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_d_h_s(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_d_h_s)8 int ostei_h_d_h_s(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_d_h_s)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__h_d_h_s);
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
27 // partition workspace
28 double * const INT__h_s_h_s = work + (SIMINT_NSHELL_SIMD * 0);
29 double * const INT__i_s_h_s = work + (SIMINT_NSHELL_SIMD * 441);
30 double * const INT__k_s_h_s = work + (SIMINT_NSHELL_SIMD * 1029);
31 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*1785);
32 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
33 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_s_s = primwork + 13;
34 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_p_s = primwork + 49;
35 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_s_s = primwork + 94;
36 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_p_s = primwork + 160;
37 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_d_s = primwork + 250;
38 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_s_s = primwork + 394;
39 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_p_s = primwork + 494;
40 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_d_s = primwork + 644;
41 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_f_s = primwork + 884;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_s_s = primwork + 1184;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_p_s = primwork + 1319;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_d_s = primwork + 1544;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_f_s = primwork + 1904;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_g_s = primwork + 2354;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_s_s = primwork + 2804;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_p_s = primwork + 2972;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_d_s = primwork + 3287;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_f_s = primwork + 3791;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_g_s = primwork + 4421;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_h_s = primwork + 5051;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_s_s = primwork + 5492;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_p_s = primwork + 5688;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_d_s = primwork + 6108;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_f_s = primwork + 6780;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_g_s = primwork + 7620;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_h_s = primwork + 8460;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_s_s = primwork + 9048;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_p_s = primwork + 9264;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_d_s = primwork + 9804;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_f_s = primwork + 10668;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_g_s = primwork + 11748;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_h_s = primwork + 12828;
65 double * const hrrwork = (double *)(primwork + 13584);
66 double * const HRR_INT__h_p_h_s = hrrwork + 0;
67 double * const HRR_INT__i_p_h_s = hrrwork + 1323;
68
69
70 // Create constants
71 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
72 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
73 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
74 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
75 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
76 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
77 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
78 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
79
80
81 ////////////////////////////////////////
82 // Loop over shells and primitives
83 ////////////////////////////////////////
84
85 real_abcd = 0;
86 istart = 0;
87 for(ab = 0; ab < P.nshell12_clip; ++ab)
88 {
89 const int iend = istart + P.nprim12[ab];
90
91 cd = 0;
92 jstart = 0;
93
94 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
95 {
96 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
97 int jend = jstart;
98 for(i = 0; i < nshellbatch; i++)
99 jend += Q.nprim12[cd+i];
100
101 // Clear the beginning of the workspace (where we are accumulating integrals)
102 memset(work, 0, SIMINT_NSHELL_SIMD * 1785 * sizeof(double));
103 abcd = 0;
104
105
106 for(i = istart; i < iend; ++i)
107 {
108 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
109
110 if(check_screen)
111 {
112 // Skip this whole thing if always insignificant
113 if((P.screen[i] * Q.screen_max) < screen_tol)
114 continue;
115 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
116 }
117
118 icd = 0;
119 iprimcd = 0;
120 nprim_icd = Q.nprim12[cd];
121 double * restrict PRIM_PTR_INT__h_s_h_s = INT__h_s_h_s + abcd * 441;
122 double * restrict PRIM_PTR_INT__i_s_h_s = INT__i_s_h_s + abcd * 588;
123 double * restrict PRIM_PTR_INT__k_s_h_s = INT__k_s_h_s + abcd * 756;
124
125
126
127 // Load these one per loop over i
128 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
129 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
130 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
131
132 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
133
134 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
135 {
136 // calculate the shell offsets
137 // these are the offset from the shell pointed to by cd
138 // for each element
139 int shelloffsets[SIMINT_SIMD_LEN] = {0};
140 int lastoffset = 0;
141 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
142
143 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
144 {
145 // Handle if the first element of the vector is a new shell
146 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
147 {
148 nprim_icd += Q.nprim12[cd + (++icd)];
149 PRIM_PTR_INT__h_s_h_s += 441;
150 PRIM_PTR_INT__i_s_h_s += 588;
151 PRIM_PTR_INT__k_s_h_s += 756;
152 }
153 iprimcd++;
154 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
155 {
156 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
157 {
158 shelloffsets[n] = shelloffsets[n-1] + 1;
159 lastoffset++;
160 nprim_icd += Q.nprim12[cd + (++icd)];
161 }
162 else
163 shelloffsets[n] = shelloffsets[n-1];
164 iprimcd++;
165 }
166 }
167 else
168 iprimcd += SIMINT_SIMD_LEN;
169
170 // Do we have to compute this vector (or has it been screened out)?
