aU, 0xe97c420fU, 0xc9f8841eU, 0x00000000U,
414     0x83098086U, 0x48322bedU, 0xac1e1170U, 0x4e6c5a72U,
415     0xfbfd0effU, 0x560f8538U, 0x1e3daed5U, 0x27362d39U,
416     0x640a0fd9U, 0x21685ca6U, 0xd19b5b54U, 0x3a24362eU,
417     0xb10c0a67U, 0x0f9357e7U, 0xd2b4ee96U, 0x9e1b9b91U,
418     0x4f80c0c5U, 0xa261dc20U, 0x695a774bU, 0x161c121aU,
419     0x0ae293baU, 0xe5c0a02aU, 0x433c22e0U, 0x1d121b17U,
420     0x0b0e090dU, 0xadf28bc7U, 0xb92db6a8U, 0xc8141ea9U,
421     0x8557f119U, 0x4caf7507U, 0xbbee99ddU, 0xfda37f60U,
422     0x9ff70126U, 0xbc5c72f5U, 0xc544663bU, 0x345bfb7eU,
423     0x768b4329U, 0xdccb23c6U, 0x68b6edfcU, 0x63b8e4f1U,
424     0xcad731dcU, 0x10426385U, 0x40139722U, 0x2084c611U,
425     0x7d854a24U, 0xf8d2bb3dU, 0x11aef932U, 0x6dc729a1U,
426     0x4b1d9e2fU, 0xf3dcb230U, 0xec0d8652U, 0xd077c1e3U,
427     0x6c2bb316U, 0x99a970b9U, 0xfa119448U, 0x2247e964U,
428     0xc4a8fc8cU, 0x1aa0f03fU, 0xd8567d2cU, 0xef223390U,
429     0xc787494eU, 0xc1d938d1U, 0xfe8ccaa2U, 0x3698d40bU,
430     0xcfa6f581U, 0x28a57adeU, 0x26dab78eU, 0xa43fadbfU,
431     0xe42c3a9dU, 0x0d507892U, 0x9b6a5fccU, 0x62547e46U,
432     0xc2f68d13U, 0xe890d8b8U, 0x5e2e39f7U, 0xf582c3afU,
433     0xbe9f5d80U, 0x7c69d093U, 0xa96fd52dU, 0xb3cf2512U,
434     0x3bc8ac99U, 0xa710187dU, 0x6ee89c63U, 0x7bdb3bbbU,
435     0x09cd2678U, 0xf46e5918U, 0x01ec9ab7U, 0xa8834f9aU,
436     0x65e6956eU, 0x7eaaffe6U, 0x0821bccfU, 0xe6ef15e8U,
437     0xd9bae79bU, 0xce4a6f36U, 0xd4ea9f09U, 0xd629b07cU,
438     0xaf31a4b2U, 0x312a3f23U, 0x30c6a594U, 0xc035a266U,
439     0x37744ebcU, 0xa6fc82caU, 0xb0e090d0U, 0x1533a7d8U,
440     0x4af10498U, 0xf741ecdaU, 0x0e7fcd50U, 0x2f1791f6U,
441     0x8d764dd6U, 0x4d43efb0U, 0x54ccaa4dU, 0xdfe49604U,
442     0xe39ed1b5U, 0x1b4c6a88U, 0xb8c12c1fU, 0x7f466551U,
443     0x049d5eeaU, 0x5d018c35U, 0x73fa8774U, 0x2efb0b41U,
444     0x5ab3671dU, 0x5292dbd2U, 0x33e91056U, 0x136dd647U,
445     0x8c9ad761U, 0x7a37a10cU, 0x8e59f814U, 0x89eb133cU,
446     0xeecea927U, 0x35b761c9U, 0xede11ce5U, 0x3c7a47b1U,
447     0x599cd2dfU, 0x3f55f273U, 0x791814ceU, 0xbf73c737U,
448     0xea53f7cdU, 0x5b5ffdaaU, 0x14df3d6fU, 0x867844dbU,
449     0x81caaff3U, 0x3eb968c4U, 0x2c382434U, 0x5fc2a340U,
450     0x72161dc3U, 0x0cbce225U, 0x8b283c49U, 0x41ff0d95U,
451     0x7139a801U, 0xde080cb3U, 0x9cd8b4e4U, 0x906456c1U,
452     0x617bcb84U, 0x70d532b6U, 0x74486c5cU, 0x42d0b857U,
453 };
454 static const u32 Td2[256] = {
455     0xa75051f4U, 0x65537e41U, 0xa4c31a17U, 0x5e963a27U,
456     0x6bcb3babU, 0x45f11f9dU, 0x58abacfaU, 0x03934be3U,
457     0xfa552030U, 0x6df6ad76U, 0x769188ccU, 0x4c25f502U,
458     0xd7fc4fe5U, 0xcbd7c52aU, 0x44802635U, 0xa38fb562U,
459     0x5a49deb1U, 0x1b6725baU, 0x0e9845eaU, 0xc0e15dfeU,
460     0x7502c32fU, 0xf012814cU, 0x97a38d46U, 0xf9c66bd3U,
461     0x5fe7038fU, 0x9c951592U, 0x7aebbf6dU, 0x59da9552U,
462     0x832dd4beU, 0x21d35874U, 0x692949e0U, 0xc8448ec9U,
463     0x896a75c2U, 0x7978f48eU, 0x3e6b9958U, 0x71dd27b9U,
464     0x4fb6bee1U, 0xad17f088U, 0xac66c920U, 0x3ab47dceU,
465     0x4a1863dfU, 0x3182e51aU, 0x33609751U, 0x7f456253U,
466     0x77e0b164U, 0xae84bb6bU, 0xa01cfe81U, 0x2b94f908U,
467     0x68587048U, 0xfd198f45U, 0x6c8794deU, 0xf8b7527bU,
468     0xd323ab73U, 0x02e2724bU, 0x8f57e31fU, 0xab2a6655U,
469     0x2807b2ebU, 0xc2032fb5U, 0x7b9a86c5U, 0x08a5d337U,
470     0x87f23028U, 0xa5b223bfU, 0x6aba0203U, 0x825ced16U,
471     0x1c2b8acfU, 0xb492a779U, 0xf2f0f307U, 0xe2a14e69U,
472     0xf4cd65daU, 0xbed50605U, 0x621fd134U, 0xfe8ac4a6U,
473     0x539d342eU, 0x55a0a2f3U, 0xe132058aU, 0xeb75a4f6U,
474     0xec390b83U, 0xefaa4060U, 0x9f065e71U, 0x1051bd6eU,
475     0x8af93e21U, 0x063d96ddU, 0x05aedd3eU, 0xbd464de6U,
476     0x8db59154U, 0x5d0571c4U, 0xd46f0406U, 0x15ff6050U,
477     0xfb241998U, 0xe997d6bdU, 0x43cc8940U, 0x9e7767d9U,
478     0x42bdb0e8U, 0x8b880789U, 0x5b38e719U, 0xeedb79c8U,
479     0x0a47a17cU, 0x0fe97c42U, 0x1ec9f884U, 0x00000000U,
480     0x86830980U, 0xed48322bU, 0x70ac1e11U, 0x724e6c5aU,
