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_f_d_g_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__f_d_g_f)8 int ostei_f_d_g_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__f_d_g_f)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__f_d_g_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__f_s_g_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__f_s_h_s = work + (SIMINT_NSHELL_SIMD * 150);
31 double * const INT__f_s_i_s = work + (SIMINT_NSHELL_SIMD * 360);
32 double * const INT__f_s_k_s = work + (SIMINT_NSHELL_SIMD * 640);
33 double * const INT__g_s_g_s = work + (SIMINT_NSHELL_SIMD * 1000);
34 double * const INT__g_s_h_s = work + (SIMINT_NSHELL_SIMD * 1225);
35 double * const INT__g_s_i_s = work + (SIMINT_NSHELL_SIMD * 1540);
36 double * const INT__g_s_k_s = work + (SIMINT_NSHELL_SIMD * 1960);
37 double * const INT__h_s_g_s = work + (SIMINT_NSHELL_SIMD * 2500);
38 double * const INT__h_s_h_s = work + (SIMINT_NSHELL_SIMD * 2815);
39 double * const INT__h_s_i_s = work + (SIMINT_NSHELL_SIMD * 3256);
40 double * const INT__h_s_k_s = work + (SIMINT_NSHELL_SIMD * 3844);
41 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*4600);
42 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_p_s = primwork + 13;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_d_s = primwork + 49;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_f_s = primwork + 115;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_g_s = primwork + 215;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_h_s = primwork + 350;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_i_s = primwork + 518;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_k_s = primwork + 714;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_s_s = primwork + 930;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_p_s = primwork + 945;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_d_s = primwork + 990;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_f_s = primwork + 1080;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_g_s = primwork + 1230;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_h_s = primwork + 1455;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_i_s = primwork + 1770;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_k_s = primwork + 2190;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_p_s = primwork + 2730;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_d_s = primwork + 2802;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_f_s = primwork + 2946;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_g_s = primwork + 3186;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_h_s = primwork + 3546;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_i_s = primwork + 4050;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_k_s = primwork + 4722;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_d_s = primwork + 5586;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_f_s = primwork + 5766;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_g_s = primwork + 6066;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_h_s = primwork + 6516;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_i_s = primwork + 7146;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_k_s = primwork + 7986;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_f_s = primwork + 9066;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_g_s = primwork + 9366;
73 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_h_s = primwork + 9816;
74 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_i_s = primwork + 10446;
75 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_k_s = primwork + 11286;
76 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_g_s = primwork + 12366;
77 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_h_s = primwork + 12681;
78 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_i_s = primwork + 13122;
79 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_k_s = primwork + 13710;
80 double * const hrrwork = (double *)(primwork + 14466);
81 double * const HRR_INT__f_p_g_s = hrrwork + 0;
82 double * const HRR_INT__f_p_h_s = hrrwork + 450;
83 double * const HRR_INT__f_p_i_s = hrrwork + 1080;
84 double * const HRR_INT__f_p_k_s = hrrwork + 1920;
85 double * const HRR_INT__f_d_g_s = hrrwork + 3000;
86 double * const HRR_INT__f_d_g_p = hrrwork + 3900;
87 double * const HRR_INT__f_d_g_d = hrrwork + 6600;
88 double * const HRR_INT__f_d_h_s = hrrwork + 12000;
89 double * const HRR_INT__f_d_h_p = hrrwork + 13260;
