1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
use crate::core_arch::arm::{uint32x4_t, uint8x16_t};

#[allow(improper_ctypes)]
extern "C" {
    #[link_name = "llvm.aarch64.crypto.aese"]
    fn vaeseq_u8_(data: uint8x16_t, key: uint8x16_t) -> uint8x16_t;
    #[link_name = "llvm.aarch64.crypto.aesd"]
    fn vaesdq_u8_(data: uint8x16_t, key: uint8x16_t) -> uint8x16_t;
    #[link_name = "llvm.aarch64.crypto.aesmc"]
    fn vaesmcq_u8_(data: uint8x16_t) -> uint8x16_t;
    #[link_name = "llvm.aarch64.crypto.aesimc"]
    fn vaesimcq_u8_(data: uint8x16_t) -> uint8x16_t;

    #[link_name = "llvm.aarch64.crypto.sha1h"]
    fn vsha1h_u32_(hash_e: u32) -> u32;
    #[link_name = "llvm.aarch64.crypto.sha1su0"]
    fn vsha1su0q_u32_(w0_3: uint32x4_t, w4_7: uint32x4_t, w8_11: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha1su1"]
    fn vsha1su1q_u32_(tw0_3: uint32x4_t, w12_15: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha1c"]
    fn vsha1cq_u32_(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha1p"]
    fn vsha1pq_u32_(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha1m"]
    fn vsha1mq_u32_(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t;

    #[link_name = "llvm.aarch64.crypto.sha256h"]
    fn vsha256hq_u32_(hash_abcd: uint32x4_t, hash_efgh: uint32x4_t, wk: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha256h2"]
    fn vsha256h2q_u32_(hash_efgh: uint32x4_t, hash_abcd: uint32x4_t, wk: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha256su0"]
    fn vsha256su0q_u32_(w0_3: uint32x4_t, w4_7: uint32x4_t) -> uint32x4_t;
    #[link_name = "llvm.aarch64.crypto.sha256su1"]
    fn vsha256su1q_u32_(tw0_3: uint32x4_t, w8_11: uint32x4_t, w12_15: uint32x4_t) -> uint32x4_t;
}

#[cfg(test)]
use stdarch_test::assert_instr;

/// AES single round encryption.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(aese))]
pub unsafe fn vaeseq_u8(data: uint8x16_t, key: uint8x16_t) -> uint8x16_t {
    vaeseq_u8_(data, key)
}

/// AES single round decryption.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(aesd))]
pub unsafe fn vaesdq_u8(data: uint8x16_t, key: uint8x16_t) -> uint8x16_t {
    vaesdq_u8_(data, key)
}

/// AES mix columns.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(aesmc))]
pub unsafe fn vaesmcq_u8(data: uint8x16_t) -> uint8x16_t {
    vaesmcq_u8_(data)
}

/// AES inverse mix columns.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(aesimc))]
pub unsafe fn vaesimcq_u8(data: uint8x16_t) -> uint8x16_t {
    vaesimcq_u8_(data)
}

/// SHA1 fixed rotate.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1h))]
pub unsafe fn vsha1h_u32(hash_e: u32) -> u32 {
    vsha1h_u32_(hash_e)
}

/// SHA1 hash update accelerator, choose.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1c))]
pub unsafe fn vsha1cq_u32(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t {
    vsha1cq_u32_(hash_abcd, hash_e, wk)
}

/// SHA1 hash update accelerator, majority.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1m))]
pub unsafe fn vsha1mq_u32(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t {
    vsha1mq_u32_(hash_abcd, hash_e, wk)
}

/// SHA1 hash update accelerator, parity.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1p))]
pub unsafe fn vsha1pq_u32(hash_abcd: uint32x4_t, hash_e: u32, wk: uint32x4_t) -> uint32x4_t {
    vsha1pq_u32_(hash_abcd, hash_e, wk)
}

/// SHA1 schedule update accelerator, first part.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1su0))]
pub unsafe fn vsha1su0q_u32(w0_3: uint32x4_t, w4_7: uint32x4_t, w8_11: uint32x4_t) -> uint32x4_t {
    vsha1su0q_u32_(w0_3, w4_7, w8_11)
}

/// SHA1 schedule update accelerator, second part.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha1su1))]
pub unsafe fn vsha1su1q_u32(tw0_3: uint32x4_t, w12_15: uint32x4_t) -> uint32x4_t {
    vsha1su1q_u32_(tw0_3, w12_15)
}

/// SHA256 hash update accelerator.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha256h))]
pub unsafe fn vsha256hq_u32(
    hash_abcd: uint32x4_t,
    hash_efgh: uint32x4_t,
    wk: uint32x4_t,
) -> uint32x4_t {
    vsha256hq_u32_(hash_abcd, hash_efgh, wk)
}

/// SHA256 hash update accelerator, upper part.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha256h2))]
pub unsafe fn vsha256h2q_u32(
    hash_efgh: uint32x4_t,
    hash_abcd: uint32x4_t,
    wk: uint32x4_t,
) -> uint32x4_t {
    vsha256h2q_u32_(hash_efgh, hash_abcd, wk)
}