481     0xfffbfd0eU, 0x38560f85U, 0xd51e3daeU, 0x3927362dU,
482     0xd9640a0fU, 0xa621685cU, 0x54d19b5bU, 0x2e3a2436U,
483     0x67b10c0aU, 0xe70f9357U, 0x96d2b4eeU, 0x919e1b9bU,
484     0xc54f80c0U, 0x20a261dcU, 0x4b695a77U, 0x1a161c12U,
485     0xba0ae293U, 0x2ae5c0a0U, 0xe0433c22U, 0x171d121bU,
486     0x0d0b0e09U, 0xc7adf28bU, 0xa8b92db6U, 0xa9c8141eU,
487     0x198557f1U, 0x074caf75U, 0xddbbee99U, 0x60fda37fU,
488     0x269ff701U, 0xf5bc5c72U, 0x3bc54466U, 0x7e345bfbU,
489     0x29768b43U, 0xc6dccb23U, 0xfc68b6edU, 0xf163b8e4U,
490     0xdccad731U, 0x85104263U, 0x22401397U, 0x112084c6U,
491     0x247d854aU, 0x3df8d2bbU, 0x3211aef9U, 0xa16dc729U,
492     0x2f4b1d9eU, 0x30f3dcb2U, 0x52ec0d86U, 0xe3d077c1U,
493     0x166c2bb3U, 0xb999a970U, 0x48fa1194U, 0x642247e9U,
494     0x8cc4a8fcU, 0x3f1aa0f0U, 0x2cd8567dU, 0x90ef2233U,
495     0x4ec78749U, 0xd1c1d938U, 0xa2fe8ccaU, 0x0b3698d4U,
496     0x81cfa6f5U, 0xde28a57aU, 0x8e26dab7U, 0xbfa43fadU,
497     0x9de42c3aU, 0x920d5078U, 0xcc9b6a5fU, 0x4662547eU,
498     0x13c2f68dU, 0xb8e890d8U, 0xf75e2e39U, 0xaff582c3U,
499     0x80be9f5dU, 0x937c69d0U, 0x2da96fd5U, 0x12b3cf25U,
500     0x993bc8acU, 0x7da71018U, 0x636ee89cU, 0xbb7bdb3bU,
501     0x7809cd26U, 0x18f46e59U, 0xb701ec9aU, 0x9aa8834fU,
502     0x6e65e695U, 0xe67eaaffU, 0xcf0821bcU, 0xe8e6ef15U,
503     0x9bd9bae7U, 0x36ce4a6fU, 0x09d4ea9fU, 0x7cd629b0U,
504     0xb2af31a4U, 0x23312a3fU, 0x9430c6a5U, 0x66c035a2U,
505     0xbc37744eU, 0xcaa6fc82U, 0xd0b0e090U, 0xd81533a7U,
506     0x984af104U, 0xdaf741ecU, 0x500e7fcdU, 0xf62f1791U,
507     0xd68d764dU, 0xb04d43efU, 0x4d54ccaaU, 0x04dfe496U,
508     0xb5e39ed1U, 0x881b4c6aU, 0x1fb8c12cU, 0x517f4665U,
509     0xea049d5eU, 0x355d018cU, 0x7473fa87U, 0x412efb0bU,
510     0x1d5ab367U, 0xd25292dbU, 0x5633e910U, 0x47136dd6U,
511     0x618c9ad7U, 0x0c7a37a1U, 0x148e59f8U, 0x3c89eb13U,
512     0x27eecea9U, 0xc935b761U, 0xe5ede11cU, 0xb13c7a47U,
513     0xdf599cd2U, 0x733f55f2U, 0xce791814U, 0x37bf73c7U,
514     0xcdea53f7U, 0xaa5b5ffdU, 0x6f14df3dU, 0xdb867844U,
515     0xf381caafU, 0xc43eb968U, 0x342c3824U, 0x405fc2a3U,
516     0xc372161dU, 0x250cbce2U, 0x498b283cU, 0x9541ff0dU,
517     0x017139a8U, 0xb3de080cU, 0xe49cd8b4U, 0xc1906456U,
518     0x84617bcbU, 0xb670d532U, 0x5c74486cU, 0x5742d0b8U,
519 };
520 static const u32 Td3[256] = {
521     0xf4a75051U, 0x4165537eU, 0x17a4c31aU, 0x275e963aU,
522     0xab6bcb3bU, 0x9d45f11fU, 0xfa58abacU, 0xe303934bU,
523     0x30fa5520U, 0x766df6adU, 0xcc769188U, 0x024c25f5U,
524     0xe5d7fc4fU, 0x2acbd7c5U, 0x35448026U, 0x62a38fb5U,
525     0xb15a49deU, 0xba1b6725U, 0xea0e9845U, 0xfec0e15dU,
526     0x2f7502c3U, 0x4cf01281U, 0x4697a38dU, 0xd3f9c66bU,
527     0x8f5fe703U, 0x929c9515U, 0x6d7aebbfU, 0x5259da95U,
528     0xbe832dd4U, 0x7421d358U, 0xe0692949U, 0xc9c8448eU,
529     0xc2896a75U, 0x8e7978f4U, 0x583e6b99U, 0xb971dd27U,
530     0xe14fb6beU, 0x88ad17f0U, 0x20ac66c9U, 0xce3ab47dU,
531     0xdf4a1863U, 0x1a3182e5U, 0x51336097U, 0x537f4562U,
532     0x6477e0b1U, 0x6bae84bbU, 0x81a01cfeU, 0x082b94f9U,
533     0x48685870U, 0x45fd198fU, 0xde6c8794U, 0x7bf8b752U,
534     0x73d323abU, 0x4b02e272U, 0x1f8f57e3U, 0x55ab2a66U,
535     0xeb2807b2U, 0xb5c2032fU, 0xc57b9a86U, 0x3708a5d3U,
536     0x2887f230U, 0xbfa5b223U, 0x036aba02U, 0x16825cedU,
537     0xcf1c2b8aU, 0x79b492a7U, 0x07f2f0f3U, 0x69e2a14eU,
538     0xdaf4cd65U, 0x05bed506U, 0x34621fd1U, 0xa6fe8ac4U,
539     0x2e539d34U, 0xf355a0a2U, 0x8ae13205U, 0xf6eb75a4U,
540     0x83ec390bU, 0x60efaa40U, 0x719f065eU, 0x6e1051bdU,
541     0x218af93eU, 0xdd063d96U, 0x3e05aeddU, 0xe6bd464dU,
542     0x548db591U, 0xc45d0571U, 0x06d46f04U, 0x5015ff60U,
543     0x98fb2419U, 0xbde997d6U, 0x4043cc89U, 0xd99e7767U,
544     0xe842bdb0U, 0x898b8807U, 0x195b38e7U, 0xc8eedb79U,
545     0x7c0a47a1U, 0x420fe97cU, 0x841ec9f8U, 0x00000000U,
546     0x80868309U, 0x2bed4832U, 0x1170ac1eU, 0x5a724e6cU,
547     0x0efffbfdU, 0x8538560fU, 0xaed51e3dU, 0x2d392736U,
548     0x0fd9640aU, 0x5ca62168U, 0x5b54d19bU, 0x362e3a24U,
