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
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
use crate::{
    field::Field,
    group::{globalz_set_table_gej, set_table_gej_var, Affine, AffineStorage, Jacobian, AFFINE_G},
    scalar::Scalar,
};
use alloc::{
    alloc::{alloc, Layout},
    boxed::Box,
    vec,
    vec::Vec,
};
use subtle::Choice;

pub const WINDOW_A: usize = 5;
pub const WINDOW_G: usize = 16;
pub const ECMULT_TABLE_SIZE_A: usize = 1 << (WINDOW_A - 2);
pub const ECMULT_TABLE_SIZE_G: usize = 1 << (WINDOW_G - 2);
pub const WNAF_BITS: usize = 256;

fn odd_multiples_table_storage_var(pre: &mut [AffineStorage], a: &Jacobian) {
    let mut prej: Vec<Jacobian> = Vec::with_capacity(pre.len());
    for _ in 0..pre.len() {
        prej.push(Jacobian::default());
    }
    let mut prea: Vec<Affine> = Vec::with_capacity(pre.len());
    for _ in 0..pre.len() {
        prea.push(Affine::default());
    }
    let mut zr: Vec<Field> = Vec::with_capacity(pre.len());
    for _ in 0..pre.len() {
        zr.push(Field::default());
    }

    odd_multiples_table(&mut prej, &mut zr, a);
    set_table_gej_var(&mut prea, &prej, &zr);

    for i in 0..pre.len() {
        pre[i] = prea[i].into();
    }
}

/// Context for accelerating the computation of a*P + b*G.
pub struct ECMultContext {
    pre_g: [AffineStorage; ECMULT_TABLE_SIZE_G],
}

impl ECMultContext {
    /// Create a new `ECMultContext` from raw values.
    ///
    /// # Safety
    /// The function is unsafe because incorrect value of `pre_g` can lead to
    /// crypto logic failure. You most likely do not want to use this function,
    /// but `ECMultContext::new_boxed`.
    pub const unsafe fn new_from_raw(pre_g: [AffineStorage; ECMULT_TABLE_SIZE_G]) -> Self {
        Self { pre_g }
    }

    /// Inspect raw values of `ECMultContext`.
    pub fn inspect_raw(&self) -> &[AffineStorage; ECMULT_TABLE_SIZE_G] {
        &self.pre_g
    }

    /// Generate a new `ECMultContext` on the heap. Note that this function is expensive.
    pub fn new_boxed() -> Box<Self> {
        // This unsafe block allocates a new, unitialized `ECMultContext` and
        // then fills in the value. This is to avoid allocating it on stack
        // because the struct is big. All values in `ECMultContext` are manually
        // initialized after allocation.
        let mut this = unsafe {
            let ptr = alloc(Layout::new::<ECMultContext>()) as *mut ECMultContext;
            let mut this = Box::from_raw(ptr);

            for i in 0..ECMULT_TABLE_SIZE_G {
                this.pre_g[i] = AffineStorage::default();
            }

            this
        };

        let mut gj = Jacobian::default();
        gj.set_ge(&AFFINE_G);
        odd_multiples_table_storage_var(&mut this.pre_g, &gj);

        this
    }
}

/// Set a batch of group elements equal to the inputs given in jacobian
/// coordinates. Not constant time.
pub fn set_all_gej_var(a: &[Jacobian]) -> Vec<Affine> {
    let mut az: Vec<Field> = Vec::with_capacity(a.len());
    for point in a {
        if !point.is_infinity() {
            az.push(point.z);
        }
    }
    let azi: Vec<Field> = inv_all_var(&az);

    let mut ret = vec![Affine::default(); a.len()];

    let mut count = 0;
    for i in 0..a.len() {
        ret[i].infinity = a[i].infinity;
        if !a[i].is_infinity() {
            ret[i].set_gej_zinv(&a[i], &azi[count]);
            count += 1;
        }
    }
    ret
}

