1 #include "simint/boys/boys.h"
2 #include "simint/ostei/gen/ostei_generated.h"
3 #include "simint/vectorization/vectorization.h"
4 #include <math.h>
5 #include <string.h>
6
7
ostei_h_p_k_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_p_k_f)8 int ostei_h_p_k_f(struct simint_multi_shellpair const P,
9 struct simint_multi_shellpair const Q,
10 double screen_tol,
11 double * const restrict work,
12 double * const restrict INT__h_p_k_f)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__h_p_k_f);
17 int ab, cd, abcd;
18 int istart, jstart;
19 int iprimcd, nprim_icd, icd;
20 const int check_screen = (screen_tol > 0.0);
21 int i, j;
22 int n;
23 int not_screened;
24 int real_abcd;
25 int iket;
26 int ibra;
27
28 // partition workspace
29 double * const INT__h_s_k_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__h_s_l_s = work + (SIMINT_NSHELL_SIMD * 756);
31 double * const INT__h_s_m_s = work + (SIMINT_NSHELL_SIMD * 1701);
32 double * const INT__h_s_n_s = work + (SIMINT_NSHELL_SIMD * 2856);
33 double * const INT__i_s_k_s = work + (SIMINT_NSHELL_SIMD * 4242);
34 double * const INT__i_s_l_s = work + (SIMINT_NSHELL_SIMD * 5250);
35 double * const INT__i_s_m_s = work + (SIMINT_NSHELL_SIMD * 6510);
36 double * const INT__i_s_n_s = work + (SIMINT_NSHELL_SIMD * 8050);
37 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*9898);
38 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
39 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_p_s = primwork + 17;
40 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_d_s = primwork + 65;
41 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_f_s = primwork + 155;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_g_s = primwork + 295;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_h_s = primwork + 490;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_i_s = primwork + 742;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_k_s = primwork + 1050;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_l_s = primwork + 1410;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_m_s = primwork + 1815;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_n_s = primwork + 2255;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_d_s = primwork + 2717;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_f_s = primwork + 2825;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_g_s = primwork + 3005;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_h_s = primwork + 3275;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_i_s = primwork + 3653;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_k_s = primwork + 4157;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_l_s = primwork + 4805;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_m_s = primwork + 5615;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_n_s = primwork + 6605;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_f_s = primwork + 7793;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_g_s = primwork + 8093;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_h_s = primwork + 8543;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_i_s = primwork + 9173;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_k_s = primwork + 10013;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_l_s = primwork + 11093;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_m_s = primwork + 12443;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_n_s = primwork + 14093;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_g_s = primwork + 16073;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_h_s = primwork + 16673;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_i_s = primwork + 17513;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_k_s = primwork + 18633;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_l_s = primwork + 20073;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_m_s = primwork + 21873;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_n_s = primwork + 24073;
73 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_h_s = primwork + 26713;
74 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_i_s = primwork + 27658;
75 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_k_s = primwork + 28918;
76 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_l_s = primwork + 30538;
77 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_m_s = primwork + 32563;
78 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_n_s = primwork + 35038;
79 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_i_s = primwork + 38008;
80 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_k_s = primwork + 39184;
81 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_l_s = primwork + 40696;
82 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_m_s = primwork + 42586;
83 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_n_s = primwork + 44896;
84 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_k_s = primwork + 47668;
85 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_l_s = primwork + 48676;
86 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_m_s = primwork + 49936;
87 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_n_s = primwork + 51476;
88 double * const hrrwork = (double *)(primwork + 53324);
89 double * const HRR_INT__h_p_k_s = hrrwork + 0;
90 double * const HRR_INT__h_p_k_p = hrrwork + 2268;
91 double * const HRR_INT__h_p_k_d = hrrwork + 9072;
92 double * const HRR_INT__h_p_l_s = hrrwork + 22680;
93 double * const HRR_INT__h_p_l_p = hrrwork + 25515;
94 double * const HRR_INT__h_p_l_d = hrrwork + 34020;
95 double * const HRR_INT__h_p_m_s = hrrwork + 51030;