171 // (not_screened != 0 means we have to do this vector)
172 if(check_screen)
173 {
174 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
175 if(vmax < screen_tol)
176 {
177 PRIM_PTR_INT__h_s_h_s += lastoffset*441;
178 PRIM_PTR_INT__i_s_h_s += lastoffset*588;
179 PRIM_PTR_INT__k_s_h_s += lastoffset*756;
180 continue;
181 }
182 }
183
184 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
185 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
186 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
187 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
188
189
190 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
191 SIMINT_DBLTYPE PQ[3];
192 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
193 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
194 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
195 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
196 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
197 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
198
199 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
200 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
201 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
202 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
203 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
204 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
205 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
206
207 // NOTE: Minus sign!
208 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
209 SIMINT_DBLTYPE aop_PQ[3];
210 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
211 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
212 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
213
214 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
215 SIMINT_DBLTYPE aoq_PQ[3];
216 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
217 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
218 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
219 // Put a minus sign here so we don't have to in RR routines
220 a_over_q = SIMINT_NEG(a_over_q);
221
222
223 //////////////////////////////////////////////
224 // Fjt function section
225 // Maximum v value: 12
226 //////////////////////////////////////////////
227 // The parameter to the Fjt function
228 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
229
230
231 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
232
233
234 boys_F_split(PRIM_INT__s_s_s_s, F_x, 12);
235 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
236 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
237 for(n = 0; n <= 12; n++)
238 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
239
240 //////////////////////////////////////////////
241 // Primitive integrals: Vertical recurrance
242 //////////////////////////////////////////////
243
244 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
245 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
246 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
247 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
248 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
249 const SIMINT_DBLTYPE vrr_const_6_over_2p = SIMINT_MUL(const_6, one_over_2p);
250 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
251 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
252 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
253 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
254 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
255 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
256 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
257 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
258 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
259 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