549     0x0a67b10cU, 0x57e70f93U, 0xee96d2b4U, 0x9b919e1bU,
550     0xc0c54f80U, 0xdc20a261U, 0x774b695aU, 0x121a161cU,
551     0x93ba0ae2U, 0xa02ae5c0U, 0x22e0433cU, 0x1b171d12U,
552     0x090d0b0eU, 0x8bc7adf2U, 0xb6a8b92dU, 0x1ea9c814U,
553     0xf1198557U, 0x75074cafU, 0x99ddbbeeU, 0x7f60fda3U,
554     0x01269ff7U, 0x72f5bc5cU, 0x663bc544U, 0xfb7e345bU,
555     0x4329768bU, 0x23c6dccbU, 0xedfc68b6U, 0xe4f163b8U,
556     0x31dccad7U, 0x63851042U, 0x97224013U, 0xc6112084U,
557     0x4a247d85U, 0xbb3df8d2U, 0xf93211aeU, 0x29a16dc7U,
558     0x9e2f4b1dU, 0xb230f3dcU, 0x8652ec0dU, 0xc1e3d077U,
559     0xb3166c2bU, 0x70b999a9U, 0x9448fa11U, 0xe9642247U,
560     0xfc8cc4a8U, 0xf03f1aa0U, 0x7d2cd856U, 0x3390ef22U,
561     0x494ec787U, 0x38d1c1d9U, 0xcaa2fe8cU, 0xd40b3698U,
562     0xf581cfa6U, 0x7ade28a5U, 0xb78e26daU, 0xadbfa43fU,
563     0x3a9de42cU, 0x78920d50U, 0x5fcc9b6aU, 0x7e466254U,
564     0x8d13c2f6U, 0xd8b8e890U, 0x39f75e2eU, 0xc3aff582U,
565     0x5d80be9fU, 0xd0937c69U, 0xd52da96fU, 0x2512b3cfU,
566     0xac993bc8U, 0x187da710U, 0x9c636ee8U, 0x3bbb7bdbU,
567     0x267809cdU, 0x5918f46eU, 0x9ab701ecU, 0x4f9aa883U,
568     0x956e65e6U, 0xffe67eaaU, 0xbccf0821U, 0x15e8e6efU,
569     0xe79bd9baU, 0x6f36ce4aU, 0x9f09d4eaU, 0xb07cd629U,
570     0xa4b2af31U, 0x3f23312aU, 0xa59430c6U, 0xa266c035U,
571     0x4ebc3774U, 0x82caa6fcU, 0x90d0b0e0U, 0xa7d81533U,
572     0x04984af1U, 0xecdaf741U, 0xcd500e7fU, 0x91f62f17U,
573     0x4dd68d76U, 0xefb04d43U, 0xaa4d54ccU, 0x9604dfe4U,
574     0xd1b5e39eU, 0x6a881b4cU, 0x2c1fb8c1U, 0x65517f46U,
575     0x5eea049dU, 0x8c355d01U, 0x877473faU, 0x0b412efbU,
576     0x671d5ab3U, 0xdbd25292U, 0x105633e9U, 0xd647136dU,
577     0xd7618c9aU, 0xa10c7a37U, 0xf8148e59U, 0x133c89ebU,
578     0xa927eeceU, 0x61c935b7U, 0x1ce5ede1U, 0x47b13c7aU,
579     0xd2df599cU, 0xf2733f55U, 0x14ce7918U, 0xc737bf73U,
580     0xf7cdea53U, 0xfdaa5b5fU, 0x3d6f14dfU, 0x44db8678U,
581     0xaff381caU, 0x68c43eb9U, 0x24342c38U, 0xa3405fc2U,
582     0x1dc37216U, 0xe2250cbcU, 0x3c498b28U, 0x0d9541ffU,
583     0xa8017139U, 0x0cb3de08U, 0xb4e49cd8U, 0x56c19064U,
584     0xcb84617bU, 0x32b670d5U, 0x6c5c7448U, 0xb85742d0U,
585 };
586 static const u8 Td4[256] = {
587     0x52U, 0x09U, 0x6aU, 0xd5U, 0x30U, 0x36U, 0xa5U, 0x38U,
588     0xbfU, 0x40U, 0xa3U, 0x9eU, 0x81U, 0xf3U, 0xd7U, 0xfbU,
589     0x7cU, 0xe3U, 0x39U, 0x82U, 0x9bU, 0x2fU, 0xffU, 0x87U,
590     0x34U, 0x8eU, 0x43U, 0x44U, 0xc4U, 0xdeU, 0xe9U, 0xcbU,
591     0x54U, 0x7bU, 0x94U, 0x32U, 0xa6U, 0xc2U, 0x23U, 0x3dU,
592     0xeeU, 0x4cU, 0x95U, 0x0bU, 0x42U, 0xfaU, 0xc3U, 0x4eU,
593     0x08U, 0x2eU, 0xa1U, 0x66U, 0x28U, 0xd9U, 0x24U, 0xb2U,
594     0x76U, 0x5bU, 0xa2U, 0x49U, 0x6dU, 0x8bU, 0xd1U, 0x25U,
595     0x72U, 0xf8U, 0xf6U, 0x64U, 0x86U, 0x68U, 0x98U, 0x16U,
596     0xd4U, 0xa4U, 0x5cU, 0xccU, 0x5dU, 0x65U, 0xb6U, 0x92U,
597     0x6cU, 0x70U, 0x48U, 0x50U, 0xfdU, 0xedU, 0xb9U, 0xdaU,
598     0x5eU, 0x15U, 0x46U, 0x57U, 0xa7U, 0x8dU, 0x9dU, 0x84U,
599     0x90U, 0xd8U, 0xabU, 0x00U, 0x8cU, 0xbcU, 0xd3U, 0x0aU,
600     0xf7U, 0xe4U, 0x58U, 0x05U, 0xb8U, 0xb3U, 0x45U, 0x06U,
601     0xd0U, 0x2cU, 0x1eU, 0x8fU, 0xcaU, 0x3fU, 0x0fU, 0x02U,
602     0xc1U, 0xafU, 0xbdU, 0x03U, 0x01U, 0x13U, 0x8aU, 0x6bU,
603     0x3aU, 0x91U, 0x11U, 0x41U, 0x4fU, 0x67U, 0xdcU, 0xeaU,
604     0x97U, 0xf2U, 0xcfU, 0xceU, 0xf0U, 0xb4U, 0xe6U, 0x73U,
605     0x96U, 0xacU, 0x74U, 0x22U, 0xe7U, 0xadU, 0x35U, 0x85U,
606     0xe2U, 0xf9U, 0x37U, 0xe8U, 0x1cU, 0x75U, 0xdfU, 0x6eU,
607     0x47U, 0xf1U, 0x1aU, 0x71U, 0x1dU, 0x29U, 0xc5U, 0x89U,
608     0x6fU, 0xb7U, 0x62U, 0x0eU, 0xaaU, 0x18U, 0xbeU, 0x1bU,
609     0xfcU, 0x56U, 0x3eU, 0x4bU, 0xc6U, 0xd2U, 0x79U, 0x20U,
610     0x9aU, 0xdbU, 0xc0U, 0xfeU, 0x78U, 0xcdU, 0x5aU, 0xf4U,
611     0x1fU, 0xddU, 0xa8U, 0x33U, 0x88U, 0x07U, 0xc7U, 0x31U,
612     0xb1U, 0x12U, 0x10U, 0x59U, 0x27U, 0x80U, 0xecU, 0x5fU,
613     0x60U, 0x51U, 0x7fU, 0xa9U, 0x19U, 0xb5U, 0x4aU, 0x0dU,
614     0x2dU, 0xe5U, 0x7aU, 0x9fU, 0x93U, 0xc9U, 0x9cU, 0xefU,