90 double * const HRR_INT__f_d_h_d = hrrwork + 17040;
91 double * const HRR_INT__f_d_i_s = hrrwork + 24600;
92 double * const HRR_INT__f_d_i_p = hrrwork + 26280;
93 double * const HRR_INT__f_d_k_s = hrrwork + 31320;
94 double * const HRR_INT__g_p_g_s = hrrwork + 33480;
95 double * const HRR_INT__g_p_h_s = hrrwork + 34155;
96 double * const HRR_INT__g_p_i_s = hrrwork + 35100;
97 double * const HRR_INT__g_p_k_s = hrrwork + 36360;
98
99
100 // Create constants
101 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
102 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
103 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
104 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
105 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
106 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
107 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
108 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
109
110
111 ////////////////////////////////////////
112 // Loop over shells and primitives
113 ////////////////////////////////////////
114
115 real_abcd = 0;
116 istart = 0;
117 for(ab = 0; ab < P.nshell12_clip; ++ab)
118 {
119 const int iend = istart + P.nprim12[ab];
120
121 cd = 0;
122 jstart = 0;
123
124 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
125 {
126 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
127 int jend = jstart;
128 for(i = 0; i < nshellbatch; i++)
129 jend += Q.nprim12[cd+i];
130
131 // Clear the beginning of the workspace (where we are accumulating integrals)
132 memset(work, 0, SIMINT_NSHELL_SIMD * 4600 * sizeof(double));
133 abcd = 0;
134
135
136 for(i = istart; i < iend; ++i)
137 {
138 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
139
140 if(check_screen)
141 {
142 // Skip this whole thing if always insignificant
143 if((P.screen[i] * Q.screen_max) < screen_tol)
144 continue;
145 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
146 }
147
148 icd = 0;
149 iprimcd = 0;
150 nprim_icd = Q.nprim12[cd];
151 double * restrict PRIM_PTR_INT__f_s_g_s = INT__f_s_g_s + abcd * 150;
152 double * restrict PRIM_PTR_INT__f_s_h_s = INT__f_s_h_s + abcd * 210;
153 double * restrict PRIM_PTR_INT__f_s_i_s = INT__f_s_i_s + abcd * 280;
154 double * restrict PRIM_PTR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
155 double * restrict PRIM_PTR_INT__g_s_g_s = INT__g_s_g_s + abcd * 225;
156 double * restrict PRIM_PTR_INT__g_s_h_s = INT__g_s_h_s + abcd * 315;
157 double * restrict PRIM_PTR_INT__g_s_i_s = INT__g_s_i_s + abcd * 420;
158 double * restrict PRIM_PTR_INT__g_s_k_s = INT__g_s_k_s + abcd * 540;
159 double * restrict PRIM_PTR_INT__h_s_g_s = INT__h_s_g_s + abcd * 315;
160 double * restrict PRIM_PTR_INT__h_s_h_s = INT__h_s_h_s + abcd * 441;
161 double * restrict PRIM_PTR_INT__h_s_i_s = INT__h_s_i_s + abcd * 588;
162 double * restrict PRIM_PTR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
163
164
165
166 // Load these one per loop over i
167 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
168 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
169 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
170
171 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
172
173 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
174 {
175 // calculate the shell offsets
176 // these are the offset from the shell pointed to by cd
177 // for each element
178 int shelloffsets[SIMINT_SIMD_LEN] = {0};
179 int lastoffset = 0;
180 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
181
182 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
183 {
184 // Handle if the first element of the vector is a new shell
185 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
186 {
187 nprim_icd += Q.nprim12[cd + (++icd)];
188 PRIM_PTR_INT__f_s_g_s += 150;
189 PRIM_PTR_INT__f_s_h_s += 210;
190 PRIM_PTR_INT__f_s_i_s += 280;
191 PRIM_PTR_INT__f_s_k_s += 360;
192 PRIM_PTR_INT__g_s_g_s += 225;
193 PRIM_PTR_INT__g_s_h_s += 315;
194 PRIM_PTR_INT__g_s_i_s += 420;
195 PRIM_PTR_INT__g_s_k_s += 540;
196 PRIM_PTR_INT__h_s_g_s += 315;
197 PRIM_PTR_INT__h_s_h_s += 441;
198 PRIM_PTR_INT__h_s_i_s += 588;
199 PRIM_PTR_INT__h_s_k_s += 756;
200 }
201 iprimcd++;
202 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
203 {
204 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
205 {
206 shelloffsets[n] = shelloffsets[n-1] + 1;
207 lastoffset++;
208 nprim_icd += Q.nprim12[cd + (++icd)];
209 }
210 else
211 shelloffsets[n] = shelloffsets[n-1];
212 iprimcd++;
213 }
214 }
215 else
216 iprimcd += SIMINT_SIMD_LEN;
217
218 // Do we have to compute this vector (or has it been screened out)?