260 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
261
262
263
264 // Forming PRIM_INT__p_s_s_s[12 * 3];
265 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
266 {
267
268 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
269 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_s_s[n * 3 + 0]);
270
271 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
272 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_s_s[n * 3 + 1]);
273
274 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
275 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_s_s[n * 3 + 2]);
276
277 }
278
279
280
281 // Forming PRIM_INT__d_s_s_s[11 * 6];
282 for(n = 0; n < 11; ++n) // loop over orders of auxiliary function
283 {
284
285 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_s_s[n * 3 + 0]);
286 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_s_s[n * 6 + 0]);
287 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__d_s_s_s[n * 6 + 0]);
288
289 PRIM_INT__d_s_s_s[n * 6 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 0]);
290 PRIM_INT__d_s_s_s[n * 6 + 1] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_s_s[n * 6 + 1]);
291
292 PRIM_INT__d_s_s_s[n * 6 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 0]);
293 PRIM_INT__d_s_s_s[n * 6 + 2] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_s_s[n * 6 + 2]);
294
295 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 1]);
296 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__d_s_s_s[n * 6 + 3]);
297 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__d_s_s_s[n * 6 + 3]);
298
299 PRIM_INT__d_s_s_s[n * 6 + 4] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 1]);
300 PRIM_INT__d_s_s_s[n * 6 + 4] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__d_s_s_s[n * 6 + 4]);
301
302 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 2]);
303 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__d_s_s_s[n * 6 + 5]);
304 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__s_s_s_s[n * 1 + 0]), PRIM_INT__d_s_s_s[n * 6 + 5]);
305
306 }
307
308
309
310 // Forming PRIM_INT__f_s_s_s[10 * 10];
311 for(n = 0; n < 10; ++n) // loop over orders of auxiliary function
312 {
313
314 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 0]);
315 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__f_s_s_s[n * 10 + 0]);
316 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_FMADD( vrr_const_2_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__p_s_s_s[n * 3 + 0]), PRIM_INT__f_s_s_s[n * 10 + 0]);
317
318 PRIM_INT__f_s_s_s[n * 10 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 0]);
319 PRIM_INT__f_s_s_s[n * 10 + 1] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__f_s_s_s[n * 10 + 1]);
320
321 PRIM_INT__f_s_s_s[n * 10 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 0]);
322 PRIM_INT__f_s_s_s[n * 10 + 2] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__f_s_s_s[n * 10 + 2]);
323
324 PRIM_INT__f_s_s_s[n * 10 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 3]);
325 PRIM_INT__f_s_s_s[n * 10 + 3] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__f_s_s_s[n * 10 + 3]);
326
327 PRIM_INT__f_s_s_s[n * 10 + 4] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 1]);
328 PRIM_INT__f_s_s_s[n * 10 + 4] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 1], PRIM_INT__f_s_s_s[n * 10 + 4]);
329
330 PRIM_INT__f_s_s_s[n * 10 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 5]);
331 PRIM_INT__f_s_s_s[n * 10 + 5] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__f_s_s_s[n * 10 + 5]);
332
333 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 3]);
334 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__f_s_s_s[n * 10 + 6]);
335 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_FMADD( vrr_const_2_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__p_s_s_s[n * 3 + 1]), PRIM_INT__f_s_s_s[n * 10 + 6]);