615     0xa0U, 0xe0U, 0x3bU, 0x4dU, 0xaeU, 0x2aU, 0xf5U, 0xb0U,
616     0xc8U, 0xebU, 0xbbU, 0x3cU, 0x83U, 0x53U, 0x99U, 0x61U,
617     0x17U, 0x2bU, 0x04U, 0x7eU, 0xbaU, 0x77U, 0xd6U, 0x26U,
618     0xe1U, 0x69U, 0x14U, 0x63U, 0x55U, 0x21U, 0x0cU, 0x7dU,
619 };
620 static const u32 rcon[] = {
621     0x01000000, 0x02000000, 0x04000000, 0x08000000,
622     0x10000000, 0x20000000, 0x40000000, 0x80000000,
623     0x1B000000, 0x36000000, /* for 128-bit blocks, Rijndael never uses more than 10 rcon values */
624 };
625
626 /**
627  * Expand the cipher key into the encryption key schedule.
628  */
629 int AES_set_encrypt_key(const unsigned char *userKey, const int bits,
630                         AES_KEY *key)
631 {
632
633     u32 *rk;
634     int i = 0;
635     u32 temp;
636
637     if (!userKey || !key)
638         return -1;
639     if (bits != 128 && bits != 192 && bits != 256)
640         return -2;
641
642     rk = key->rd_key;
643
644     if (bits==128)
645         key->rounds = 10;
646     else if (bits==192)
647         key->rounds = 12;
648     else
649         key->rounds = 14;
650
651     rk[0] = GETU32(userKey     );
652     rk[1] = GETU32(userKey +  4);
653     rk[2] = GETU32(userKey +  8);
654     rk[3] = GETU32(userKey + 12);
655     if (bits == 128) {
656         while (1) {
657             temp  = rk[3];
658             rk[4] = rk[0] ^
659                 (Te2[(temp >> 16) & 0xff] & 0xff000000) ^
660                 (Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
661                 (Te0[(temp      ) & 0xff] & 0x0000ff00) ^
662                 (Te1[(temp >> 24)       ] & 0x000000ff) ^
663                 rcon[i];
664             rk[5] = rk[1] ^ rk[4];
665             rk[6] = rk[2] ^ rk[5];
666             rk[7] = rk[3] ^ rk[6];
667             if (++i == 10) {
668                 return 0;
669             }
670             rk += 4;
671         }
672     }
673     rk[4] = GETU32(userKey + 16);
674     rk[5] = GETU32(userKey + 20);
675     if (bits == 192) {
676         while (1) {
677             temp = rk[ 5];
678             rk[ 6] = rk[ 0] ^
679                 (Te2[(temp >> 16) & 0xff] & 0xff000000) ^
680                 (Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
681                 (Te0[(temp      ) & 0xff] & 0x0000ff00) ^
682                 (Te1[(temp >> 24)       ] & 0x000000ff) ^
683                 rcon[i];
684             rk[ 7] = rk[ 1] ^ rk[ 6];
685             rk[ 8] = rk[ 2] ^ rk[ 7];
686             rk[ 9] = rk[ 3] ^ rk[ 8];
687             if (++i == 8) {
688                 return 0;
689             }
690             rk[10] = rk[ 4] ^ rk[ 9];
691             rk[11] = rk[ 5] ^ rk[10];
692             rk += 6;
693         }
694     }
695     rk[6] = GETU32(userKey + 24);
696     rk[7] = GETU32(userKey + 28);
697     if (bits == 256) {
698         while (1) {
699             temp = rk[ 7];
700             rk[ 8] = rk[ 0] ^
701                 (Te2[(temp >> 16) & 0xff] & 0xff000000) ^
702                 (Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
703                 (Te0[(temp      ) & 0xff] & 0x0000ff00) ^
704                 (Te1[(temp >> 24)       ] & 0x000000ff) ^
705                 rcon[i];
706             rk[ 9] = rk[ 1] ^ rk[ 8];
707             rk[10] = rk[ 2] ^ rk[ 9];
708             rk[11] = rk[ 3] ^ rk[10];
709             if (++i == 7) {
710                 return 0;
711             }
712             temp = rk[11];
713             rk[12] = rk[ 4] ^
714                 (Te2[(temp >> 24)       ] & 0xff000000) ^
715                 (Te3[(temp >> 16) & 0xff] & 0x00ff0000) ^
716                 (Te0[(temp >>  8) & 0xff] & 0x0000ff00) ^
717                 (Te1[(temp      ) & 0xff] & 0x000000ff);
718             rk[13] = rk[ 5] ^ rk[12];
719             rk[14] = rk[ 6] ^ rk[13];
720             rk[15] = rk[ 7] ^ rk[14];
721
722             rk += 8;
723             }
724     }
725     return 0;
726 }
727
728 /**
729  * Expand the cipher key into the decryption key schedule.