96 double * const HRR_INT__h_p_m_p = hrrwork + 54495;
97 double * const HRR_INT__h_p_n_s = hrrwork + 64890;
98
99
100 // Create constants
101 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
102 const SIMINT_DBLTYPE const_10 = SIMINT_DBLSET1(10);
103 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
104 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
105 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
106 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
107 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
108 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
109 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
110 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
111 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
112
113
114 ////////////////////////////////////////
115 // Loop over shells and primitives
116 ////////////////////////////////////////
117
118 real_abcd = 0;
119 istart = 0;
120 for(ab = 0; ab < P.nshell12_clip; ++ab)
121 {
122 const int iend = istart + P.nprim12[ab];
123
124 cd = 0;
125 jstart = 0;
126
127 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
128 {
129 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
130 int jend = jstart;
131 for(i = 0; i < nshellbatch; i++)
132 jend += Q.nprim12[cd+i];
133
134 // Clear the beginning of the workspace (where we are accumulating integrals)
135 memset(work, 0, SIMINT_NSHELL_SIMD * 9898 * sizeof(double));
136 abcd = 0;
137
138
139 for(i = istart; i < iend; ++i)
140 {
141 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
142
143 if(check_screen)
144 {
145 // Skip this whole thing if always insignificant
146 if((P.screen[i] * Q.screen_max) < screen_tol)
147 continue;
148 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
149 }
150
151 icd = 0;
152 iprimcd = 0;
153 nprim_icd = Q.nprim12[cd];
154 double * restrict PRIM_PTR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
155 double * restrict PRIM_PTR_INT__h_s_l_s = INT__h_s_l_s + abcd * 945;
156 double * restrict PRIM_PTR_INT__h_s_m_s = INT__h_s_m_s + abcd * 1155;
157 double * restrict PRIM_PTR_INT__h_s_n_s = INT__h_s_n_s + abcd * 1386;
158 double * restrict PRIM_PTR_INT__i_s_k_s = INT__i_s_k_s + abcd * 1008;
159 double * restrict PRIM_PTR_INT__i_s_l_s = INT__i_s_l_s + abcd * 1260;
160 double * restrict PRIM_PTR_INT__i_s_m_s = INT__i_s_m_s + abcd * 1540;
161 double * restrict PRIM_PTR_INT__i_s_n_s = INT__i_s_n_s + abcd * 1848;
162
163
164
165 // Load these one per loop over i
166 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
167 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
168 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
169
170 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
171
172 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
173 {
174 // calculate the shell offsets
175 // these are the offset from the shell pointed to by cd
176 // for each element
177 int shelloffsets[SIMINT_SIMD_LEN] = {0};
178 int lastoffset = 0;
179 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
180
181 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
182 {
183 // Handle if the first element of the vector is a new shell
184 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
185 {
186 nprim_icd += Q.nprim12[cd + (++icd)];
187 PRIM_PTR_INT__h_s_k_s += 756;
188 PRIM_PTR_INT__h_s_l_s += 945;
189 PRIM_PTR_INT__h_s_m_s += 1155;
190 PRIM_PTR_INT__h_s_n_s += 1386;
191 PRIM_PTR_INT__i_s_k_s += 1008;
192 PRIM_PTR_INT__i_s_l_s += 1260;
193 PRIM_PTR_INT__i_s_m_s += 1540;
194 PRIM_PTR_INT__i_s_n_s += 1848;
195 }
196 iprimcd++;
197 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
198 {
199 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
200 {
201 shelloffsets[n] = shelloffsets[n-1] + 1;
202 lastoffset++;
203 nprim_icd += Q.nprim12[cd + (++icd)];
204 }
205 else
206 shelloffsets[n] = shelloffsets[n-1];
207 iprimcd++;
208 }
209 }
210 else
211 iprimcd += SIMINT_SIMD_LEN;
212
213 // Do we have to compute this vector (or has it been screened out)?
214 // (not_screened != 0 means we have to do this vector)
215 if(check_screen)
216 {
217 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
218 if(vmax < screen_tol)
219 {
220 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
221 PRIM_PTR_INT__h_s_l_s += lastoffset*945;
222 PRIM_PTR_INT__h_s_m_s += lastoffset*1155;
223 PRIM_PTR_INT__h_s_n_s += lastoffset*1386;
224 PRIM_PTR_INT__i_s_k_s += lastoffset*1008;
225 PRIM_PTR_INT__i_s_l_s += lastoffset*1260;
226 PRIM_PTR_INT__i_s_m_s += lastoffset*1540;
227 PRIM_PTR_INT__i_s_n_s += lastoffset*1848;
228 continue;
229 }
230 }
231
232 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
233 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
234 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
235 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
236
237
238 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
239 SIMINT_DBLTYPE PQ[3];
240 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
241 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
242 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
243 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
244 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
245 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
246
247 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
248 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
249 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
250 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
251 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
252 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
253 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
254
255 // NOTE: Minus sign!