336
337 PRIM_INT__f_s_s_s[n * 10 + 7] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 3]);
338 PRIM_INT__f_s_s_s[n * 10 + 7] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__f_s_s_s[n * 10 + 7]);
339
340 PRIM_INT__f_s_s_s[n * 10 + 8] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 5]);
341 PRIM_INT__f_s_s_s[n * 10 + 8] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__f_s_s_s[n * 10 + 8]);
342
343 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 5]);
344 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__f_s_s_s[n * 10 + 9]);
345 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_FMADD( vrr_const_2_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__p_s_s_s[n * 3 + 2]), PRIM_INT__f_s_s_s[n * 10 + 9]);
346
347 }
348
349
350 VRR_I_g_s_s_s(
351 PRIM_INT__g_s_s_s,
352 PRIM_INT__f_s_s_s,
353 PRIM_INT__d_s_s_s,
354 P_PA,
355 a_over_p,
356 aop_PQ,
357 one_over_2p,
358 9);
359
360
361 VRR_I_h_s_s_s(
362 PRIM_INT__h_s_s_s,
363 PRIM_INT__g_s_s_s,
364 PRIM_INT__f_s_s_s,
365 P_PA,
366 a_over_p,
367 aop_PQ,
368 one_over_2p,
369 8);
370
371
372 ostei_general_vrr_K(5, 0, 1, 0, 5,
373 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
374 PRIM_INT__h_s_s_s, NULL, NULL, PRIM_INT__g_s_s_s, NULL, PRIM_INT__h_s_p_s);
375
376
377 VRR_K_g_s_p_s(
378 PRIM_INT__g_s_p_s,
379 PRIM_INT__g_s_s_s,
380 PRIM_INT__f_s_s_s,
381 Q_PA,
382 aoq_PQ,
383 one_over_2pq,
384 5);
385
386
387 ostei_general_vrr_K(5, 0, 2, 0, 4,
388 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
389 PRIM_INT__h_s_p_s, PRIM_INT__h_s_s_s, NULL, PRIM_INT__g_s_p_s, NULL, PRIM_INT__h_s_d_s);
390
391
392 VRR_K_f_s_p_s(
393 PRIM_INT__f_s_p_s,
394 PRIM_INT__f_s_s_s,
395 PRIM_INT__d_s_s_s,
396 Q_PA,
397 aoq_PQ,
398 one_over_2pq,
399 5);
400
401
402 ostei_general_vrr_K(4, 0, 2, 0, 4,
403 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
404 PRIM_INT__g_s_p_s, PRIM_INT__g_s_s_s, NULL, PRIM_INT__f_s_p_s, NULL, PRIM_INT__g_s_d_s);
405
406
407 ostei_general_vrr_K(5, 0, 3, 0, 3,
408 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
409 PRIM_INT__h_s_d_s, PRIM_INT__h_s_p_s, NULL, PRIM_INT__g_s_d_s, NULL, PRIM_INT__h_s_f_s);
410
411
412
413 // Forming PRIM_INT__d_s_p_s[5 * 18];
414 for(n = 0; n < 5; ++n) // loop over orders of auxiliary function
415 {
416
417 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 0]);
418 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__d_s_p_s[n * 18 + 0]);
419 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_p_s[n * 18 + 0]);
420
421 PRIM_INT__d_s_p_s[n * 18 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 0]);
422 PRIM_INT__d_s_p_s[n * 18 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__d_s_p_s[n * 18 + 1]);
423
424 PRIM_INT__d_s_p_s[n * 18 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 0]);
425 PRIM_INT__d_s_p_s[n * 18 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 0], PRIM_INT__d_s_p_s[n * 18 + 2]);
426
427 PRIM_INT__d_s_p_s[n * 18 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 1]);
428 PRIM_INT__d_s_p_s[n * 18 + 3] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 1], PRIM_INT__d_s_p_s[n * 18 + 3]);
429 PRIM_INT__d_s_p_s[n * 18 + 3] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__d_s_p_s[n * 18 + 3]);
430
431 PRIM_INT__d_s_p_s[n * 18 + 4] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 1]);
432 PRIM_INT__d_s_p_s[n * 18 + 4] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 1], PRIM_INT__d_s_p_s[n * 18 + 4]);
433 PRIM_INT__d_s_p_s[n * 18 + 4] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_p_s[n * 18 + 4]);
434
435 PRIM_INT__d_s_p_s[n * 18 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 1]);
436 PRIM_INT__d_s_p_s[n * 18 + 5] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 1], PRIM_INT__d_s_p_s[n * 18 + 5]);