730  */
731 int AES_set_decrypt_key(const unsigned char *userKey, const int bits,
732                         AES_KEY *key)
733 {
734
735     u32 *rk;
736     int i, j, status;
737     u32 temp;
738
739     /* first, start with an encryption schedule */
740     status = AES_set_encrypt_key(userKey, bits, key);
741     if (status < 0)
742         return status;
743
744     rk = key->rd_key;
745
746     /* invert the order of the round keys: */
747     for (i = 0, j = 4*(key->rounds); i < j; i += 4, j -= 4) {
748         temp = rk[i    ]; rk[i    ] = rk[j    ]; rk[j    ] = temp;
749         temp = rk[i + 1]; rk[i + 1] = rk[j + 1]; rk[j + 1] = temp;
750         temp = rk[i + 2]; rk[i + 2] = rk[j + 2]; rk[j + 2] = temp;
751         temp = rk[i + 3]; rk[i + 3] = rk[j + 3]; rk[j + 3] = temp;
752     }
753     /* apply the inverse MixColumn transform to all round keys but the first and the last: */
754     for (i = 1; i < (key->rounds); i++) {
755         rk += 4;
756         rk[0] =
757             Td0[Te1[(rk[0] >> 24)       ] & 0xff] ^
758             Td1[Te1[(rk[0] >> 16) & 0xff] & 0xff] ^
759             Td2[Te1[(rk[0] >>  8) & 0xff] & 0xff] ^
760             Td3[Te1[(rk[0]      ) & 0xff] & 0xff];
761         rk[1] =
762             Td0[Te1[(rk[1] >> 24)       ] & 0xff] ^
763             Td1[Te1[(rk[1] >> 16) & 0xff] & 0xff] ^
764             Td2[Te1[(rk[1] >>  8) & 0xff] & 0xff] ^
765             Td3[Te1[(rk[1]      ) & 0xff] & 0xff];
766         rk[2] =
767             Td0[Te1[(rk[2] >> 24)       ] & 0xff] ^
768             Td1[Te1[(rk[2] >> 16) & 0xff] & 0xff] ^
769             Td2[Te1[(rk[2] >>  8) & 0xff] & 0xff] ^
770             Td3[Te1[(rk[2]      ) & 0xff] & 0xff];
771         rk[3] =
772             Td0[Te1[(rk[3] >> 24)       ] & 0xff] ^
773             Td1[Te1[(rk[3] >> 16) & 0xff] & 0xff] ^
774             Td2[Te1[(rk[3] >>  8) & 0xff] & 0xff] ^
775             Td3[Te1[(rk[3]      ) & 0xff] & 0xff];
776     }
777     return 0;
778 }
779
780 /*
781  * Encrypt a single block
782  * in and out can overlap
783  */
784 void AES_encrypt(const unsigned char *in, unsigned char *out,
785                  const AES_KEY *key) {
786
787     const u32 *rk;
788     u32 s0, s1, s2, s3, t0, t1, t2, t3;
789 #ifndef FULL_UNROLL
790     int r;
791 #endif /* ?FULL_UNROLL */
792
793     assert(in && out && key);
794     rk = key->rd_key;
795
796     /*
797      * map byte array block to cipher state
798      * and add initial round key:
799      */
800     s0 = GETU32(in     ) ^ rk[0];
801     s1 = GETU32(in +  4) ^ rk[1];
802     s2 = GETU32(in +  8) ^ rk[2];
803     s3 = GETU32(in + 12) ^ rk[3];
804 #ifdef FULL_UNROLL
805     /* round 1: */
806     t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[ 4];
807     t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[ 5];
808     t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[ 6];
809     t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[ 7];
810     /* round 2: */
811     s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[ 8];
812     s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[ 9];
813     s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[10];
814     s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[11];
815     /* round 3: */
816     t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[12];
817     t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[13];
818     t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[14];
819     t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[15];
820     /* round 4: */
821     s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[16];
822     s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[17];
823     s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[18];
824     s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[19];
825     /* round 5: */
826     t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[20];
827     t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[21];
828     t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[22];
829     t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[23];
830     /* round 6: */
831     s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[24];
832     s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[25];
833     s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[26];
834     s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[27];
835     /* round 7: */
836     t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[28];
837     t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[29];
838     t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[30];
839     t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[31];
840     /* round 8: */
841     s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[32];
842     s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[33];
843     s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[34];
844     s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[35];
845     /* round 9: */
846     t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[36];
847     t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[37];
848     t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[38];
849     t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[39];
850     if (key->rounds > 10) {
851         /* round 10: */
852         s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[40];
853         s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[41];
854         s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[42];
855         s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[43];
856         /* round 11: */
857         t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[44];
858         t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[45];
859         t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[46];
860         t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[47];
861         if (key->rounds > 12) {
862             /* round 12: */
863             s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[48];
864             s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[49];
865             s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[50];
866             s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[51];
867             /* round 13: */
868             t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[52];
869             t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[53];
870             t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[54];
871             t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[55];
872         }
873     }
874     rk += key->rounds << 2;
875 #else  /* !FULL_UNROLL */