/// Calculate the (modular) inverses of a batch of field
/// elements. Requires the inputs' magnitudes to be at most 8. The
/// output magnitudes are 1 (but not guaranteed to be
/// normalized).
pub fn inv_all_var(fields: &[Field]) -> Vec<Field> {
    if fields.is_empty() {
        return Vec::new();
    }

    let mut ret = Vec::with_capacity(fields.len());
    ret.push(fields[0]);

    for i in 1..fields.len() {
        ret.push(Field::default());
        ret[i] = ret[i - 1] * fields[i];
    }

    let mut u = ret[fields.len() - 1].inv_var();

    for i in (1..fields.len()).rev() {
        let j = i;
        let i = i - 1;
        ret[j] = ret[i] * u;
        u *= fields[j];
    }

    ret[0] = u;
    ret
}

const GEN_BLIND: Scalar = Scalar([
    2217680822, 850875797, 1046150361, 1330484644, 4015777837, 2466086288, 2052467175, 2084507480,
]);
const GEN_INITIAL: Jacobian = Jacobian {
    x: Field::new_raw(
        586608, 43357028, 207667908, 262670128, 142222828, 38529388, 267186148, 45417712,
        115291924, 13447464,
    ),
    y: Field::new_raw(
        12696548, 208302564, 112025180, 191752716, 143238548, 145482948, 228906000, 69755164,
        243572800, 210897016,
    ),
    z: Field::new_raw(
        3685368, 75404844, 20246216, 5748944, 73206666, 107661790, 110806176, 73488774, 5707384,
        104448710,
    ),
    infinity: false,
};

/// Context for accelerating the computation of a*G.
pub struct ECMultGenContext {
    prec: [[AffineStorage; 16]; 64],
    blind: Scalar,
    initial: Jacobian,
}

impl ECMultGenContext {
    /// Create a new `ECMultGenContext` from raw values.
    ///
    /// # Safety
    /// The function is unsafe because incorrect value of `pre_g` can lead to
    /// crypto logic failure. You most likely do not want to use this function,
    /// but `ECMultGenContext::new_boxed`.
    pub const unsafe fn new_from_raw(prec: [[AffineStorage; 16]; 64]) -> Self {
        Self {
            prec,
            blind: GEN_BLIND,
            initial: GEN_INITIAL,
        }
    }

    /// Inspect `ECMultGenContext` values.
    pub fn inspect_raw(&self) -> &[[AffineStorage; 16]; 64] {
        &self.prec
    }

    /// Generate a new `ECMultGenContext` on the heap. Note that this function is expensive.
    pub fn new_boxed() -> Box<Self> {
        // This unsafe block allocates a new, unitialized `ECMultGenContext` and
        // then fills in the value. This is to avoid allocating it on stack
        // because the struct is big. All values in `ECMultGenContext` are
        // manually initialized after allocation.
        let mut this = unsafe {
            let ptr = alloc(Layout::new::<ECMultGenContext>()) as *mut ECMultGenContext;
            let mut this = Box::from_raw(ptr);

            for j in 0..64 {
                for i in 0..16 {
                    this.prec[j][i] = AffineStorage::default();
                }
            }

            this.blind = GEN_BLIND;
            this.initial = GEN_INITIAL;

            this
        };

        let mut gj = Jacobian::default();
        gj.set_ge(&AFFINE_G);

        // Construct a group element with no known corresponding scalar (nothing up my sleeve).
        let mut nums_32 = [0u8; 32];
        debug_assert!(b"The scalar for this x is unknown".len() == 32);
        for (i, v) in b"The scalar for this x is unknown".iter().enumerate() {
            nums_32[i] = *v;
        }
        let mut nums_x = Field::default();
        assert!(nums_x.set_b32(&nums_32));
        let mut nums_ge = Affine::default();
        assert!(nums_ge.set_xo_var(&nums_x, false));
        let mut nums_gej = Jacobian::default();
        nums_gej.set_ge(&nums_ge);
        nums_gej = nums_gej.add_ge_var(&AFFINE_G, None);