437
438 PRIM_INT__d_s_p_s[n * 18 + 6] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 2]);
439 PRIM_INT__d_s_p_s[n * 18 + 6] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 2], PRIM_INT__d_s_p_s[n * 18 + 6]);
440 PRIM_INT__d_s_p_s[n * 18 + 6] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__d_s_p_s[n * 18 + 6]);
441
442 PRIM_INT__d_s_p_s[n * 18 + 7] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 2]);
443 PRIM_INT__d_s_p_s[n * 18 + 7] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 2], PRIM_INT__d_s_p_s[n * 18 + 7]);
444
445 PRIM_INT__d_s_p_s[n * 18 + 8] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 2]);
446 PRIM_INT__d_s_p_s[n * 18 + 8] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 2], PRIM_INT__d_s_p_s[n * 18 + 8]);
447 PRIM_INT__d_s_p_s[n * 18 + 8] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__d_s_p_s[n * 18 + 8]);
448
449 PRIM_INT__d_s_p_s[n * 18 + 9] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 3]);
450 PRIM_INT__d_s_p_s[n * 18 + 9] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__d_s_p_s[n * 18 + 9]);
451
452 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 3]);
453 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__d_s_p_s[n * 18 + 10]);
454 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__d_s_p_s[n * 18 + 10]);
455
456 PRIM_INT__d_s_p_s[n * 18 + 11] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 3]);
457 PRIM_INT__d_s_p_s[n * 18 + 11] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__d_s_p_s[n * 18 + 11]);
458
459 PRIM_INT__d_s_p_s[n * 18 + 12] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 4]);
460 PRIM_INT__d_s_p_s[n * 18 + 12] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 4], PRIM_INT__d_s_p_s[n * 18 + 12]);
461
462 PRIM_INT__d_s_p_s[n * 18 + 13] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 4]);
463 PRIM_INT__d_s_p_s[n * 18 + 13] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 4], PRIM_INT__d_s_p_s[n * 18 + 13]);
464 PRIM_INT__d_s_p_s[n * 18 + 13] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__d_s_p_s[n * 18 + 13]);
465
466 PRIM_INT__d_s_p_s[n * 18 + 14] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 4]);
467 PRIM_INT__d_s_p_s[n * 18 + 14] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 4], PRIM_INT__d_s_p_s[n * 18 + 14]);
468 PRIM_INT__d_s_p_s[n * 18 + 14] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__d_s_p_s[n * 18 + 14]);
469
470 PRIM_INT__d_s_p_s[n * 18 + 15] = SIMINT_MUL(Q_PA[0], PRIM_INT__d_s_s_s[n * 6 + 5]);
471 PRIM_INT__d_s_p_s[n * 18 + 15] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__d_s_p_s[n * 18 + 15]);
472
473 PRIM_INT__d_s_p_s[n * 18 + 16] = SIMINT_MUL(Q_PA[1], PRIM_INT__d_s_s_s[n * 6 + 5]);
474 PRIM_INT__d_s_p_s[n * 18 + 16] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__d_s_p_s[n * 18 + 16]);
475
476 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_MUL(Q_PA[2], PRIM_INT__d_s_s_s[n * 6 + 5]);
477 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__d_s_p_s[n * 18 + 17]);
478 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__d_s_p_s[n * 18 + 17]);
479
480 }
481
482
483 VRR_K_f_s_d_s(
484 PRIM_INT__f_s_d_s,
485 PRIM_INT__f_s_p_s,
486 PRIM_INT__f_s_s_s,
487 PRIM_INT__d_s_p_s,
488 Q_PA,
489 a_over_q,
490 aoq_PQ,
491 one_over_2pq,
492 one_over_2q,
493 4);
494
495
496 ostei_general_vrr_K(4, 0, 3, 0, 3,
497 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
498 PRIM_INT__g_s_d_s, PRIM_INT__g_s_p_s, NULL, PRIM_INT__f_s_d_s, NULL, PRIM_INT__g_s_f_s);
499
500
501 ostei_general_vrr_K(5, 0, 4, 0, 2,
502 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
503 PRIM_INT__h_s_f_s, PRIM_INT__h_s_d_s, NULL, PRIM_INT__g_s_f_s, NULL, PRIM_INT__h_s_g_s);
504
505
506
507 // Forming PRIM_INT__p_s_p_s[5 * 9];