256 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
257 SIMINT_DBLTYPE aop_PQ[3];
258 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
259 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
260 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
261
262 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
263 SIMINT_DBLTYPE aoq_PQ[3];
264 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
265 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
266 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
267 // Put a minus sign here so we don't have to in RR routines
268 a_over_q = SIMINT_NEG(a_over_q);
269
270
271 //////////////////////////////////////////////
272 // Fjt function section
273 // Maximum v value: 16
274 //////////////////////////////////////////////
275 // The parameter to the Fjt function
276 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
277
278
279 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
280
281
282 boys_F_split(PRIM_INT__s_s_s_s, F_x, 16);
283 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
284 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
285 for(n = 0; n <= 16; n++)
286 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
287
288 //////////////////////////////////////////////
289 // Primitive integrals: Vertical recurrance
290 //////////////////////////////////////////////
291
292 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
293 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
294 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
295 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
296 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
297 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
298 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
299 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
300 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
301 const SIMINT_DBLTYPE vrr_const_5_over_2q = SIMINT_MUL(const_5, one_over_2q);
302 const SIMINT_DBLTYPE vrr_const_6_over_2q = SIMINT_MUL(const_6, one_over_2q);
303 const SIMINT_DBLTYPE vrr_const_7_over_2q = SIMINT_MUL(const_7, one_over_2q);
304 const SIMINT_DBLTYPE vrr_const_8_over_2q = SIMINT_MUL(const_8, one_over_2q);
305 const SIMINT_DBLTYPE vrr_const_9_over_2q = SIMINT_MUL(const_9, one_over_2q);
306 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
307 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
308 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
309 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
310 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
311 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
312 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
313 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
314 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
315 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
316
317
318
319 // Forming PRIM_INT__s_s_p_s[16 * 3];
320 for(n = 0; n < 16; ++n) // loop over orders of auxiliary function
321 {
322
323 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
324 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_p_s[n * 3 + 0]);
325
326 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
327 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_p_s[n * 3 + 1]);
328
329 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
330 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_p_s[n * 3 + 2]);
331
332 }
333
334
335
336 // Forming PRIM_INT__s_s_d_s[15 * 6];
337 for(n = 0; n < 15; ++n) // loop over orders of auxiliary function
338 {
339
340 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
341 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_d_s[n * 6 + 0]);
342 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_FMADD( vrr_const_1_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__s_s_d_s[n * 6 + 0]);
343
344 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 0]);
345 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_d_s[n * 6 + 1]);
346
347 PRIM_INT__s_s_d_s[n * 6 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 0]);
348 PRIM_INT__s_s_d_s[n * 6 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_d_s[n * 6 + 2]);
349
350 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 1]);
351 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_d_s[n * 6 + 3]);
352 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_FMADD( vrr_const_1_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__s_s_d_s[n * 6 + 3]);
353
354 PRIM_INT__s_s_d_s[n * 6 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 1]);
