36 for (i = 1; i + 4 <= na; i += 4) {
37 __uint128_t
a0 = (__uint128_t)numa[i] * x + cl;
38 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + (
mp_limb_t)(
a0 >> 64);
39 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + (
mp_limb_t)(
a1 >> 64);
40 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + (
mp_limb_t)(
a2 >> 64);
48 __uint128_t a = (__uint128_t)numa[i] * x + cl;
54 for (j = 1; j + 1 < na; j++) {
57 for (i = j + 1; i + 4 <= na; i += 4) {
58 __uint128_t
a0 = (__uint128_t)numa[i] * x + dst[i + j] + cl;
59 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + dst[i + j + 1] + (
mp_limb_t)(
a0 >> 64);
60 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + dst[i + j + 2] + (
mp_limb_t)(
a1 >> 64);
61 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + dst[i + j + 3] + (
mp_limb_t)(
a2 >> 64);
69 __uint128_t a = (__uint128_t)numa[i] * x + dst[i + j] + cl;
77 for (i = 1; i + 4 <= 2 * na - 1; i += 4) {
78 __uint128_t
a0 = (__uint128_t)dst[i] * 2 + cl;
79 __uint128_t
a1 = (__uint128_t)dst[i + 1] * 2 + (
mp_limb_t)(
a0 >> 64);
80 __uint128_t
a2 = (__uint128_t)dst[i + 2] * 2 + (
mp_limb_t)(
a1 >> 64);
81 __uint128_t
a3 = (__uint128_t)dst[i + 3] * 2 + (
mp_limb_t)(
a2 >> 64);
88 for (; i < 2 * na - 1; i++) {
89 __uint128_t a = (__uint128_t)dst[i] * 2 + cl;
96 for (i = 0; i < na; i++) {
97 __uint128_t
t = (__uint128_t)numa[i] * numa[i] + dst[2 * i] + cl;
100 t = (__uint128_t)dst[2 * i + 1] + cl;
122 for (i = 0; i + 4 <= na; i += 4) {
123 __uint128_t
a0 = (__uint128_t)numa[i + 0] * x + cl;
124 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + (uint64_t)(
a0 >> 64);
125 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + (uint64_t)(
a1 >> 64);
126 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + (uint64_t)(
a2 >> 64);
127 dst[i + 0] = (uint64_t)
a0;
128 dst[i + 1] = (uint64_t)
a1;
129 dst[i + 2] = (uint64_t)
a2;
130 dst[i + 3] = (uint64_t)
a3;
131 cl = (uint64_t)(
a3 >> 64);
133 for (; i < na; i++) {
134 __uint128_t a = (__uint128_t)numa[i] * x + cl;
135 dst[i] = (uint64_t)a;
136 cl = (uint64_t)(a >> 64);
144 for (; nb >= 2; j += 2, dp += 2, nb -= 2) {
147 for (i = 0; i + 4 <= na; i += 4) {
148 __uint128_t
a0 = (__uint128_t)numa[i + 0] * x + dp[i + 0] + cl;
149 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + dp[i + 1] + (uint64_t)(
a0 >> 64);
150 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + dp[i + 2] + (uint64_t)(
a1 >> 64);
151 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + dp[i + 3] + (uint64_t)(
a2 >> 64);
152 dp[i + 0] = (uint64_t)
a0;
153 dp[i + 1] = (uint64_t)
a1;
154 dp[i + 2] = (uint64_t)
a2;
155 dp[i + 3] = (uint64_t)
a3;
156 cl = (uint64_t)(
a3 >> 64);
158 for (; i < na; i++) {
159 __uint128_t a = (__uint128_t)numa[i] * x + dp[i] + cl;
161 cl = (uint64_t)(a >> 64);
167 for (i = 0; i + 4 <= na; i += 4) {
168 __uint128_t
a0 = (__uint128_t)numa[i + 0] * x + dp[i + 1] + cl;
169 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + dp[i + 2] + (uint64_t)(
a0 >> 64);
170 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + dp[i + 3] + (uint64_t)(
a1 >> 64);
171 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + dp[i + 4] + (uint64_t)(
a2 >> 64);
172 dp[i + 1] = (uint64_t)
a0;
173 dp[i + 2] = (uint64_t)
a1;
174 dp[i + 3] = (uint64_t)
a2;
175 dp[i + 4] = (uint64_t)
a3;
176 cl = (uint64_t)(
a3 >> 64);
178 for (; i < na; i++) {
179 __uint128_t a = (__uint128_t)numa[i] * x + dp[i + 1] + cl;
180 dp[i + 1] = (uint64_t)a;
181 cl = (uint64_t)(a >> 64);
185 for (; nb >= 1; j++, dp++, nb--) {
188 for (i = 0; i + 4 <= na; i += 4) {
189 __uint128_t
a0 = (__uint128_t)numa[i + 0] * x + dp[i + 0] + cl;
190 __uint128_t
a1 = (__uint128_t)numa[i + 1] * x + dp[i + 1] + (uint64_t)(
a0 >> 64);
191 __uint128_t
a2 = (__uint128_t)numa[i + 2] * x + dp[i + 2] + (uint64_t)(
a1 >> 64);
192 __uint128_t
a3 = (__uint128_t)numa[i + 3] * x + dp[i + 3] + (uint64_t)(
a2 >> 64);
193 dp[i + 0] = (uint64_t)
a0;
194 dp[i + 1] = (uint64_t)
a1;
195 dp[i + 2] = (uint64_t)
a2;
196 dp[i + 3] = (uint64_t)
a3;
197 cl = (uint64_t)(
a3 >> 64);
199 for (; i < na; i++) {
200 __uint128_t a = (__uint128_t)numa[i] * x + dp[i] + cl;
202 cl = (uint64_t)(a >> 64);
void lmmp_sqr_basecase_(mp_ptr dst, mp_srcptr numa, mp_size_t na)
基础平方运算 [dst,2*na] = [numa,na]^2
void lmmp_mul_basecase_(mp_ptr dst, mp_srcptr numa, mp_size_t na, mp_srcptr numb, mp_size_t nb)
基础乘法运算 [dst,na+nb] = [numa,na] * [numb,nb]