219 // (not_screened != 0 means we have to do this vector)
220 if(check_screen)
221 {
222 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
223 if(vmax < screen_tol)
224 {
225 PRIM_PTR_INT__f_s_g_s += lastoffset*150;
226 PRIM_PTR_INT__f_s_h_s += lastoffset*210;
227 PRIM_PTR_INT__f_s_i_s += lastoffset*280;
228 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
229 PRIM_PTR_INT__g_s_g_s += lastoffset*225;
230 PRIM_PTR_INT__g_s_h_s += lastoffset*315;
231 PRIM_PTR_INT__g_s_i_s += lastoffset*420;
232 PRIM_PTR_INT__g_s_k_s += lastoffset*540;
233 PRIM_PTR_INT__h_s_g_s += lastoffset*315;
234 PRIM_PTR_INT__h_s_h_s += lastoffset*441;
235 PRIM_PTR_INT__h_s_i_s += lastoffset*588;
236 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
237 continue;
238 }
239 }
240
241 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
242 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
243 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
244 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
245
246
247 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
248 SIMINT_DBLTYPE PQ[3];
249 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
250 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
251 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
252 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
253 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
254 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
255
256 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
257 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
258 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
259 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
260 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
261 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
262 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
263
264 // NOTE: Minus sign!
265 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
266 SIMINT_DBLTYPE aop_PQ[3];
267 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
268 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
269 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
270
271 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
272 SIMINT_DBLTYPE aoq_PQ[3];
273 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
274 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
275 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
276 // Put a minus sign here so we don't have to in RR routines
277 a_over_q = SIMINT_NEG(a_over_q);
278
279
280 //////////////////////////////////////////////
281 // Fjt function section
282 // Maximum v value: 12
283 //////////////////////////////////////////////
284 // The parameter to the Fjt function
285 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
286
287
288 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
289
290
291 boys_F_split(PRIM_INT__s_s_s_s, F_x, 12);
292 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
293 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
294 for(n = 0; n <= 12; n++)
295 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
296
297 //////////////////////////////////////////////
298 // Primitive integrals: Vertical recurrance
299 //////////////////////////////////////////////
300
301 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
302 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
303 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
304 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
305 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
306 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
307 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
308 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
309 const SIMINT_DBLTYPE vrr_const_5_over_2q = SIMINT_MUL(const_5, one_over_2q);
310 const SIMINT_DBLTYPE vrr_const_6_over_2q = SIMINT_MUL(const_6, one_over_2q);
311 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
312 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
313 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
314 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
315 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
316 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
317 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
318
319
320
321 // Forming PRIM_INT__s_s_p_s[12 * 3];
322 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
323 {
324
325 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
326 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]);
327