355 PRIM_INT__s_s_d_s[n * 6 + 4] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_d_s[n * 6 + 4]);
356
357 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
358 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__s_s_d_s[n * 6 + 5]);
359 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_FMADD( vrr_const_1_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__s_s_d_s[n * 6 + 5]);
360
361 }
362
363
364
365 // Forming PRIM_INT__s_s_f_s[14 * 10];
366 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
367 {
368
369 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
370 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__s_s_f_s[n * 10 + 0]);
371 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_FMADD( vrr_const_2_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_p_s[n * 3 + 0]), PRIM_INT__s_s_f_s[n * 10 + 0]);
372
373 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
374 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__s_s_f_s[n * 10 + 1]);
375
376 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
377 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__s_s_f_s[n * 10 + 2]);
378
379 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
380 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__s_s_f_s[n * 10 + 3]);
381
382 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
383 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__s_s_f_s[n * 10 + 4]);
384
385 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
386 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__s_s_f_s[n * 10 + 5]);
387
388 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
389 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__s_s_f_s[n * 10 + 6]);
390 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_FMADD( vrr_const_2_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_p_s[n * 3 + 1]), PRIM_INT__s_s_f_s[n * 10 + 6]);
391
392 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
393 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__s_s_f_s[n * 10 + 7]);
394
395 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
396 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__s_s_f_s[n * 10 + 8]);
397
398 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
399 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__s_s_f_s[n * 10 + 9]);
400 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_FMADD( vrr_const_2_over_2q, SIMINT_FMADD(a_over_q, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__s_s_p_s[n * 3 + 2]), PRIM_INT__s_s_f_s[n * 10 + 9]);
401
402 }
403
404
405 VRR_K_s_s_g_s(
406 PRIM_INT__s_s_g_s,
407 PRIM_INT__s_s_f_s,
408 PRIM_INT__s_s_d_s,
409 Q_PA,
410 a_over_q,
411 aoq_PQ,
412 one_over_2q,
413 13);
414
415
416 VRR_K_s_s_h_s(
417 PRIM_INT__s_s_h_s,
418 PRIM_INT__s_s_g_s,
419 PRIM_INT__s_s_f_s,
420 Q_PA,
421 a_over_q,
422 aoq_PQ,
423 one_over_2q,
424 12);
425
426
427 ostei_general_vrr1_K(6, 11,
428 one_over_2q, a_over_q, aoq_PQ, Q_PA,
429 PRIM_INT__s_s_h_s, PRIM_INT__s_s_g_s, PRIM_INT__s_s_i_s);
430
431
432 ostei_general_vrr1_K(7, 10,
433 one_over_2q, a_over_q, aoq_PQ, Q_PA,
434 PRIM_INT__s_s_i_s, PRIM_INT__s_s_h_s, PRIM_INT__s_s_k_s);
435
436
437 ostei_general_vrr_I(1, 0, 7, 0, 6,
438 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
439 PRIM_INT__s_s_k_s, NULL, NULL, PRIM_INT__s_s_i_s, NULL, PRIM_INT__p_s_k_s);
440
441
442 ostei_general_vrr_I(1, 0, 6, 0, 6,
443 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
444 PRIM_INT__s_s_i_s, NULL, NULL, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_i_s);
445
446
447 ostei_general_vrr_I(2, 0, 7, 0, 5,
448 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
449 PRIM_INT__p_s_k_s, PRIM_INT__s_s_k_s, NULL, PRIM_INT__p_s_i_s, NULL, PRIM_INT__d_s_k_s);
450
451
452 ostei_general_vrr_I(1, 0, 5, 0, 6,
453 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
454 PRIM_INT__s_s_h_s, NULL, NULL, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_h_s);
455
456
457 ostei_general_vrr_I(2, 0, 6, 0, 5,
458 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
459 PRIM_INT__p_s_i_s, PRIM_INT__s_s_i_s, NULL, PRIM_INT__p_s_h_s, NULL, PRIM_INT__d_s_i_s);
460
461
462 ostei_general_vrr_I(3, 0, 7, 0, 4,
463 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
464 PRIM_INT__d_s_k_s, PRIM_INT__p_s_k_s, NULL, PRIM_INT__d_s_i_s, NULL, PRIM_INT__f_s_k_s);
465
466
467 VRR_I_p_s_g_s(
468 PRIM_INT__p_s_g_s,
469 PRIM_INT__s_s_g_s,
470 PRIM_INT__s_s_f_s,
471 P_PA,
472 aop_PQ,
473 one_over_2pq,
474 6);
475
476
477 ostei_general_vrr_I(2, 0, 5, 0, 5,
478 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