        // Compute prec.
        let mut precj: Vec<Jacobian> = Vec::with_capacity(1024);
        for _ in 0..1024 {
            precj.push(Jacobian::default());
        }
        let mut gbase = gj;
        let mut numsbase = nums_gej;
        for j in 0..64 {
            precj[j * 16] = numsbase;
            for i in 1..16 {
                precj[j * 16 + i] = precj[j * 16 + i - 1].add_var(&gbase, None);
            }
            for _ in 0..4 {
                gbase = gbase.double_var(None);
            }
            numsbase = numsbase.double_var(None);
            if j == 62 {
                numsbase = numsbase.neg();
                numsbase = numsbase.add_var(&nums_gej, None);
            }
        }
        let prec = set_all_gej_var(&precj);

        for j in 0..64 {
            for i in 0..16 {
                let pg: AffineStorage = prec[j * 16 + i].into();
                this.prec[j][i] = pg;
            }
        }

        this
    }
}

pub fn odd_multiples_table(prej: &mut [Jacobian], zr: &mut [Field], a: &Jacobian) {
    debug_assert!(prej.len() == zr.len());
    debug_assert!(!prej.is_empty());
    debug_assert!(!a.is_infinity());

    let d = a.double_var(None);
    let d_ge = Affine {
        x: d.x,
        y: d.y,
        infinity: false,
    };

    let mut a_ge = Affine::default();
    a_ge.set_gej_zinv(a, &d.z);
    prej[0].x = a_ge.x;
    prej[0].y = a_ge.y;
    prej[0].z = a.z;
    prej[0].infinity = false;

    zr[0] = d.z;
    for i in 1..prej.len() {
        prej[i] = prej[i - 1].add_ge_var(&d_ge, Some(&mut zr[i]));
    }

    let l = prej.last().unwrap().z * d.z;
    prej.last_mut().unwrap().z = l;
}

fn odd_multiples_table_globalz_windowa(
    pre: &mut [Affine; ECMULT_TABLE_SIZE_A],
    globalz: &mut Field,
    a: &Jacobian,
) {
    let mut prej: [Jacobian; ECMULT_TABLE_SIZE_A] = Default::default();
    let mut zr: [Field; ECMULT_TABLE_SIZE_A] = Default::default();

    odd_multiples_table(&mut prej, &mut zr, a);
    globalz_set_table_gej(pre, globalz, &prej, &zr);
}

fn table_get_ge(r: &mut Affine, pre: &[Affine], n: i32, w: usize) {
    debug_assert!(n & 1 == 1);
    debug_assert!(n >= -((1 << (w - 1)) - 1));
    debug_assert!(n <= ((1 << (w - 1)) - 1));
    if n > 0 {
        *r = pre[((n - 1) / 2) as usize];
    } else {
        *r = pre[((-n - 1) / 2) as usize].neg();
    }
}

fn table_get_ge_const(r: &mut Affine, pre: &[Affine], n: i32, w: usize) {
    let abs_n = n * (if n > 0 { 1 } else { 0 } * 2 - 1);
    let idx_n = abs_n / 2;
    debug_assert!(n & 1 == 1);
    debug_assert!(n >= -((1 << (w - 1)) - 1));
    debug_assert!(n <= ((1 << (w - 1)) - 1));
    for m in 0..pre.len() {
        let flag = m == idx_n as usize;
        r.x.cmov(&pre[m].x, flag);
        r.y.cmov(&pre[m].y, flag);
    }
    r.infinity = false;
    let neg_y = r.y.neg(1);
    r.y.cmov(&neg_y, n != abs_n);
}

fn table_get_ge_storage(r: &mut Affine, pre: &[AffineStorage], n: i32, w: usize) {
    debug_assert!(n & 1 == 1);
    debug_assert!(n >= -((1 << (w - 1)) - 1));
    debug_assert!(n <= ((1 << (w - 1)) - 1));
    if n > 0 {
        *r = pre[((n - 1) / 2) as usize].into();
    } else {
        *r = pre[((-n - 1) / 2) as usize].into();
        *r = r.neg();
    }
}

pub fn ecmult_wnaf(wnaf: &mut [i32], a: &Scalar, w: usize) -> i32 {
    let mut s = *a;
    let mut last_set_bit: i32 = -1;
    let mut bit = 0;
    let mut sign = 1;
    let mut carry = 0;

    debug_assert!(wnaf.len() <= 256);
    debug_assert!(w >= 2 && w <= 31);