508 for(n = 0; n < 5; ++n) // loop over orders of auxiliary function
509 {
510
511 PRIM_INT__p_s_p_s[n * 9 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__p_s_s_s[n * 3 + 0]);
512 PRIM_INT__p_s_p_s[n * 9 + 0] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 0]);
513 PRIM_INT__p_s_p_s[n * 9 + 0] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_p_s[n * 9 + 0]);
514
515 PRIM_INT__p_s_p_s[n * 9 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__p_s_s_s[n * 3 + 0]);
516 PRIM_INT__p_s_p_s[n * 9 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 1]);
517
518 PRIM_INT__p_s_p_s[n * 9 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__p_s_s_s[n * 3 + 0]);
519 PRIM_INT__p_s_p_s[n * 9 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 2]);
520
521 PRIM_INT__p_s_p_s[n * 9 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__p_s_s_s[n * 3 + 1]);
522 PRIM_INT__p_s_p_s[n * 9 + 3] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 3]);
523
524 PRIM_INT__p_s_p_s[n * 9 + 4] = SIMINT_MUL(Q_PA[1], PRIM_INT__p_s_s_s[n * 3 + 1]);
525 PRIM_INT__p_s_p_s[n * 9 + 4] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 4]);
526 PRIM_INT__p_s_p_s[n * 9 + 4] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_p_s[n * 9 + 4]);
527
528 PRIM_INT__p_s_p_s[n * 9 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__p_s_s_s[n * 3 + 1]);
529 PRIM_INT__p_s_p_s[n * 9 + 5] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 5]);
530
531 PRIM_INT__p_s_p_s[n * 9 + 6] = SIMINT_MUL(Q_PA[0], PRIM_INT__p_s_s_s[n * 3 + 2]);
532 PRIM_INT__p_s_p_s[n * 9 + 6] = SIMINT_FMADD( aoq_PQ[0], PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 6]);
533
534 PRIM_INT__p_s_p_s[n * 9 + 7] = SIMINT_MUL(Q_PA[1], PRIM_INT__p_s_s_s[n * 3 + 2]);
535 PRIM_INT__p_s_p_s[n * 9 + 7] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 7]);
536
537 PRIM_INT__p_s_p_s[n * 9 + 8] = SIMINT_MUL(Q_PA[2], PRIM_INT__p_s_s_s[n * 3 + 2]);
538 PRIM_INT__p_s_p_s[n * 9 + 8] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__p_s_s_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 8]);
539 PRIM_INT__p_s_p_s[n * 9 + 8] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_s_s[(n+1) * 1 + 0], PRIM_INT__p_s_p_s[n * 9 + 8]);
540
541 }
542
543
544 VRR_K_d_s_d_s(
545 PRIM_INT__d_s_d_s,
546 PRIM_INT__d_s_p_s,
547 PRIM_INT__d_s_s_s,
548 PRIM_INT__p_s_p_s,
549 Q_PA,
550 a_over_q,
551 aoq_PQ,
552 one_over_2pq,
553 one_over_2q,
554 4);
555
556
557 ostei_general_vrr_K(3, 0, 3, 0, 3,
558 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
559 PRIM_INT__f_s_d_s, PRIM_INT__f_s_p_s, NULL, PRIM_INT__d_s_d_s, NULL, PRIM_INT__f_s_f_s);
560
561
562 ostei_general_vrr_K(4, 0, 4, 0, 2,
563 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
564 PRIM_INT__g_s_f_s, PRIM_INT__g_s_d_s, NULL, PRIM_INT__f_s_f_s, NULL, PRIM_INT__g_s_g_s);
565
566
567 ostei_general_vrr_K(5, 0, 5, 0, 1,
568 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
569 PRIM_INT__h_s_g_s, PRIM_INT__h_s_f_s, NULL, PRIM_INT__g_s_g_s, NULL, PRIM_INT__h_s_h_s);
570
571
572 ostei_general_vrr1_I(6, 7,
573 one_over_2p, a_over_p, aop_PQ, P_PA,
574 PRIM_INT__h_s_s_s, PRIM_INT__g_s_s_s, PRIM_INT__i_s_s_s);
575
576
577 ostei_general_vrr_K(6, 0, 1, 0, 5,
578 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
579 PRIM_INT__i_s_s_s, NULL, NULL, PRIM_INT__h_s_s_s, NULL, PRIM_INT__i_s_p_s);
580
581
582 ostei_general_vrr_K(6, 0, 2, 0, 4,
583 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
584 PRIM_INT__i_s_p_s, PRIM_INT__i_s_s_s, NULL, PRIM_INT__h_s_p_s, NULL, PRIM_INT__i_s_d_s);
585
586
587 ostei_general_vrr_K(6, 0, 3, 0, 3,
588 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
589 PRIM_INT__i_s_d_s, PRIM_INT__i_s_p_s, NULL, PRIM_INT__h_s_d_s, NULL, PRIM_INT__i_s_f_s);
590
591
592 ostei_general_vrr_K(6, 0, 4, 0, 2,