479 PRIM_INT__p_s_h_s, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_g_s, NULL, PRIM_INT__d_s_h_s);
480
481
482 ostei_general_vrr_I(3, 0, 6, 0, 4,
483 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
484 PRIM_INT__d_s_i_s, PRIM_INT__p_s_i_s, NULL, PRIM_INT__d_s_h_s, NULL, PRIM_INT__f_s_i_s);
485
486
487 ostei_general_vrr_I(4, 0, 7, 0, 3,
488 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
489 PRIM_INT__f_s_k_s, PRIM_INT__d_s_k_s, NULL, PRIM_INT__f_s_i_s, NULL, PRIM_INT__g_s_k_s);
490
491
492 VRR_I_p_s_f_s(
493 PRIM_INT__p_s_f_s,
494 PRIM_INT__s_s_f_s,
495 PRIM_INT__s_s_d_s,
496 P_PA,
497 aop_PQ,
498 one_over_2pq,
499 6);
500
501
502 ostei_general_vrr_I(2, 0, 4, 0, 5,
503 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
504 PRIM_INT__p_s_g_s, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_f_s, NULL, PRIM_INT__d_s_g_s);
505
506
507 ostei_general_vrr_I(3, 0, 5, 0, 4,
508 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
509 PRIM_INT__d_s_h_s, PRIM_INT__p_s_h_s, NULL, PRIM_INT__d_s_g_s, NULL, PRIM_INT__f_s_h_s);
510
511
512 ostei_general_vrr_I(4, 0, 6, 0, 3,
513 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
514 PRIM_INT__f_s_i_s, PRIM_INT__d_s_i_s, NULL, PRIM_INT__f_s_h_s, NULL, PRIM_INT__g_s_i_s);
515
516
517 ostei_general_vrr_I(5, 0, 7, 0, 2,
518 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
519 PRIM_INT__g_s_k_s, PRIM_INT__f_s_k_s, NULL, PRIM_INT__g_s_i_s, NULL, PRIM_INT__h_s_k_s);
520
521
522 ostei_general_vrr1_K(8, 9,
523 one_over_2q, a_over_q, aoq_PQ, Q_PA,
524 PRIM_INT__s_s_k_s, PRIM_INT__s_s_i_s, PRIM_INT__s_s_l_s);
525
526
527 ostei_general_vrr_I(1, 0, 8, 0, 6,
528 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
529 PRIM_INT__s_s_l_s, NULL, NULL, PRIM_INT__s_s_k_s, NULL, PRIM_INT__p_s_l_s);
530
531
532 ostei_general_vrr_I(2, 0, 8, 0, 5,
533 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
534 PRIM_INT__p_s_l_s, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_k_s, NULL, PRIM_INT__d_s_l_s);
535
536
537 ostei_general_vrr_I(3, 0, 8, 0, 4,
538 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
539 PRIM_INT__d_s_l_s, PRIM_INT__p_s_l_s, NULL, PRIM_INT__d_s_k_s, NULL, PRIM_INT__f_s_l_s);
540
541
542 ostei_general_vrr_I(4, 0, 8, 0, 3,
543 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
544 PRIM_INT__f_s_l_s, PRIM_INT__d_s_l_s, NULL, PRIM_INT__f_s_k_s, NULL, PRIM_INT__g_s_l_s);
545
546
547 ostei_general_vrr_I(5, 0, 8, 0, 2,
548 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
549 PRIM_INT__g_s_l_s, PRIM_INT__f_s_l_s, NULL, PRIM_INT__g_s_k_s, NULL, PRIM_INT__h_s_l_s);
550
551
552
553 // Forming PRIM_INT__p_s_d_s[6 * 18];
554 for(n = 0; n < 6; ++n) // loop over orders of auxiliary function
555 {
556
557 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
558 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 0]);
559 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 0]);
560
561 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 1]);
562 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 1]);
563 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 1]);
564
565 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 2]);
566 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 2]);
567 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 2]);
568
569 PRIM_INT__p_s_d_s[n * 18 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
570 PRIM_INT__p_s_d_s[n * 18 + 3] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 3]);
571
572 PRIM_INT__p_s_d_s[n * 18 + 4] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 4]);
573 PRIM_INT__p_s_d_s[n * 18 + 4] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 4]);
574
575 PRIM_INT__p_s_d_s[n * 18 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
576 PRIM_INT__p_s_d_s[n * 18 + 5] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 5]);
577
578 PRIM_INT__p_s_d_s[n * 18 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
579 PRIM_INT__p_s_d_s[n * 18 + 6] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 6]);
580
581 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 1]);
582 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 7]);
583 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 7]);
584
585 PRIM_INT__p_s_d_s[n * 18 + 8] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 2]);
586 PRIM_INT__p_s_d_s[n * 18 + 8] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 8]);
587
588 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
589 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 9]);
590 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 9]);
591
592 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 4]);