328 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
329 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]);
330
331 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
332 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]);
333
334 }
335
336
337
338 // Forming PRIM_INT__s_s_d_s[11 * 6];
339 for(n = 0; n < 11; ++n) // loop over orders of auxiliary function
340 {
341
342 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
343 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]);
344 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]);
345
346 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 0]);
347 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]);
348
349 PRIM_INT__s_s_d_s[n * 6 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 0]);
350 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]);
351
352 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 1]);
353 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]);
354 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]);
355
356 PRIM_INT__s_s_d_s[n * 6 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 1]);
357 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]);
358
359 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
360 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]);
361 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]);
362
363 }
364
365
366
367 // Forming PRIM_INT__s_s_f_s[10 * 10];
368 for(n = 0; n < 10; ++n) // loop over orders of auxiliary function
369 {
370
371 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
372 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]);
373 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]);
374
375 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
376 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]);
377
378 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
379 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]);
380
381 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
382 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]);
383
384 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
385 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]);
386
387 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
388 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]);
389
390 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
391 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]);
392 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]);
393
394 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
395 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]);
396
397 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
398 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]);
399
400 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
401 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]);
402 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]);
403
404 }
405
406
407 VRR_K_s_s_g_s(
408 PRIM_INT__s_s_g_s,
409 PRIM_INT__s_s_f_s,
410 PRIM_INT__s_s_d_s,
411 Q_PA,
412 a_over_q,
413 aoq_PQ,
414 one_over_2q,
415 9);
416
417
418 VRR_I_p_s_g_s(
419 PRIM_INT__p_s_g_s,
420 PRIM_INT__s_s_g_s,
421 PRIM_INT__s_s_f_s,
422 P_PA,
423 aop_PQ,
424 one_over_2pq,
425 5);
426
427
428 VRR_I_p_s_f_s(
429 PRIM_INT__p_s_f_s,
430 PRIM_INT__s_s_f_s,
431 PRIM_INT__s_s_d_s,
432 P_PA,
433 aop_PQ,
434 one_over_2pq,
435 5);
436
437
438 ostei_general_vrr_I(2, 0, 4, 0, 4,
439 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
440 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);
441
442
443
444 // Forming PRIM_INT__p_s_d_s[5 * 18];
445 for(n = 0; n < 5; ++n) // loop over orders of auxiliary function
446 {
447
448 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
449 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]);
450 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]);
451
452 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 1]);
453 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]);
454 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]);
455
456 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 2]);
457 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]);
458 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]);
459
460 PRIM_INT__p_s_d_s[n * 18 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
461 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]);
462
463 PRIM_INT__p_s_d_s[n * 18 + 4] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 4]);
464 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]);