    for i in 0..wnaf.len() {
        wnaf[i] = 0;
    }

    if s.bits(255, 1) > 0 {
        s = -s;
        sign = -1;
    }

    while bit < wnaf.len() {
        let mut now;
        let mut word;
        if s.bits(bit, 1) == carry as u32 {
            bit += 1;
            continue;
        }

        now = w;
        if now > wnaf.len() - bit {
            now = wnaf.len() - bit;
        }

        word = (s.bits_var(bit, now) as i32) + carry;

        carry = (word >> (w - 1)) & 1;
        word -= carry << w;

        wnaf[bit] = sign * word;
        last_set_bit = bit as i32;

        bit += now;
    }
    debug_assert!(carry == 0);
    debug_assert!({
        let mut t = true;
        while bit < 256 {
            t = t && (s.bits(bit, 1) == 0);
            bit += 1;
        }
        t
    });
    last_set_bit + 1
}

pub fn ecmult_wnaf_const(wnaf: &mut [i32], a: &Scalar, w: usize) -> i32 {
    let mut s = *a;
    let mut word = 0;

    /* Note that we cannot handle even numbers by negating them to be
     * odd, as is done in other implementations, since if our scalars
     * were specified to have width < 256 for performance reasons,
     * their negations would have width 256 and we'd lose any
     * performance benefit. Instead, we use a technique from Section
     * 4.2 of the Okeya/Tagaki paper, which is to add either 1 (for
     * even) or 2 (for odd) to the number we are encoding, returning a
     * skew value indicating this, and having the caller compensate
     * after doing the multiplication. */

    /* Negative numbers will be negated to keep their bit
     * representation below the maximum width */
    let flip = s.is_high();
    /* We add 1 to even numbers, 2 to odd ones, noting that negation
     * flips parity */
    let bit = flip ^ !s.is_even();
    /* We add 1 to even numbers, 2 to odd ones, noting that negation
     * flips parity */
    let neg_s = -s;
    let not_neg_one = !neg_s.is_one();
    s.cadd_bit(if bit { 1 } else { 0 }, not_neg_one);
    /* If we had negative one, flip == 1, s.d[0] == 0, bit == 1, so
     * caller expects that we added two to it and flipped it. In fact
     * for -1 these operations are identical. We only flipped, but
     * since skewing is required (in the sense that the skew must be 1
     * or 2, never zero) and flipping is not, we need to change our
     * flags to claim that we only skewed. */
    let mut global_sign = if flip { -1 } else { 1 };
    s.cond_neg_assign(Choice::from(flip as u8));
    global_sign *= if not_neg_one { 1 } else { 0 } * 2 - 1;
    let skew = 1 << (if bit { 1 } else { 0 });

    let mut u_last: i32 = s.shr_int(w) as i32;
    let mut u: i32 = 0;
    while word * w < WNAF_BITS {
        u = s.shr_int(w) as i32;
        let even = (u & 1) == 0;
        let sign = 2 * (if u_last > 0 { 1 } else { 0 }) - 1;
        u += sign * if even { 1 } else { 0 };
        u_last -= sign * if even { 1 } else { 0 } * (1 << w);

        wnaf[word] = (u_last as i32 * global_sign as i32) as i32;
        word += 1;

        u_last = u;
    }
    wnaf[word] = u * global_sign as i32;

    debug_assert!(s.is_zero());
    let wnaf_size = (WNAF_BITS + w - 1) / w;
    debug_assert!(word == wnaf_size);

    skew
}

impl ECMultContext {
    pub fn ecmult(&self, r: &mut Jacobian, a: &Jacobian, na: &Scalar, ng: &Scalar) {
        let mut tmpa = Affine::default();
        let mut pre_a: [Affine; ECMULT_TABLE_SIZE_A] = Default::default();
        let mut z = Field::default();
        let mut wnaf_na = [0i32; 256];
        let mut wnaf_ng = [0i32; 256];
        let bits_na = ecmult_wnaf(&mut wnaf_na, na, WINDOW_A);
        let mut bits = bits_na;
        odd_multiples_table_globalz_windowa(&mut pre_a, &mut z, a);