593 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 10]);
594 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 10]);
595
596 PRIM_INT__p_s_d_s[n * 18 + 11] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
597 PRIM_INT__p_s_d_s[n * 18 + 11] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 11]);
598
599 PRIM_INT__p_s_d_s[n * 18 + 12] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
600 PRIM_INT__p_s_d_s[n * 18 + 12] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 12]);
601
602 PRIM_INT__p_s_d_s[n * 18 + 13] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
603 PRIM_INT__p_s_d_s[n * 18 + 13] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 13]);
604
605 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 2]);
606 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 14]);
607 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 14]);
608
609 PRIM_INT__p_s_d_s[n * 18 + 15] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
610 PRIM_INT__p_s_d_s[n * 18 + 15] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 15]);
611
612 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 4]);
613 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 16]);
614 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 16]);
615
616 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
617 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 17]);
618 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 17]);
619
620 }
621
622
623 VRR_I_d_s_f_s(
624 PRIM_INT__d_s_f_s,
625 PRIM_INT__p_s_f_s,
626 PRIM_INT__s_s_f_s,
627 PRIM_INT__p_s_d_s,
628 P_PA,
629 a_over_p,
630 aop_PQ,
631 one_over_2p,
632 one_over_2pq,
633 5);
634
635
636 ostei_general_vrr_I(3, 0, 4, 0, 4,
637 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
638 PRIM_INT__d_s_g_s, PRIM_INT__p_s_g_s, NULL, PRIM_INT__d_s_f_s, NULL, PRIM_INT__f_s_g_s);
639
640
641 ostei_general_vrr_I(4, 0, 5, 0, 3,
642 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
643 PRIM_INT__f_s_h_s, PRIM_INT__d_s_h_s, NULL, PRIM_INT__f_s_g_s, NULL, PRIM_INT__g_s_h_s);
644
645
646 ostei_general_vrr_I(5, 0, 6, 0, 2,
647 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
648 PRIM_INT__g_s_i_s, PRIM_INT__f_s_i_s, NULL, PRIM_INT__g_s_h_s, NULL, PRIM_INT__h_s_i_s);
649
650
651 ostei_general_vrr_I(6, 0, 7, 0, 1,
652 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
653 PRIM_INT__h_s_k_s, PRIM_INT__g_s_k_s, NULL, PRIM_INT__h_s_i_s, NULL, PRIM_INT__i_s_k_s);
654
655
656 ostei_general_vrr1_K(9, 8,
657 one_over_2q, a_over_q, aoq_PQ, Q_PA,
658 PRIM_INT__s_s_l_s, PRIM_INT__s_s_k_s, PRIM_INT__s_s_m_s);
659
660
661 ostei_general_vrr_I(1, 0, 9, 0, 6,
662 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
663 PRIM_INT__s_s_m_s, NULL, NULL, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_m_s);
664
665
666 ostei_general_vrr_I(2, 0, 9, 0, 5,
667 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
668 PRIM_INT__p_s_m_s, PRIM_INT__s_s_m_s, NULL, PRIM_INT__p_s_l_s, NULL, PRIM_INT__d_s_m_s);
669
670
671 ostei_general_vrr_I(3, 0, 9, 0, 4,
672 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
673 PRIM_INT__d_s_m_s, PRIM_INT__p_s_m_s, NULL, PRIM_INT__d_s_l_s, NULL, PRIM_INT__f_s_m_s);
674
675
676 ostei_general_vrr_I(4, 0, 9, 0, 3,
677 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
678 PRIM_INT__f_s_m_s, PRIM_INT__d_s_m_s, NULL, PRIM_INT__f_s_l_s, NULL, PRIM_INT__g_s_m_s);
679
680
681 ostei_general_vrr_I(5, 0, 9, 0, 2,
682 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
683 PRIM_INT__g_s_m_s, PRIM_INT__f_s_m_s, NULL, PRIM_INT__g_s_l_s, NULL, PRIM_INT__h_s_m_s);
684
685
686 ostei_general_vrr_I(6, 0, 8, 0, 1,
687 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
688 PRIM_INT__h_s_l_s, PRIM_INT__g_s_l_s, NULL, PRIM_INT__h_s_k_s, NULL, PRIM_INT__i_s_l_s);
689
690
691 ostei_general_vrr1_K(10, 7,
692 one_over_2q, a_over_q, aoq_PQ, Q_PA,
693 PRIM_INT__s_s_m_s, PRIM_INT__s_s_l_s, PRIM_INT__s_s_n_s);
694
695
696 ostei_general_vrr_I(1, 0, 10, 0, 6,
697 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
698 PRIM_INT__s_s_n_s, NULL, NULL, PRIM_INT__s_s_m_s, NULL, PRIM_INT__p_s_n_s);
699
700
701 ostei_general_vrr_I(2, 0, 10, 0, 5,
702 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
703 PRIM_INT__p_s_n_s, PRIM_INT__s_s_n_s, NULL, PRIM_INT__p_s_m_s, NULL, PRIM_INT__d_s_n_s);
704
705
706 ostei_general_vrr_I(3, 0, 10, 0, 4,
707 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
708 PRIM_INT__d_s_n_s, PRIM_INT__p_s_n_s, NULL, PRIM_INT__d_s_m_s, NULL, PRIM_INT__f_s_n_s);
709
710
711 ostei_general_vrr_I(4, 0, 10, 0, 3,