876     /*
877      * Nr - 1 full rounds:
878      */
879     r = key->rounds >> 1;
880     for (;;) {
881         t0 =
882             Te0[(s0 >> 24)       ] ^
883             Te1[(s1 >> 16) & 0xff] ^
884             Te2[(s2 >>  8) & 0xff] ^
885             Te3[(s3      ) & 0xff] ^
886             rk[4];
887         t1 =
888             Te0[(s1 >> 24)       ] ^
889             Te1[(s2 >> 16) & 0xff] ^
890             Te2[(s3 >>  8) & 0xff] ^
891             Te3[(s0      ) & 0xff] ^
892             rk[5];
893         t2 =
894             Te0[(s2 >> 24)       ] ^
895             Te1[(s3 >> 16) & 0xff] ^
896             Te2[(s0 >>  8) & 0xff] ^
897             Te3[(s1      ) & 0xff] ^
898             rk[6];
899         t3 =
900             Te0[(s3 >> 24)       ] ^
901             Te1[(s0 >> 16) & 0xff] ^
902             Te2[(s1 >>  8) & 0xff] ^
903             Te3[(s2      ) & 0xff] ^
904             rk[7];
905
906         rk += 8;
907         if (--r == 0) {
908             break;
909         }
910
911         s0 =
912             Te0[(t0 >> 24)       ] ^
913             Te1[(t1 >> 16) & 0xff] ^
914             Te2[(t2 >>  8) & 0xff] ^
915             Te3[(t3      ) & 0xff] ^
916             rk[0];
917         s1 =
918             Te0[(t1 >> 24)       ] ^
919             Te1[(t2 >> 16) & 0xff] ^
920             Te2[(t3 >>  8) & 0xff] ^
921             Te3[(t0      ) & 0xff] ^
922             rk[1];
923         s2 =
924             Te0[(t2 >> 24)       ] ^
925             Te1[(t3 >> 16) & 0xff] ^
926             Te2[(t0 >>  8) & 0xff] ^
927             Te3[(t1      ) & 0xff] ^
928             rk[2];
929         s3 =
930             Te0[(t3 >> 24)       ] ^
931             Te1[(t0 >> 16) & 0xff] ^
932             Te2[(t1 >>  8) & 0xff] ^
933             Te3[(t2      ) & 0xff] ^
934             rk[3];
935     }
936 #endif /* ?FULL_UNROLL */
937     /*
938      * apply last round and
939      * map cipher state to byte array block:
940      */
941     s0 =
942         (Te2[(t0 >> 24)       ] & 0xff000000) ^
943         (Te3[(t1 >> 16) & 0xff] & 0x00ff0000) ^
944         (Te0[(t2 >>  8) & 0xff] & 0x0000ff00) ^
945         (Te1[(t3      ) & 0xff] & 0x000000ff) ^
946         rk[0];
947     PUTU32(out     , s0);
948     s1 =
949         (Te2[(t1 >> 24)       ] & 0xff000000) ^
950         (Te3[(t2 >> 16) & 0xff] & 0x00ff0000) ^
951         (Te0[(t3 >>  8) & 0xff] & 0x0000ff00) ^
952         (Te1[(t0      ) & 0xff] & 0x000000ff) ^
953         rk[1];
954     PUTU32(out +  4, s1);
955     s2 =
956         (Te2[(t2 >> 24)       ] & 0xff000000) ^
957         (Te3[(t3 >> 16) & 0xff] & 0x00ff0000) ^
958         (Te0[(t0 >>  8) & 0xff] & 0x0000ff00) ^
959         (Te1[(t1      ) & 0xff] & 0x000000ff) ^
960         rk[2];
961     PUTU32(out +  8, s2);
962     s3 =
963         (Te2[(t3 >> 24)       ] & 0xff000000) ^
964         (Te3[(t0 >> 16) & 0xff] & 0x00ff0000) ^
965         (Te0[(t1 >>  8) & 0xff] & 0x0000ff00) ^
966         (Te1[(t2      ) & 0xff] & 0x000000ff) ^
967         rk[3];
968     PUTU32(out + 12, s3);
969 }
970
971 /*
972  * Decrypt a single block
973  * in and out can overlap
974  */
975 void AES_decrypt(const unsigned char *in, unsigned char *out,
976                  const AES_KEY *key)
977 {
978
979     const u32 *rk;
980     u32 s0, s1, s2, s3, t0, t1, t2, t3;
981 #ifndef FULL_UNROLL
982     int r;
983 #endif /* ?FULL_UNROLL */
984
985     assert(in && out && key);
986     rk = key->rd_key;
987
988     /*
989      * map byte array block to cipher state
990      * and add initial round key:
991      */
992     s0 = GETU32(in     ) ^ rk[0];
993     s1 = GETU32(in +  4) ^ rk[1];
994     s2 = GETU32(in +  8) ^ rk[2];
995     s3 = GETU32(in + 12) ^ rk[3];
996 #ifdef FULL_UNROLL
997     /* round 1: */
998     t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[ 4];
999     t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[ 5];
1000     t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[ 6];
1001     t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[ 7];
1002     /* round 2: */
1003     s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[ 8];
1004     s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[ 9];
1005     s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[10];
1006     s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[11];
1007     /* round 3: */
1008     t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[12];
1009     t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[13];
1010     t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[14];
1011     t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[15];
1012     /* round 4: */
1013     s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[16];
1014     s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[17];
1015     s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[18];
1016     s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[19];
1017     /* round 5: */
1018     t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[20];
1019     t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[21];
1020     t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[22];
1021     t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[23];
1022     /* round 6: */
1023     s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[24];
1024     s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[25];
1025     s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[26];
1026     s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[27];
1027     /* round 7: */
1028     t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[28];
1029     t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[29];
1030     t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[30];
1031     t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[31];
1032     /* round 8: */
1033     s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[32];
1034     s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[33];
1035     s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[34];
1036     s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[35];
1037     /* round 9: */
1038     t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[36];
1039     t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[37];
1040     t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[38];
1041     t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[39];
1042     if (key->rounds > 10) {
1043         /* round 10: */
1044         s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[40];
1045         s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[41];
1046         s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[42];
1047         s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[43];