        let bits_ng = ecmult_wnaf(&mut wnaf_ng, &ng, WINDOW_G);
        if bits_ng > bits {
            bits = bits_ng;
        }

        r.set_infinity();
        for i in (0..bits).rev() {
            let mut n;
            *r = r.double_var(None);

            n = wnaf_na[i as usize];
            if i < bits_na && n != 0 {
                table_get_ge(&mut tmpa, &pre_a, n, WINDOW_A);
                *r = r.add_ge_var(&tmpa, None);
            }
            n = wnaf_ng[i as usize];
            if i < bits_ng && n != 0 {
                table_get_ge_storage(&mut tmpa, &self.pre_g, n, WINDOW_G);
                *r = r.add_zinv_var(&tmpa, &z);
            }
        }

        if !r.is_infinity() {
            r.z *= &z;
        }
    }

    pub fn ecmult_const(&self, r: &mut Jacobian, a: &Affine, scalar: &Scalar) {
        const WNAF_SIZE: usize = (WNAF_BITS + (WINDOW_A - 1) - 1) / (WINDOW_A - 1);

        let mut tmpa = Affine::default();
        let mut pre_a: [Affine; ECMULT_TABLE_SIZE_A] = Default::default();
        let mut z = Field::default();

        let mut wnaf_1 = [0i32; 1 + WNAF_SIZE];

        let sc = *scalar;
        let skew_1 = ecmult_wnaf_const(&mut wnaf_1, &sc, WINDOW_A - 1);

        /* Calculate odd multiples of a.  All multiples are brought to
         * the same Z 'denominator', which is stored in Z. Due to
         * secp256k1' isomorphism we can do all operations pretending
         * that the Z coordinate was 1, use affine addition formulae,
         * and correct the Z coordinate of the result once at the end.
         */
        r.set_ge(a);
        odd_multiples_table_globalz_windowa(&mut pre_a, &mut z, r);
        for i in 0..ECMULT_TABLE_SIZE_A {
            pre_a[i].y.normalize_weak();
        }

        /* first loop iteration (separated out so we can directly set
         * r, rather than having it start at infinity, get doubled
         * several times, then have its new value added to it) */
        let i = wnaf_1[WNAF_SIZE];
        debug_assert!(i != 0);
        table_get_ge_const(&mut tmpa, &pre_a, i, WINDOW_A);
        r.set_ge(&tmpa);

        /* remaining loop iterations */
        for i in (0..WNAF_SIZE).rev() {
            for _ in 0..(WINDOW_A - 1) {
                let r2 = *r;
                r.double_nonzero_in_place(&r2, None);
            }

            let n = wnaf_1[i];
            table_get_ge_const(&mut tmpa, &pre_a, n, WINDOW_A);
            debug_assert!(n != 0);
            *r = r.add_ge(&tmpa);
        }

        r.z *= &z;

        /* Correct for wNAF skew */
        let mut correction = *a;
        let mut correction_1_stor: AffineStorage;
        let a2_stor: AffineStorage;
        let mut tmpj = Jacobian::default();
        tmpj.set_ge(&correction);
        tmpj = tmpj.double_var(None);
        correction.set_gej(&tmpj);
        correction_1_stor = (*a).into();
        a2_stor = correction.into();

        /* For odd numbers this is 2a (so replace it), for even ones a (so no-op) */
        correction_1_stor.cmov(&a2_stor, skew_1 == 2);

        /* Apply the correction */
        correction = correction_1_stor.into();
        correction = correction.neg();
        *r = r.add_ge(&correction)
    }
}

impl ECMultGenContext {
    pub fn ecmult_gen(&self, r: &mut Jacobian, gn: &Scalar) {
        let mut adds = AffineStorage::default();
        *r = self.initial;

        let mut gnb = gn + &self.blind;
        let mut add = Affine::default();
        add.infinity = false;

        for j in 0..64 {
            let mut bits = gnb.bits(j * 4, 4);
            for i in 0..16 {
                adds.cmov(&self.prec[j][i], i as u32 == bits);
            }
            add = adds.into();
            *r = r.add_ge(&add);
            #[allow(unused_assignments)]
            {
                bits = 0;
            }
        }
        add.clear();
        gnb.clear();
    }
}