712 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
713 PRIM_INT__f_s_n_s, PRIM_INT__d_s_n_s, NULL, PRIM_INT__f_s_m_s, NULL, PRIM_INT__g_s_n_s);
714
715
716 ostei_general_vrr_I(5, 0, 10, 0, 2,
717 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
718 PRIM_INT__g_s_n_s, PRIM_INT__f_s_n_s, NULL, PRIM_INT__g_s_m_s, NULL, PRIM_INT__h_s_n_s);
719
720
721 ostei_general_vrr_I(6, 0, 9, 0, 1,
722 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
723 PRIM_INT__h_s_m_s, PRIM_INT__g_s_m_s, NULL, PRIM_INT__h_s_l_s, NULL, PRIM_INT__i_s_m_s);
724
725
726 ostei_general_vrr_I(6, 0, 10, 0, 1,
727 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
728 PRIM_INT__h_s_n_s, PRIM_INT__g_s_n_s, NULL, PRIM_INT__h_s_m_s, NULL, PRIM_INT__i_s_n_s);
729
730
731
732
733 ////////////////////////////////////
734 // Accumulate contracted integrals
735 ////////////////////////////////////
736 if(lastoffset == 0)
737 {
738 contract_all(756, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
739 contract_all(945, PRIM_INT__h_s_l_s, PRIM_PTR_INT__h_s_l_s);
740 contract_all(1155, PRIM_INT__h_s_m_s, PRIM_PTR_INT__h_s_m_s);
741 contract_all(1386, PRIM_INT__h_s_n_s, PRIM_PTR_INT__h_s_n_s);
742 contract_all(1008, PRIM_INT__i_s_k_s, PRIM_PTR_INT__i_s_k_s);
743 contract_all(1260, PRIM_INT__i_s_l_s, PRIM_PTR_INT__i_s_l_s);
744 contract_all(1540, PRIM_INT__i_s_m_s, PRIM_PTR_INT__i_s_m_s);
745 contract_all(1848, PRIM_INT__i_s_n_s, PRIM_PTR_INT__i_s_n_s);
746 }
747 else
748 {
749 contract(756, shelloffsets, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
750 contract(945, shelloffsets, PRIM_INT__h_s_l_s, PRIM_PTR_INT__h_s_l_s);
751 contract(1155, shelloffsets, PRIM_INT__h_s_m_s, PRIM_PTR_INT__h_s_m_s);
752 contract(1386, shelloffsets, PRIM_INT__h_s_n_s, PRIM_PTR_INT__h_s_n_s);
753 contract(1008, shelloffsets, PRIM_INT__i_s_k_s, PRIM_PTR_INT__i_s_k_s);
754 contract(1260, shelloffsets, PRIM_INT__i_s_l_s, PRIM_PTR_INT__i_s_l_s);
755 contract(1540, shelloffsets, PRIM_INT__i_s_m_s, PRIM_PTR_INT__i_s_m_s);
756 contract(1848, shelloffsets, PRIM_INT__i_s_n_s, PRIM_PTR_INT__i_s_n_s);
757 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
758 PRIM_PTR_INT__h_s_l_s += lastoffset*945;
759 PRIM_PTR_INT__h_s_m_s += lastoffset*1155;
760 PRIM_PTR_INT__h_s_n_s += lastoffset*1386;
761 PRIM_PTR_INT__i_s_k_s += lastoffset*1008;
762 PRIM_PTR_INT__i_s_l_s += lastoffset*1260;
763 PRIM_PTR_INT__i_s_m_s += lastoffset*1540;
764 PRIM_PTR_INT__i_s_n_s += lastoffset*1848;
765 }
766
767 } // close loop over j
768 } // close loop over i
769
770 //Advance to the next batch
771 jstart = SIMINT_SIMD_ROUND(jend);
772
773 //////////////////////////////////////////////
774 // Contracted integrals: Horizontal recurrance
775 //////////////////////////////////////////////
776
777
778 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
779
780
781 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
782 {
783 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
784
785 // set up HRR pointers
786 double const * restrict HRR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
787 double const * restrict HRR_INT__h_s_l_s = INT__h_s_l_s + abcd * 945;
788 double const * restrict HRR_INT__h_s_m_s = INT__h_s_m_s + abcd * 1155;
789 double const * restrict HRR_INT__h_s_n_s = INT__h_s_n_s + abcd * 1386;
790 double const * restrict HRR_INT__i_s_k_s = INT__i_s_k_s + abcd * 1008;
791 double const * restrict HRR_INT__i_s_l_s = INT__i_s_l_s + abcd * 1260;
792 double const * restrict HRR_INT__i_s_m_s = INT__i_s_m_s + abcd * 1540;
793 double const * restrict HRR_INT__i_s_n_s = INT__i_s_n_s + abcd * 1848;
794 double * restrict HRR_INT__h_p_k_f = INT__h_p_k_f + real_abcd * 22680;
795
796 // form INT__h_p_k_s
797 ostei_general_hrr_J(5, 1, 7, 0, hAB, HRR_INT__i_s_k_s, HRR_INT__h_s_k_s, HRR_INT__h_p_k_s);
798
799 // form INT__h_p_l_s
800 ostei_general_hrr_J(5, 1, 8, 0, hAB, HRR_INT__i_s_l_s, HRR_INT__h_s_l_s, HRR_INT__h_p_l_s);
801
802 // form INT__h_p_m_s
803 ostei_general_hrr_J(5, 1, 9, 0, hAB, HRR_INT__i_s_m_s, HRR_INT__h_s_m_s, HRR_INT__h_p_m_s);
804
805 // form INT__h_p_n_s
806 ostei_general_hrr_J(5, 1, 10, 0, hAB, HRR_INT__i_s_n_s, HRR_INT__h_s_n_s, HRR_INT__h_p_n_s);
807
808 // form INT__h_p_k_p
809 ostei_general_hrr_L(5, 1, 7, 1, hCD, HRR_INT__h_p_l_s, HRR_INT__h_p_k_s, HRR_INT__h_p_k_p);
810
811 // form INT__h_p_l_p
812 ostei_general_hrr_L(5, 1, 8, 1, hCD, HRR_INT__h_p_m_s, HRR_INT__h_p_l_s, HRR_INT__h_p_l_p);
813
814 // form INT__h_p_m_p
815 ostei_general_hrr_L(5, 1, 9, 1, hCD, HRR_INT__h_p_n_s, HRR_INT__h_p_m_s, HRR_INT__h_p_m_p);
816
817 // form INT__h_p_k_d