593 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
594 PRIM_INT__i_s_f_s, PRIM_INT__i_s_d_s, NULL, PRIM_INT__h_s_f_s, NULL, PRIM_INT__i_s_g_s);
595
596
597 ostei_general_vrr_K(6, 0, 5, 0, 1,
598 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
599 PRIM_INT__i_s_g_s, PRIM_INT__i_s_f_s, NULL, PRIM_INT__h_s_g_s, NULL, PRIM_INT__i_s_h_s);
600
601
602 ostei_general_vrr1_I(7, 6,
603 one_over_2p, a_over_p, aop_PQ, P_PA,
604 PRIM_INT__i_s_s_s, PRIM_INT__h_s_s_s, PRIM_INT__k_s_s_s);
605
606
607 ostei_general_vrr_K(7, 0, 1, 0, 5,
608 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
609 PRIM_INT__k_s_s_s, NULL, NULL, PRIM_INT__i_s_s_s, NULL, PRIM_INT__k_s_p_s);
610
611
612 ostei_general_vrr_K(7, 0, 2, 0, 4,
613 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
614 PRIM_INT__k_s_p_s, PRIM_INT__k_s_s_s, NULL, PRIM_INT__i_s_p_s, NULL, PRIM_INT__k_s_d_s);
615
616
617 ostei_general_vrr_K(7, 0, 3, 0, 3,
618 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
619 PRIM_INT__k_s_d_s, PRIM_INT__k_s_p_s, NULL, PRIM_INT__i_s_d_s, NULL, PRIM_INT__k_s_f_s);
620
621
622 ostei_general_vrr_K(7, 0, 4, 0, 2,
623 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
624 PRIM_INT__k_s_f_s, PRIM_INT__k_s_d_s, NULL, PRIM_INT__i_s_f_s, NULL, PRIM_INT__k_s_g_s);
625
626
627 ostei_general_vrr_K(7, 0, 5, 0, 1,
628 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
629 PRIM_INT__k_s_g_s, PRIM_INT__k_s_f_s, NULL, PRIM_INT__i_s_g_s, NULL, PRIM_INT__k_s_h_s);
630
631
632
633
634 ////////////////////////////////////
635 // Accumulate contracted integrals
636 ////////////////////////////////////
637 if(lastoffset == 0)
638 {
639 contract_all(441, PRIM_INT__h_s_h_s, PRIM_PTR_INT__h_s_h_s);
640 contract_all(588, PRIM_INT__i_s_h_s, PRIM_PTR_INT__i_s_h_s);
641 contract_all(756, PRIM_INT__k_s_h_s, PRIM_PTR_INT__k_s_h_s);
642 }
643 else
644 {
645 contract(441, shelloffsets, PRIM_INT__h_s_h_s, PRIM_PTR_INT__h_s_h_s);
646 contract(588, shelloffsets, PRIM_INT__i_s_h_s, PRIM_PTR_INT__i_s_h_s);
647 contract(756, shelloffsets, PRIM_INT__k_s_h_s, PRIM_PTR_INT__k_s_h_s);
648 PRIM_PTR_INT__h_s_h_s += lastoffset*441;
649 PRIM_PTR_INT__i_s_h_s += lastoffset*588;
650 PRIM_PTR_INT__k_s_h_s += lastoffset*756;
651 }
652
653 } // close loop over j
654 } // close loop over i
655
656 //Advance to the next batch
657 jstart = SIMINT_SIMD_ROUND(jend);
658
659 //////////////////////////////////////////////
660 // Contracted integrals: Horizontal recurrance
661 //////////////////////////////////////////////
662
663
664 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
665
666
667 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
668 {
669
670 // set up HRR pointers
671 double const * restrict HRR_INT__h_s_h_s = INT__h_s_h_s + abcd * 441;
672 double const * restrict HRR_INT__i_s_h_s = INT__i_s_h_s + abcd * 588;
673 double const * restrict HRR_INT__k_s_h_s = INT__k_s_h_s + abcd * 756;
674 double * restrict HRR_INT__h_d_h_s = INT__h_d_h_s + real_abcd * 2646;
675
676 // form INT__h_p_h_s
677 ostei_general_hrr_J(5, 1, 5, 0, hAB, HRR_INT__i_s_h_s, HRR_INT__h_s_h_s, HRR_INT__h_p_h_s);
678
679 // form INT__i_p_h_s
680 ostei_general_hrr_J(6, 1, 5, 0, hAB, HRR_INT__k_s_h_s, HRR_INT__i_s_h_s, HRR_INT__i_p_h_s);
681
682 // form INT__h_d_h_s
683 ostei_general_hrr_J(5, 2, 5, 0, hAB, HRR_INT__i_p_h_s, HRR_INT__h_p_h_s, HRR_INT__h_d_h_s);
684
685
686 } // close HRR loop
687
688
689 } // close loop cdbatch
690
691 istart = iend;
692 } // close loop over ab
693
694 return P.nshell12_clip * Q.nshell12_clip;
695 }
696
ostei_d_h_h_s(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_h_h_s)697 int ostei_d_h_h_s(struct simint_multi_shellpair const P,
698 struct simint_multi_shellpair const Q,
699 double screen_tol,
700 double * const restrict work,
701 double * const restrict INT__d_h_h_s)
702 {