1048         /* round 11: */
1049         t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[44];
1050         t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[45];
1051         t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[46];
1052         t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[47];
1053         if (key->rounds > 12) {
1054             /* round 12: */
1055             s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[48];
1056             s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[49];
1057             s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[50];
1058             s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[51];
1059             /* round 13: */
1060             t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[52];
1061             t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[53];
1062             t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[54];
1063             t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[55];
1064         }
1065     }
1066     rk += key->rounds << 2;
1067 #else  /* !FULL_UNROLL */
1068     /*
1069      * Nr - 1 full rounds:
1070      */
1071     r = key->rounds >> 1;
1072     for (;;) {
1073         t0 =
1074             Td0[(s0 >> 24)       ] ^
1075             Td1[(s3 >> 16) & 0xff] ^
1076             Td2[(s2 >>  8) & 0xff] ^
1077             Td3[(s1      ) & 0xff] ^
1078             rk[4];
1079         t1 =
1080             Td0[(s1 >> 24)       ] ^
1081             Td1[(s0 >> 16) & 0xff] ^
1082             Td2[(s3 >>  8) & 0xff] ^
1083             Td3[(s2      ) & 0xff] ^
1084             rk[5];
1085         t2 =
1086             Td0[(s2 >> 24)       ] ^
1087             Td1[(s1 >> 16) & 0xff] ^
1088             Td2[(s0 >>  8) & 0xff] ^
1089             Td3[(s3      ) & 0xff] ^
1090             rk[6];
1091         t3 =
1092             Td0[(s3 >> 24)       ] ^
1093             Td1[(s2 >> 16) & 0xff] ^
1094             Td2[(s1 >>  8) & 0xff] ^
1095             Td3[(s0      ) & 0xff] ^
1096             rk[7];
1097
1098         rk += 8;
1099         if (--r == 0) {
1100             break;
1101         }
1102
1103         s0 =
1104             Td0[(t0 >> 24)       ] ^
1105             Td1[(t3 >> 16) & 0xff] ^
1106             Td2[(t2 >>  8) & 0xff] ^
1107             Td3[(t1      ) & 0xff] ^
1108             rk[0];
1109         s1 =
1110             Td0[(t1 >> 24)       ] ^
1111             Td1[(t0 >> 16) & 0xff] ^
1112             Td2[(t3 >>  8) & 0xff] ^
1113             Td3[(t2      ) & 0xff] ^
1114             rk[1];
1115         s2 =
1116             Td0[(t2 >> 24)       ] ^
1117             Td1[(t1 >> 16) & 0xff] ^
1118             Td2[(t0 >>  8) & 0xff] ^
1119             Td3[(t3      ) & 0xff] ^
1120             rk[2];
1121         s3 =
1122             Td0[(t3 >> 24)       ] ^
1123             Td1[(t2 >> 16) & 0xff] ^
1124             Td2[(t1 >>  8) & 0xff] ^
1125             Td3[(t0      ) & 0xff] ^
1126             rk[3];
1127     }
1128 #endif /* ?FULL_UNROLL */
1129     /*
1130      * apply last round and
1131      * map cipher state to byte array block:
1132      */
1133     s0 =
1134         ((u32)Td4[(t0 >> 24)       ] << 24) ^
1135         ((u32)Td4[(t3 >> 16) & 0xff] << 16) ^
1136         ((u32)Td4[(t2 >>  8) & 0xff] <<  8) ^
1137         ((u32)Td4[(t1      ) & 0xff])       ^
1138         rk[0];
1139     PUTU32(out     , s0);
1140     s1 =
1141         ((u32)Td4[(t1 >> 24)       ] << 24) ^
1142         ((u32)Td4[(t0 >> 16) & 0xff] << 16) ^
1143         ((u32)Td4[(t3 >>  8) & 0xff] <<  8) ^
1144         ((u32)Td4[(t2      ) & 0xff])       ^
1145         rk[1];
1146     PUTU32(out +  4, s1);
1147     s2 =
1148         ((u32)Td4[(t2 >> 24)       ] << 24) ^
1149         ((u32)Td4[(t1 >> 16) & 0xff] << 16) ^
1150         ((u32)Td4[(t0 >>  8) & 0xff] <<  8) ^
1151         ((u32)Td4[(t3      ) & 0xff])       ^
1152         rk[2];
1153     PUTU32(out +  8, s2);
1154     s3 =
1155         ((u32)Td4[(t3 >> 24)       ] << 24) ^
1156         ((u32)Td4[(t2 >> 16) & 0xff] << 16) ^
1157         ((u32)Td4[(t1 >>  8) & 0xff] <<  8) ^
1158         ((u32)Td4[(t0      ) & 0xff])       ^
1159         rk[3];
1160     PUTU32(out + 12, s3);
1161 }
1162
1163 #else /* AES_ASM */
1164
1165 static const u8 Te4[256] = {
1166     0x63U, 0x7cU, 0x77U, 0x7bU, 0xf2U, 0x6bU, 0x6fU, 0xc5U,
1167     0x30U, 0x01U, 0x67U, 0x2bU, 0xfeU, 0xd7U, 0xabU, 0x76U,
1168     0xcaU, 0x82U, 0xc9U, 0x7dU, 0xfaU, 0x59U, 0x47U, 0xf0U,
1169     0xadU, 0xd4U, 0xa2U, 0xafU, 0x9cU, 0xa4U, 0x72U, 0xc0U,
1170     0xb7U, 0xfdU, 0x93U, 0x26U, 0x36U, 0x3fU, 0xf7U, 0xccU,
1171     0x34U, 0xa5U, 0xe5U, 0xf1U, 0x71U, 0xd8U, 0x31U, 0x15U,
1172     0x04U, 0xc7U, 0x23U, 0xc3U, 0x18U, 0x96U, 0x05U, 0x9aU,
1173     0x07U, 0x12U, 0x80U, 0xe2U, 0xebU, 0x27U, 0xb2U, 0x75U,
1174     0x09U, 0x83U, 0x2cU, 0x1aU, 0x1bU, 0x6eU, 0x5aU, 0xa0U,
1175     0x52U, 0x3bU, 0xd6U, 0xb3U, 0x29U, 0xe3U, 0x2fU, 0x84U,
1176     0x53U, 0xd1U, 0x00U, 0xedU, 0x20U, 0xfcU, 0xb1U, 0x5bU,
1177     0x6aU, 0xcbU, 0xbeU, 0x39U, 0x4aU, 0x4cU, 0x58U, 0xcfU,
1178     0xd0U, 0xefU, 0xaaU, 0xfbU, 0x43U, 0x4dU, 0x33U, 0x85U,
1179     0x45U, 0xf9U, 0x02U, 0x7fU, 0x50U, 0x3cU, 0x9fU, 0xa8U,
1180     0x51U, 0xa3U, 0x40U, 0x8fU, 0x92U, 0x9dU, 0x38U, 0xf5U,
1181     0xbcU, 0xb6U, 0xdaU, 0x21U, 0x10U, 0xffU, 0xf3U, 0xd2U,
1182     0xcdU, 0x0cU, 0x13U, 0xecU, 0x5fU, 0x97U, 0x44U, 0x17U,
1183     0xc4U, 0xa7U, 0x7eU, 0x3dU, 0x64U, 0x5dU, 0x19U, 0x73U,
1184     0x60U, 0x81U, 0x4fU, 0xdcU, 0x22U, 0x2aU, 0x90U, 0x88U,
1185     0x46U, 0xeeU, 0xb8U, 0x14U, 0xdeU, 0x5eU, 0x0bU, 0xdbU,
1186     0xe0U, 0x32U, 0x3aU, 0x0aU, 0x49U, 0x06U, 0x24U, 0x5cU,
1187     0xc2U, 0xd3U, 0xacU, 0x62U, 0x91U, 0x95U, 0xe4U, 0x79U,
1188     0xe7U, 0xc8U, 0x37U, 0x6dU, 0x8dU, 0xd5U, 0x4eU, 0xa9U,
1189     0x6cU, 0x56U, 0xf4U, 0xeaU, 0x65U, 0x7aU, 0xaeU, 0x08U,
1190     0xbaU, 0x78U, 0x25U, 0x2eU, 0x1cU, 0xa6U, 0xb4U, 0xc6U,
1191     0xe8U, 0xddU, 0x74U, 0x1fU, 0x4bU, 0xbdU, 0x8bU, 0x8aU,
1192     0x70U, 0x3eU, 0xb5U, 0x66U, 0x48U, 0x03U, 0xf6U, 0x0eU,
1193     0x61U, 0x35U, 0x57U, 0xb9U, 0x86U, 0xc1U, 0x1dU, 0x9eU,
1194     0xe1U, 0xf8U, 0x98U, 0x11U, 0x69U, 0xd9U, 0x8eU, 0x94U,
1195     0x9bU, 0x1eU, 0x87U, 0xe9U, 0xceU, 0x55U, 0x28U, 0xdfU,
1196     0x8cU, 0xa1U, 0x89U, 0x0dU, 0xbfU, 0xe6U, 0x42U, 0x68U,
1197     0x41U, 0x99U, 0x2dU, 0x0fU, 0xb0U, 0x54U, 0xbbU, 0x16U
1198 };
1199 static const u32 rcon[] = {
1200     0x01000000, 0x02000000, 0x04000000, 0x08000000,
1201     0x10000000, 0x20000000, 0x40000000, 0x80000000,
1202     0x1B000000, 0x36000000, /* for 128-bit blocks, Rijndael never uses more than 10 rcon values */
1203 };
1204
1205 /**
1206  * Expand the cipher key into the encryption key schedule.