/// SHA256 schedule update accelerator, first part.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha256su0))]
pub unsafe fn vsha256su0q_u32(w0_3: uint32x4_t, w4_7: uint32x4_t) -> uint32x4_t {
    vsha256su0q_u32_(w0_3, w4_7)
}

/// SHA256 schedule update accelerator, second part.
#[inline]
#[target_feature(enable = "crypto")]
#[cfg_attr(test, assert_instr(sha256su1))]
pub unsafe fn vsha256su1q_u32(
    tw0_3: uint32x4_t,
    w8_11: uint32x4_t,
    w12_15: uint32x4_t,
) -> uint32x4_t {
    vsha256su1q_u32_(tw0_3, w8_11, w12_15)
}

#[cfg(test)]
mod tests {
    use crate::core_arch::{aarch64::*, simd::*};
    use std::mem;
    use stdarch_test::simd_test;

    #[simd_test(enable = "crypto")]
    unsafe fn test_vaeseq_u8() {
        let data = mem::transmute(u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8));
        let key = mem::transmute(u8x16::new(0, 1, 2, 3, 4, 5, 6, 7, 0, 1, 2, 3, 4, 5, 6, 7));
        let r: u8x16 = mem::transmute(vaeseq_u8(data, key));
        assert_eq!(
            r,
            u8x16::new(
                124, 123, 124, 118, 124, 123, 124, 197, 124, 123, 124, 118, 124, 123, 124, 197
            )
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vaesdq_u8() {
        let data = mem::transmute(u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8));
        let key = mem::transmute(u8x16::new(0, 1, 2, 3, 4, 5, 6, 7, 0, 1, 2, 3, 4, 5, 6, 7));
        let r: u8x16 = mem::transmute(vaesdq_u8(data, key));
        assert_eq!(
            r,
            u8x16::new(9, 213, 9, 251, 9, 213, 9, 56, 9, 213, 9, 251, 9, 213, 9, 56)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vaesmcq_u8() {
        let data = mem::transmute(u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8));
        let r: u8x16 = mem::transmute(vaesmcq_u8(data));
        assert_eq!(
            r,
            u8x16::new(3, 4, 9, 10, 15, 8, 21, 30, 3, 4, 9, 10, 15, 8, 21, 30)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vaesimcq_u8() {
        let data = mem::transmute(u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8));
        let r: u8x16 = mem::transmute(vaesimcq_u8(data));
        assert_eq!(
            r,
            u8x16::new(43, 60, 33, 50, 103, 80, 125, 70, 43, 60, 33, 50, 103, 80, 125, 70)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1h_u32() {
        assert_eq!(vsha1h_u32(0x1234), 0x048d);
        assert_eq!(vsha1h_u32(0x5678), 0x159e);
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1su0q_u32() {
        let r: u32x4 = mem::transmute(vsha1su0q_u32(
            mem::transmute(u32x4::new(0x1234_u32, 0x5678_u32, 0x9abc_u32, 0xdef0_u32)),
            mem::transmute(u32x4::new(0x1234_u32, 0x5678_u32, 0x9abc_u32, 0xdef0_u32)),
            mem::transmute(u32x4::new(0x1234_u32, 0x5678_u32, 0x9abc_u32, 0xdef0_u32)),
        ));
        assert_eq!(r, u32x4::new(0x9abc, 0xdef0, 0x1234, 0x5678));
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1su1q_u32() {
        let r: u32x4 = mem::transmute(vsha1su1q_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x00008898, 0x00019988, 0x00008898, 0x0000acd0)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1cq_u32() {
        let r: u32x4 = mem::transmute(vsha1cq_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            0x1234,
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x8a32cbd8, 0x0c518a96, 0x0018a081, 0x0000c168)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1pq_u32() {
        let r: u32x4 = mem::transmute(vsha1pq_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            0x1234,
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x469f0ba3, 0x0a326147, 0x80145d7f, 0x00009f47)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha1mq_u32() {
        let r: u32x4 = mem::transmute(vsha1mq_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            0x1234,
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0xaa39693b, 0x0d51bf84, 0x001aa109, 0x0000d278)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha256hq_u32() {
        let r: u32x4 = mem::transmute(vsha256hq_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x05e9aaa8, 0xec5f4c02, 0x20a1ea61, 0x28738cef)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha256h2q_u32() {
        let r: u32x4 = mem::transmute(vsha256h2q_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x3745362e, 0x2fb51d00, 0xbd4c529b, 0x968b8516)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha256su0q_u32() {
        let r: u32x4 = mem::transmute(vsha256su0q_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0xe59e1c97, 0x5eaf68da, 0xd7bcb51f, 0x6c8de152)
        );
    }

    #[simd_test(enable = "crypto")]
    unsafe fn test_vsha256su1q_u32() {
        let r: u32x4 = mem::transmute(vsha256su1q_u32(
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
            mem::transmute(u32x4::new(0x1234, 0x5678, 0x9abc, 0xdef0)),
        ));
        assert_eq!(
            r,
            u32x4::new(0x5e09e8d2, 0x74a6f16b, 0xc966606b, 0xa686ee9f)
        );
    }
}