26 const uint32_t MASK = 0x1fffff;
28 int64_t s0 = MASK &
load_3(s);
29 int64_t s1 = MASK & (
load_4(s + 2) >> 5);
30 int64_t s2 = MASK & (
load_3(s + 5) >> 2);
31 int64_t s3 = MASK & (
load_4(s + 7) >> 7);
32 int64_t s4 = MASK & (
load_4(s + 10) >> 4);
33 int64_t s5 = MASK & (
load_3(s + 13) >> 1);
34 int64_t s6 = MASK & (
load_4(s + 15) >> 6);
35 int64_t s7 = MASK & (
load_3(s + 18) >> 3);
36 int64_t s8 = MASK &
load_3(s + 21);
37 int64_t s9 = MASK & (
load_4(s + 23) >> 5);
38 int64_t s10 = MASK & (
load_3(s + 26) >> 2);
39 int64_t s11 = MASK & (
load_4(s + 28) >> 7);
40 int64_t s12 = MASK & (
load_4(s + 31) >> 4);
41 int64_t s13 = MASK & (
load_3(s + 34) >> 1);
42 int64_t s14 = MASK & (
load_4(s + 36) >> 6);
43 int64_t s15 = MASK & (
load_3(s + 39) >> 3);
44 int64_t s16 = MASK &
load_3(s + 42);
45 int64_t s17 = MASK & (
load_4(s + 44) >> 5);
46 int64_t s18 = MASK & (
load_3(s + 47) >> 2);
47 int64_t s19 = MASK & (
load_4(s + 49) >> 7);
48 int64_t s20 = MASK & (
load_4(s + 52) >> 4);
49 int64_t s21 = MASK & (
load_3(s + 55) >> 1);
50 int64_t s22 = MASK & (
load_4(s + 57) >> 6);
51 int64_t s23 = (
load_4(s + 60) >> 3);
53 redc_mul(s11, s12, s13, s14, s15, s16, s23);
54 redc_mul(s10, s11, s12, s13, s14, s15, s22);
55 redc_mul(s9, s10, s11, s12, s13, s14, s21);
56 redc_mul(s8, s9, s10, s11, s12, s13, s20);
57 redc_mul(s7, s8, s9, s10, s11, s12, s19);
58 redc_mul(s6, s7, s8, s9, s10, s11, s18);
73 redc_mul(s5, s6, s7, s8, s9, s10, s17);
74 redc_mul(s4, s5, s6, s7, s8, s9, s16);
75 redc_mul(s3, s4, s5, s6, s7, s8, s15);
76 redc_mul(s2, s3, s4, s5, s6, s7, s14);
77 redc_mul(s1, s2, s3, s4, s5, s6, s13);
78 redc_mul(s0, s1, s2, s3, s4, s5, s12);
94 redc_mul(s0, s1, s2, s3, s4, s5, s12);
109 redc_mul(s0, s1, s2, s3, s4, s5, s12);
124 s[0] =
static_cast<uint8_t
>(s0 >> 0);
125 s[1] =
static_cast<uint8_t
>(s0 >> 8);
126 s[2] =
static_cast<uint8_t
>((s0 >> 16) | (s1 << 5));
127 s[3] =
static_cast<uint8_t
>(s1 >> 3);
128 s[4] =
static_cast<uint8_t
>(s1 >> 11);
129 s[5] =
static_cast<uint8_t
>((s1 >> 19) | (s2 << 2));
130 s[6] =
static_cast<uint8_t
>(s2 >> 6);
131 s[7] =
static_cast<uint8_t
>((s2 >> 14) | (s3 << 7));
132 s[8] =
static_cast<uint8_t
>(s3 >> 1);
133 s[9] =
static_cast<uint8_t
>(s3 >> 9);
134 s[10] =
static_cast<uint8_t
>((s3 >> 17) | (s4 << 4));
135 s[11] =
static_cast<uint8_t
>(s4 >> 4);
136 s[12] =
static_cast<uint8_t
>(s4 >> 12);
137 s[13] =
static_cast<uint8_t
>((s4 >> 20) | (s5 << 1));
138 s[14] =
static_cast<uint8_t
>(s5 >> 7);
139 s[15] =
static_cast<uint8_t
>((s5 >> 15) | (s6 << 6));
140 s[16] =
static_cast<uint8_t
>(s6 >> 2);
141 s[17] =
static_cast<uint8_t
>(s6 >> 10);
142 s[18] =
static_cast<uint8_t
>((s6 >> 18) | (s7 << 3));
143 s[19] =
static_cast<uint8_t
>(s7 >> 5);
144 s[20] =
static_cast<uint8_t
>(s7 >> 13);
145 s[21] =
static_cast<uint8_t
>(s8 >> 0);
146 s[22] =
static_cast<uint8_t
>(s8 >> 8);
147 s[23] =
static_cast<uint8_t
>((s8 >> 16) | (s9 << 5));
148 s[24] =
static_cast<uint8_t
>(s9 >> 3);
149 s[25] =
static_cast<uint8_t
>(s9 >> 11);
150 s[26] =
static_cast<uint8_t
>((s9 >> 19) | (s10 << 2));
151 s[27] =
static_cast<uint8_t
>(s10 >> 6);
152 s[28] =
static_cast<uint8_t
>((s10 >> 14) | (s11 << 7));
153 s[29] =
static_cast<uint8_t
>(s11 >> 1);
154 s[30] =
static_cast<uint8_t
>(s11 >> 9);
155 s[31] =
static_cast<uint8_t
>(s11 >> 17);