1207  */
1208 int AES_set_encrypt_key(const unsigned char *userKey, const int bits,
1209                         AES_KEY *key)
1210 {
1211     u32 *rk;
1212     int i = 0;
1213     u32 temp;
1214
1215     if (!userKey || !key)
1216         return -1;
1217     if (bits != 128 && bits != 192 && bits != 256)
1218         return -2;
1219
1220     rk = key->rd_key;
1221
1222     if (bits==128)
1223         key->rounds = 10;
1224     else if (bits==192)
1225         key->rounds = 12;
1226     else
1227         key->rounds = 14;
1228
1229     rk[0] = GETU32(userKey     );
1230     rk[1] = GETU32(userKey +  4);
1231     rk[2] = GETU32(userKey +  8);
1232     rk[3] = GETU32(userKey + 12);
1233     if (bits == 128) {
1234         while (1) {
1235             temp  = rk[3];
1236             rk[4] = rk[0] ^
1237                 ((u32)Te4[(temp >> 16) & 0xff] << 24) ^
1238                 ((u32)Te4[(temp >>  8) & 0xff] << 16) ^
1239                 ((u32)Te4[(temp      ) & 0xff] << 8) ^
1240                 ((u32)Te4[(temp >> 24)       ]) ^
1241                 rcon[i];
1242             rk[5] = rk[1] ^ rk[4];
1243             rk[6] = rk[2] ^ rk[5];
1244             rk[7] = rk[3] ^ rk[6];
1245             if (++i == 10) {
1246                 return 0;
1247             }
1248             rk += 4;
1249         }
1250     }
1251     rk[4] = GETU32(userKey + 16);
1252     rk[5] = GETU32(userKey + 20);
1253     if (bits == 192) {
1254         while (1) {
1255             temp = rk[ 5];
1256             rk[ 6] = rk[ 0] ^
1257                 ((u32)Te4[(temp >> 16) & 0xff] << 24) ^
1258                 ((u32)Te4[(temp >>  8) & 0xff] << 16) ^
1259                 ((u32)Te4[(temp      ) & 0xff] << 8) ^
1260                 ((u32)Te4[(temp >> 24)       ]) ^
1261                 rcon[i];
1262             rk[ 7] = rk[ 1] ^ rk[ 6];
1263             rk[ 8] = rk[ 2] ^ rk[ 7];
1264             rk[ 9] = rk[ 3] ^ rk[ 8];
1265             if (++i == 8) {
1266                 return 0;
1267             }
1268             rk[10] = rk[ 4] ^ rk[ 9];
1269             rk[11] = rk[ 5] ^ rk[10];
1270             rk += 6;
1271         }
1272     }
1273     rk[6] = GETU32(userKey + 24);
1274     rk[7] = GETU32(userKey + 28);
1275     if (bits == 256) {
1276         while (1) {
1277             temp = rk[ 7];
1278             rk[ 8] = rk[ 0] ^
1279                 ((u32)Te4[(temp >> 16) & 0xff] << 24) ^
1280                 ((u32)Te4[(temp >>  8) & 0xff] << 16) ^
1281                 ((u32)Te4[(temp      ) & 0xff] << 8) ^
1282                 ((u32)Te4[(temp >> 24)       ]) ^
1283                 rcon[i];
1284             rk[ 9] = rk[ 1] ^ rk[ 8];
1285             rk[10] = rk[ 2] ^ rk[ 9];
1286             rk[11] = rk[ 3] ^ rk[10];
1287             if (++i == 7) {
1288                 return 0;
1289             }
1290             temp = rk[11];
1291             rk[12] = rk[ 4] ^
1292                 ((u32)Te4[(temp >> 24)       ] << 24) ^
1293                 ((u32)Te4[(temp >> 16) & 0xff] << 16) ^
1294                 ((u32)Te4[(temp >>  8) & 0xff] << 8) ^
1295                 ((u32)Te4[(temp      ) & 0xff]);
1296             rk[13] = rk[ 5] ^ rk[12];
1297             rk[14] = rk[ 6] ^ rk[13];
1298             rk[15] = rk[ 7] ^ rk[14];
1299
1300             rk += 8;
1301         }
1302     }
1303     return 0;
1304 }
1305
1306 /**
1307  * Expand the cipher key into the decryption key schedule.
1308  */
1309 int AES_set_decrypt_key(const unsigned char *userKey, const int bits,
1310                         AES_KEY *key)
1311 {
1312
1313     u32 *rk;
1314     int i, j, status;
1315     u32 temp;
1316
1317     /* first, start with an encryption schedule */
1318     status = AES_set_encrypt_key(userKey, bits, key);
1319     if (status < 0)
1320         return status;
1321
1322     rk = key->rd_key;
1323
1324     /* invert the order of the round keys: */
1325     for (i = 0, j = 4*(key->rounds); i < j; i += 4, j -= 4) {
1326         temp = rk[i    ]; rk[i    ] = rk[j    ]; rk[j    ] = temp;
1327         temp = rk[i + 1]; rk[i + 1] = rk[j + 1]; rk[j + 1] = temp;
1328         temp = rk[i + 2]; rk[i + 2] = rk[j + 2]; rk[j + 2] = temp;
1329         temp = rk[i + 3]; rk[i + 3] = rk[j + 3]; rk[j + 3] = temp;
1330     }
1331     /* apply the inverse MixColumn transform to all round keys but the first and the last: */
1332     for (i = 1; i < (key->rounds); i++) {
1333         rk += 4;
1334         for (j = 0; j < 4; j++) {
1335             u32 tp1, tp2, tp4, tp8, tp9, tpb, tpd, tpe, m;
1336
1337             tp1 = rk[j];
1338             m = tp1 & 0x80808080;
1339             tp2 = ((tp1 & 0x7f7f7f7f) << 1) ^
1340                 ((m - (m >> 7)) & 0x1b1b1b1b);
1341             m = tp2 & 0x80808080;
1342             tp4 = ((tp2 & 0x7f7f7f7f) << 1) ^
1343                 ((m - (m >> 7)) & 0x1b1b1b1b);
1344             m = tp4 & 0x80808080;
1345             tp8 = ((tp4 & 0x7f7f7f7f) << 1) ^
1346                 ((m - (m >> 7)) & 0x1b1b1b1b);
1347             tp9 = tp8 ^ tp1;
1348             tpb = tp9 ^ tp2;
1349             tpd = tp9 ^ tp4;
1350             tpe = tp8 ^ tp4 ^ tp2;
1351 #if defined(ROTATE)
1352             rk[j] = tpe ^ ROTATE(tpd,16) ^
1353                 ROTATE(tp9,24) ^ ROTATE(tpb,8);
1354 #else
1355             rk[j] = tpe ^ (tpd >> 16) ^ (tpd << 16) ^ 
1356                 (tp9 >> 8) ^ (tp9 << 24) ^
1357                 (tpb >> 24) ^ (tpb << 8);
1358 #endif
1359         }
1360     }
1361     return 0;
1362 }
1363
1364 #endif /* AES_ASM */