465
466 PRIM_INT__p_s_d_s[n * 18 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
467 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]);
468
469 PRIM_INT__p_s_d_s[n * 18 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
470 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]);
471
472 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 1]);
473 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]);
474 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]);
475
476 PRIM_INT__p_s_d_s[n * 18 + 8] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 2]);
477 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]);
478
479 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
480 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]);
481 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]);
482
483 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 4]);
484 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]);
485 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]);
486
487 PRIM_INT__p_s_d_s[n * 18 + 11] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
488 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]);
489
490 PRIM_INT__p_s_d_s[n * 18 + 12] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
491 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]);
492
493 PRIM_INT__p_s_d_s[n * 18 + 13] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
494 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]);
495
496 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 2]);
497 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]);
498 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]);
499
500 PRIM_INT__p_s_d_s[n * 18 + 15] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
501 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]);
502
503 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 4]);
504 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]);
505 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]);
506
507 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
508 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]);
509 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]);
510
511 }
512
513
514 VRR_I_d_s_f_s(
515 PRIM_INT__d_s_f_s,
516 PRIM_INT__p_s_f_s,
517 PRIM_INT__s_s_f_s,
518 PRIM_INT__p_s_d_s,
519 P_PA,
520 a_over_p,
521 aop_PQ,
522 one_over_2p,
523 one_over_2pq,
524 4);
525
526
527 ostei_general_vrr_I(3, 0, 4, 0, 3,
528 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
529 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);
530
531
532 VRR_K_s_s_h_s(
533 PRIM_INT__s_s_h_s,
534 PRIM_INT__s_s_g_s,
535 PRIM_INT__s_s_f_s,
536 Q_PA,
537 a_over_q,
538 aoq_PQ,
539 one_over_2q,
540 8);
541
542
543 ostei_general_vrr_I(1, 0, 5, 0, 5,
544 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
545 PRIM_INT__s_s_h_s, NULL, NULL, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_h_s);
546
547
548 ostei_general_vrr_I(2, 0, 5, 0, 4,
549 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
550 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);
551
552
553 ostei_general_vrr_I(3, 0, 5, 0, 3,
554 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
555 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);
556
557
558
559 // Forming PRIM_INT__p_s_p_s[5 * 9];
560 for(n = 0; n < 5; ++n) // loop over orders of auxiliary function
561 {
562
563 PRIM_INT__p_s_p_s[n * 9 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
564 PRIM_INT__p_s_p_s[n * 9 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 0]);
565 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]);
566
567 PRIM_INT__p_s_p_s[n * 9 + 1] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_p_s[n * 3 + 1]);
568 PRIM_INT__p_s_p_s[n * 9 + 1] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 1]);
569
570 PRIM_INT__p_s_p_s[n * 9 + 2] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_p_s[n * 3 + 2]);
571 PRIM_INT__p_s_p_s[n * 9 + 2] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 2]);
572
573 PRIM_INT__p_s_p_s[n * 9 + 3] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_p_s[n * 3 + 0]);
574 PRIM_INT__p_s_p_s[n * 9 + 3] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 3]);
575
576 PRIM_INT__p_s_p_s[n * 9 + 4] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_p_s[n * 3 + 1]);
577 PRIM_INT__p_s_p_s[n * 9 + 4] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 4]);
578 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]);
579
580 PRIM_INT__p_s_p_s[n * 9 + 5] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_p_s[n * 3 + 2]);