818 ostei_general_hrr_L(5, 1, 7, 2, hCD, HRR_INT__h_p_l_p, HRR_INT__h_p_k_p, HRR_INT__h_p_k_d);
819
820 // form INT__h_p_l_d
821 ostei_general_hrr_L(5, 1, 8, 2, hCD, HRR_INT__h_p_m_p, HRR_INT__h_p_l_p, HRR_INT__h_p_l_d);
822
823 // form INT__h_p_k_f
824 ostei_general_hrr_L(5, 1, 7, 3, hCD, HRR_INT__h_p_l_d, HRR_INT__h_p_k_d, HRR_INT__h_p_k_f);
825
826
827 } // close HRR loop
828
829
830 } // close loop cdbatch
831
832 istart = iend;
833 } // close loop over ab
834
835 return P.nshell12_clip * Q.nshell12_clip;
836 }
837
ostei_p_h_k_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__p_h_k_f)838 int ostei_p_h_k_f(struct simint_multi_shellpair const P,
839 struct simint_multi_shellpair const Q,
840 double screen_tol,
841 double * const restrict work,
842 double * const restrict INT__p_h_k_f)
843 {
844 double P_AB[3*P.nshell12];
845 struct simint_multi_shellpair P_tmp = P;
846 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
847 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
848 P_tmp.AB_x = P_AB;
849 P_tmp.AB_y = P_AB + P.nshell12;
850 P_tmp.AB_z = P_AB + 2*P.nshell12;
851
852 for(int i = 0; i < P.nshell12; i++)
853 {
854 P_tmp.AB_x[i] = -P.AB_x[i];
855 P_tmp.AB_y[i] = -P.AB_y[i];
856 P_tmp.AB_z[i] = -P.AB_z[i];
857 }
858
859 int ret = ostei_h_p_k_f(P_tmp, Q, screen_tol, work, INT__p_h_k_f);
860 double buffer[22680] SIMINT_ALIGN_ARRAY_DBL;
861
862 for(int q = 0; q < ret; q++)
863 {
864 int idx = 0;
865 for(int a = 0; a < 3; ++a)
866 for(int b = 0; b < 21; ++b)
867 for(int c = 0; c < 36; ++c)
868 for(int d = 0; d < 10; ++d)
869 buffer[idx++] = INT__p_h_k_f[q*22680+b*1080+a*360+c*10+d];
870
871 memcpy(INT__p_h_k_f+q*22680, buffer, 22680*sizeof(double));
872 }
873
874 return ret;
875 }
876
ostei_h_p_f_k(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_p_f_k)877 int ostei_h_p_f_k(struct simint_multi_shellpair const P,
878 struct simint_multi_shellpair const Q,
879 double screen_tol,
880 double * const restrict work,
881 double * const restrict INT__h_p_f_k)
882 {
883 double Q_AB[3*Q.nshell12];
884 struct simint_multi_shellpair Q_tmp = Q;
885 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
886 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
887 Q_tmp.AB_x = Q_AB;
888 Q_tmp.AB_y = Q_AB + Q.nshell12;
889 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
890
891 for(int i = 0; i < Q.nshell12; i++)
892 {
893 Q_tmp.AB_x[i] = -Q.AB_x[i];
894 Q_tmp.AB_y[i] = -Q.AB_y[i];
895 Q_tmp.AB_z[i] = -Q.AB_z[i];
896 }
897
898 int ret = ostei_h_p_k_f(P, Q_tmp, screen_tol, work, INT__h_p_f_k);
899 double buffer[22680] SIMINT_ALIGN_ARRAY_DBL;
900
901 for(int q = 0; q < ret; q++)
902 {
903 int idx = 0;
904 for(int a = 0; a < 21; ++a)
905 for(int b = 0; b < 3; ++b)
906 for(int c = 0; c < 10; ++c)
907 for(int d = 0; d < 36; ++d)
908 buffer[idx++] = INT__h_p_f_k[q*22680+a*1080+b*360+d*10+c];
909
910 memcpy(INT__h_p_f_k+q*22680, buffer, 22680*sizeof(double));
911 }
912
913 return ret;
914 }
915
ostei_p_h_f_k(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__p_h_f_k)916 int ostei_p_h_f_k(struct simint_multi_shellpair const P,
917 struct simint_multi_shellpair const Q,
918 double screen_tol,
919 double * const restrict work,
920 double * const restrict INT__p_h_f_k)
921 {
922 double P_AB[3*P.nshell12];
923 struct simint_multi_shellpair P_tmp = P;
924 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
925 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
926 P_tmp.AB_x = P_AB;
927 P_tmp.AB_y = P_AB + P.nshell12;
928 P_tmp.AB_z = P_AB + 2*P.nshell12;
929
930 for(int i = 0; i < P.nshell12; i++)
931 {
932 P_tmp.AB_x[i] = -P.AB_x[i];
933 P_tmp.AB_y[i] = -P.AB_y[i];
934 P_tmp.AB_z[i] = -P.AB_z[i];
935 }
936
937 double Q_AB[3*Q.nshell12];
938 struct simint_multi_shellpair Q_tmp = Q;
939 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
940 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
941 Q_tmp.AB_x = Q_AB;
942 Q_tmp.AB_y = Q_AB + Q.nshell12;
943 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
944
945 for(int i = 0; i < Q.nshell12; i++)
946 {
947 Q_tmp.AB_x[i] = -Q.AB_x[i];
948 Q_tmp.AB_y[i] = -Q.AB_y[i];
949 Q_tmp.AB_z[i] = -Q.AB_z[i];
950 }
951
952 int ret = ostei_h_p_k_f(P_tmp, Q_tmp, screen_tol, work, INT__p_h_f_k);
953 double buffer[22680] SIMINT_ALIGN_ARRAY_DBL;
954
955 for(int q = 0; q < ret; q++)
956 {
957 int idx = 0;
958 for(int a = 0; a < 3; ++a)
959 for(int b = 0; b < 21; ++b)
960 for(int c = 0; c < 10; ++c)
961 for(int d = 0; d < 36; ++d)
962 buffer[idx++] = INT__p_h_f_k[q*22680+b*1080+a*360+d*10+c];
963
964 memcpy(INT__p_h_f_k+q*22680, buffer, 22680*sizeof(double));
965 }
966
967 return ret;
968 }
969
970