703 double P_AB[3*P.nshell12];
704 struct simint_multi_shellpair P_tmp = P;
705 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
706 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
707 P_tmp.AB_x = P_AB;
708 P_tmp.AB_y = P_AB + P.nshell12;
709 P_tmp.AB_z = P_AB + 2*P.nshell12;
710
711 for(int i = 0; i < P.nshell12; i++)
712 {
713 P_tmp.AB_x[i] = -P.AB_x[i];
714 P_tmp.AB_y[i] = -P.AB_y[i];
715 P_tmp.AB_z[i] = -P.AB_z[i];
716 }
717
718 int ret = ostei_h_d_h_s(P_tmp, Q, screen_tol, work, INT__d_h_h_s);
719 double buffer[2646] SIMINT_ALIGN_ARRAY_DBL;
720
721 for(int q = 0; q < ret; q++)
722 {
723 int idx = 0;
724 for(int a = 0; a < 6; ++a)
725 for(int b = 0; b < 21; ++b)
726 for(int c = 0; c < 21; ++c)
727 for(int d = 0; d < 1; ++d)
728 buffer[idx++] = INT__d_h_h_s[q*2646+b*126+a*21+c*1+d];
729
730 memcpy(INT__d_h_h_s+q*2646, buffer, 2646*sizeof(double));
731 }
732
733 return ret;
734 }
735
ostei_h_d_s_h(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_d_s_h)736 int ostei_h_d_s_h(struct simint_multi_shellpair const P,
737 struct simint_multi_shellpair const Q,
738 double screen_tol,
739 double * const restrict work,
740 double * const restrict INT__h_d_s_h)
741 {
742 double Q_AB[3*Q.nshell12];
743 struct simint_multi_shellpair Q_tmp = Q;
744 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
745 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
746 Q_tmp.AB_x = Q_AB;
747 Q_tmp.AB_y = Q_AB + Q.nshell12;
748 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
749
750 for(int i = 0; i < Q.nshell12; i++)
751 {
752 Q_tmp.AB_x[i] = -Q.AB_x[i];
753 Q_tmp.AB_y[i] = -Q.AB_y[i];
754 Q_tmp.AB_z[i] = -Q.AB_z[i];
755 }
756
757 int ret = ostei_h_d_h_s(P, Q_tmp, screen_tol, work, INT__h_d_s_h);
758 double buffer[2646] SIMINT_ALIGN_ARRAY_DBL;
759
760 for(int q = 0; q < ret; q++)
761 {
762 int idx = 0;
763 for(int a = 0; a < 21; ++a)
764 for(int b = 0; b < 6; ++b)
765 for(int c = 0; c < 1; ++c)
766 for(int d = 0; d < 21; ++d)
767 buffer[idx++] = INT__h_d_s_h[q*2646+a*126+b*21+d*1+c];
768
769 memcpy(INT__h_d_s_h+q*2646, buffer, 2646*sizeof(double));
770 }
771
772 return ret;
773 }
774
ostei_d_h_s_h(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_h_s_h)775 int ostei_d_h_s_h(struct simint_multi_shellpair const P,
776 struct simint_multi_shellpair const Q,
777 double screen_tol,
778 double * const restrict work,
779 double * const restrict INT__d_h_s_h)
780 {
781 double P_AB[3*P.nshell12];
782 struct simint_multi_shellpair P_tmp = P;
783 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
784 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
785 P_tmp.AB_x = P_AB;
786 P_tmp.AB_y = P_AB + P.nshell12;
787 P_tmp.AB_z = P_AB + 2*P.nshell12;
788
789 for(int i = 0; i < P.nshell12; i++)
790 {
791 P_tmp.AB_x[i] = -P.AB_x[i];
792 P_tmp.AB_y[i] = -P.AB_y[i];
793 P_tmp.AB_z[i] = -P.AB_z[i];
794 }
795
796 double Q_AB[3*Q.nshell12];
797 struct simint_multi_shellpair Q_tmp = Q;
798 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
799 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
800 Q_tmp.AB_x = Q_AB;
801 Q_tmp.AB_y = Q_AB + Q.nshell12;
802 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
803
804 for(int i = 0; i < Q.nshell12; i++)
805 {
806 Q_tmp.AB_x[i] = -Q.AB_x[i];
807 Q_tmp.AB_y[i] = -Q.AB_y[i];
808 Q_tmp.AB_z[i] = -Q.AB_z[i];
809 }
810
811 int ret = ostei_h_d_h_s(P_tmp, Q_tmp, screen_tol, work, INT__d_h_s_h);
812 double buffer[2646] SIMINT_ALIGN_ARRAY_DBL;
813
814 for(int q = 0; q < ret; q++)
815 {
816 int idx = 0;
817 for(int a = 0; a < 6; ++a)
818 for(int b = 0; b < 21; ++b)
819 for(int c = 0; c < 1; ++c)
820 for(int d = 0; d < 21; ++d)
821 buffer[idx++] = INT__d_h_s_h[q*2646+b*126+a*21+d*1+c];
822
823 memcpy(INT__d_h_s_h+q*2646, buffer, 2646*sizeof(double));
824 }
825
826 return ret;
827 }
828
829