581 PRIM_INT__p_s_p_s[n * 9 + 5] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 5]);
582
583 PRIM_INT__p_s_p_s[n * 9 + 6] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_p_s[n * 3 + 0]);
584 PRIM_INT__p_s_p_s[n * 9 + 6] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_p_s[n * 9 + 6]);
585
586 PRIM_INT__p_s_p_s[n * 9 + 7] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_p_s[n * 3 + 1]);
587 PRIM_INT__p_s_p_s[n * 9 + 7] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_p_s[n * 9 + 7]);
588
589 PRIM_INT__p_s_p_s[n * 9 + 8] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
590 PRIM_INT__p_s_p_s[n * 9 + 8] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_p_s[n * 9 + 8]);
591 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]);
592
593 }
594
595
596 VRR_I_d_s_d_s(
597 PRIM_INT__d_s_d_s,
598 PRIM_INT__p_s_d_s,
599 PRIM_INT__s_s_d_s,
600 PRIM_INT__p_s_p_s,
601 P_PA,
602 a_over_p,
603 aop_PQ,
604 one_over_2p,
605 one_over_2pq,
606 4);
607
608
609 ostei_general_vrr_I(3, 0, 3, 0, 3,
610 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
611 PRIM_INT__d_s_f_s, PRIM_INT__p_s_f_s, NULL, PRIM_INT__d_s_d_s, NULL, PRIM_INT__f_s_f_s);
612
613
614 ostei_general_vrr_I(4, 0, 4, 0, 2,
615 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
616 PRIM_INT__f_s_g_s, PRIM_INT__d_s_g_s, NULL, PRIM_INT__f_s_f_s, NULL, PRIM_INT__g_s_g_s);
617
618
619 ostei_general_vrr1_K(6, 7,
620 one_over_2q, a_over_q, aoq_PQ, Q_PA,
621 PRIM_INT__s_s_h_s, PRIM_INT__s_s_g_s, PRIM_INT__s_s_i_s);
622
623
624 ostei_general_vrr_I(1, 0, 6, 0, 5,
625 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
626 PRIM_INT__s_s_i_s, NULL, NULL, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_i_s);
627
628
629 ostei_general_vrr_I(2, 0, 6, 0, 4,
630 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
631 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);
632
633
634 ostei_general_vrr_I(3, 0, 6, 0, 3,
635 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
636 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);
637
638
639 ostei_general_vrr_I(4, 0, 5, 0, 2,
640 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
641 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);
642
643
644
645 // Forming PRIM_INT__p_s_s_s[5 * 3];
646 for(n = 0; n < 5; ++n) // loop over orders of auxiliary function
647 {
648
649 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
650 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]);
651
652 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
653 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]);
654
655 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
656 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]);
657
658 }
659
660
661
662 // Forming PRIM_INT__d_s_p_s[4 * 18];
663 for(n = 0; n < 4; ++n) // loop over orders of auxiliary function
664 {
665
666 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_p_s[n * 9 + 0]);
667 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__p_s_p_s[(n+1) * 9 + 0], PRIM_INT__d_s_p_s[n * 18 + 0]);
668 PRIM_INT__d_s_p_s[n * 18 + 0] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_p_s[n * 3 + 0]), PRIM_INT__d_s_p_s[n * 18 + 0]);
669 PRIM_INT__d_s_p_s[n * 18 + 0] = 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 + 0]);
670
671 PRIM_INT__d_s_p_s[n * 18 + 1] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_p_s[n * 9 + 1]);
672 PRIM_INT__d_s_p_s[n * 18 + 1] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__p_s_p_s[(n+1) * 9 + 1], PRIM_INT__d_s_p_s[n * 18 + 1]);
673 PRIM_INT__d_s_p_s[n * 18 + 1] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_p_s[n * 3 + 1]), PRIM_INT__d_s_p_s[n * 18 + 1]);
674
675 PRIM_INT__d_s_p_s[n * 18 + 2] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_p_s[n * 9 + 2]);
676 PRIM_INT__d_s_p_s[n * 18 + 2] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__p_s_p_s[(n+1) * 9 + 2], PRIM_INT__d_s_p_s[n * 18 + 2]);
677 PRIM_INT__d_s_p_s[n * 18 + 2] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__s_s_p_s[n * 3 + 2]), PRIM_INT__d_s_p_s[n * 18 + 2]);
678
679 PRIM_INT__d_s_p_s[n * 18 + 9] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_p_s[n * 9 + 3]);
680 PRIM_INT__d_s_p_s[n * 18 + 9] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__p_s_p_s[(n+1) * 9 + 3], PRIM_INT__d_s_p_s[n * 18 + 9]);
681 PRIM_INT__d_s_p_s[n * 18 + 9] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_p_s[n * 3 + 0]), PRIM_INT__d_s_p_s[n * 18 + 9]);
682
683 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_p_s[n * 9 + 4]);
684 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__p_s_p_s[(n+1) * 9 + 4], PRIM_INT__d_s_p_s[n * 18 + 10]);
685 PRIM_INT__d_s_p_s[n * 18 + 10] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_p_s[n * 3 + 1]), PRIM_INT__d_s_p_s[n * 18 + 10]);
686 PRIM_INT__d_s_p_s[n * 18 + 10] = 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 + 10]);
687
688 PRIM_INT__d_s_p_s[n * 18 + 11] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_p_s[n * 9 + 5]);
689 PRIM_INT__d_s_p_s[n * 18 + 11] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__p_s_p_s[(n+1) * 9 + 5], PRIM_INT__d_s_p_s[n * 18 + 11]);
690 PRIM_INT__d_s_p_s[n * 18 + 11] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__s_s_p_s[n * 3 + 2]), PRIM_INT__d_s_p_s[n * 18 + 11]);
691
692 PRIM_INT__d_s_p_s[n * 18 + 15] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_p_s[n * 9 + 6]);
693 PRIM_INT__d_s_p_s[n * 18 + 15] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_p_s[(n+1) * 9 + 6], PRIM_INT__d_s_p_s[n * 18 + 15]);
694 PRIM_INT__d_s_p_s[n * 18 + 15] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_p_s[n * 3 + 0]), PRIM_INT__d_s_p_s[n * 18 + 15]);
695
696 PRIM_INT__d_s_p_s[n * 18 + 16] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_p_s[n * 9 + 7]);
697 PRIM_INT__d_s_p_s[n * 18 + 16] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_p_s[(n+1) * 9 + 7], PRIM_INT__d_s_p_s[n * 18 + 16]);
698 PRIM_INT__d_s_p_s[n * 18 + 16] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_p_s[n * 3 + 1]), PRIM_INT__d_s_p_s[n * 18 + 16]);
699
700 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_p_s[n * 9 + 8]);
701 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__p_s_p_s[(n+1) * 9 + 8], PRIM_INT__d_s_p_s[n * 18 + 17]);
702 PRIM_INT__d_s_p_s[n * 18 + 17] = SIMINT_FMADD( vrr_const_1_over_2p, SIMINT_FMADD(a_over_p, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__s_s_p_s[n * 3 + 2]), PRIM_INT__d_s_p_s[n * 18 + 17]);
703 PRIM_INT__d_s_p_s[n * 18 + 17] = 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 + 17]);
704
705 }
706
707
708 VRR_I_f_s_d_s(
709 PRIM_INT__f_s_d_s,
710 PRIM_INT__d_s_d_s,
711 PRIM_INT__p_s_d_s,
712 PRIM_INT__d_s_p_s,
713 P_PA,
714 a_over_p,
715 aop_PQ,
716 one_over_2p,
717 one_over_2pq,
718 3);
719
720
721 ostei_general_vrr_I(4, 0, 3, 0, 2,
722 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
723 PRIM_INT__f_s_f_s, PRIM_INT__d_s_f_s, NULL, PRIM_INT__f_s_d_s, NULL, PRIM_INT__g_s_f_s);
724
725
726 ostei_general_vrr_I(5, 0, 4, 0, 1,
727 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
728 PRIM_INT__g_s_g_s, PRIM_INT__f_s_g_s, NULL, PRIM_INT__g_s_f_s, NULL, PRIM_INT__h_s_g_s);
729
730
731 ostei_general_vrr1_K(7, 6,
732 one_over_2q, a_over_q, aoq_PQ, Q_PA,
733 PRIM_INT__s_s_i_s, PRIM_INT__s_s_h_s, PRIM_INT__s_s_k_s);
734
735
736 ostei_general_vrr_I(1, 0, 7, 0, 5,
737 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
738 PRIM_INT__s_s_k_s, NULL, NULL, PRIM_INT__s_s_i_s, NULL, PRIM_INT__p_s_k_s);
739
740
741 ostei_general_vrr_I(2, 0, 7, 0, 4,
742 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
743 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);
744
745
746 ostei_general_vrr_I(3, 0, 7, 0, 3,
747 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
748 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);
749
750
751 ostei_general_vrr_I(4, 0, 6, 0, 2,
752 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
753 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);
754
755
756 ostei_general_vrr_I(5, 0, 5, 0, 1,
757 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
758 PRIM_INT__g_s_h_s, PRIM_INT__f_s_h_s, NULL, PRIM_INT__g_s_g_s, NULL, PRIM_INT__h_s_h_s);
759
760
761 ostei_general_vrr_I(4, 0, 7, 0, 2,
762 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
763 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);
764
765
766 ostei_general_vrr_I(5, 0, 6, 0, 1,
767 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
768 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);
769
770
771 ostei_general_vrr_I(5, 0, 7, 0, 1,
772 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
773 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);
774
775
776
777
778 ////////////////////////////////////
779 // Accumulate contracted integrals
780 ////////////////////////////////////
781 if(lastoffset == 0)
782 {
783 contract_all(150, PRIM_INT__f_s_g_s, PRIM_PTR_INT__f_s_g_s);
784 contract_all(210, PRIM_INT__f_s_h_s, PRIM_PTR_INT__f_s_h_s);
785 contract_all(280, PRIM_INT__f_s_i_s, PRIM_PTR_INT__f_s_i_s);
786 contract_all(360, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
787 contract_all(225, PRIM_INT__g_s_g_s, PRIM_PTR_INT__g_s_g_s);
788 contract_all(315, PRIM_INT__g_s_h_s, PRIM_PTR_INT__g_s_h_s);
789 contract_all(420, PRIM_INT__g_s_i_s, PRIM_PTR_INT__g_s_i_s);
790 contract_all(540, PRIM_INT__g_s_k_s, PRIM_PTR_INT__g_s_k_s);
791 contract_all(315, PRIM_INT__h_s_g_s, PRIM_PTR_INT__h_s_g_s);
792 contract_all(441, PRIM_INT__h_s_h_s, PRIM_PTR_INT__h_s_h_s);
793 contract_all(588, PRIM_INT__h_s_i_s, PRIM_PTR_INT__h_s_i_s);
794 contract_all(756, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
795 }
796 else
797 {
798 contract(150, shelloffsets, PRIM_INT__f_s_g_s, PRIM_PTR_INT__f_s_g_s);
799 contract(210, shelloffsets, PRIM_INT__f_s_h_s, PRIM_PTR_INT__f_s_h_s);
800 contract(280, shelloffsets, PRIM_INT__f_s_i_s, PRIM_PTR_INT__f_s_i_s);
801 contract(360, shelloffsets, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
802 contract(225, shelloffsets, PRIM_INT__g_s_g_s, PRIM_PTR_INT__g_s_g_s);
803 contract(315, shelloffsets, PRIM_INT__g_s_h_s, PRIM_PTR_INT__g_s_h_s);
804 contract(420, shelloffsets, PRIM_INT__g_s_i_s, PRIM_PTR_INT__g_s_i_s);
805 contract(540, shelloffsets, PRIM_INT__g_s_k_s, PRIM_PTR_INT__g_s_k_s);
806 contract(315, shelloffsets, PRIM_INT__h_s_g_s, PRIM_PTR_INT__h_s_g_s);
807 contract(441, shelloffsets, PRIM_INT__h_s_h_s, PRIM_PTR_INT__h_s_h_s);
808 contract(588, shelloffsets, PRIM_INT__h_s_i_s, PRIM_PTR_INT__h_s_i_s);
809 contract(756, shelloffsets, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
810 PRIM_PTR_INT__f_s_g_s += lastoffset*150;
811 PRIM_PTR_INT__f_s_h_s += lastoffset*210;
812 PRIM_PTR_INT__f_s_i_s += lastoffset*280;
813 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
814 PRIM_PTR_INT__g_s_g_s += lastoffset*225;
815 PRIM_PTR_INT__g_s_h_s += lastoffset*315;
816 PRIM_PTR_INT__g_s_i_s += lastoffset*420;
817 PRIM_PTR_INT__g_s_k_s += lastoffset*540;
818 PRIM_PTR_INT__h_s_g_s += lastoffset*315;
819 PRIM_PTR_INT__h_s_h_s += lastoffset*441;
820 PRIM_PTR_INT__h_s_i_s += lastoffset*588;
821 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
822 }
823
824 } // close loop over j
825 } // close loop over i
826
827 //Advance to the next batch
828 jstart = SIMINT_SIMD_ROUND(jend);
829
830 //////////////////////////////////////////////
831 // Contracted integrals: Horizontal recurrance
832 //////////////////////////////////////////////
833
834
835 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
836
837
838 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
839 {
840 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
841
842 // set up HRR pointers
843 double const * restrict HRR_INT__f_s_g_s = INT__f_s_g_s + abcd * 150;
844 double const * restrict HRR_INT__f_s_h_s = INT__f_s_h_s + abcd * 210;
845 double const * restrict HRR_INT__f_s_i_s = INT__f_s_i_s + abcd * 280;
846 double const * restrict HRR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
847 double const * restrict HRR_INT__g_s_g_s = INT__g_s_g_s + abcd * 225;
848 double const * restrict HRR_INT__g_s_h_s = INT__g_s_h_s + abcd * 315;
849 double const * restrict HRR_INT__g_s_i_s = INT__g_s_i_s + abcd * 420;
850 double const * restrict HRR_INT__g_s_k_s = INT__g_s_k_s + abcd * 540;
851 double const * restrict HRR_INT__h_s_g_s = INT__h_s_g_s + abcd * 315;
852 double const * restrict HRR_INT__h_s_h_s = INT__h_s_h_s + abcd * 441;
853 double const * restrict HRR_INT__h_s_i_s = INT__h_s_i_s + abcd * 588;
854 double const * restrict HRR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
855 double * restrict HRR_INT__f_d_g_f = INT__f_d_g_f + real_abcd * 9000;
856
857 // form INT__f_p_g_s
858 HRR_J_f_p(
859 HRR_INT__f_p_g_s,
860 HRR_INT__f_s_g_s,
861 HRR_INT__g_s_g_s,
862 hAB, 15);
863
864 // form INT__f_p_h_s
865 HRR_J_f_p(
866 HRR_INT__f_p_h_s,
867 HRR_INT__f_s_h_s,
868 HRR_INT__g_s_h_s,
869 hAB, 21);
870
871 // form INT__f_p_i_s
872 HRR_J_f_p(
873 HRR_INT__f_p_i_s,
874 HRR_INT__f_s_i_s,
875 HRR_INT__g_s_i_s,
876 hAB, 28);
877
878 // form INT__f_p_k_s
879 HRR_J_f_p(
880 HRR_INT__f_p_k_s,
881 HRR_INT__f_s_k_s,
882 HRR_INT__g_s_k_s,
883 hAB, 36);
884
885 // form INT__g_p_g_s
886 HRR_J_g_p(
887 HRR_INT__g_p_g_s,
888 HRR_INT__g_s_g_s,
889 HRR_INT__h_s_g_s,
890 hAB, 15);
891
892 // form INT__g_p_h_s
893 HRR_J_g_p(
894 HRR_INT__g_p_h_s,
895 HRR_INT__g_s_h_s,
896 HRR_INT__h_s_h_s,
897 hAB, 21);
898
899 // form INT__g_p_i_s
900 HRR_J_g_p(
901 HRR_INT__g_p_i_s,
902 HRR_INT__g_s_i_s,
903 HRR_INT__h_s_i_s,
904 hAB, 28);
905
906 // form INT__g_p_k_s
907 HRR_J_g_p(
908 HRR_INT__g_p_k_s,
909 HRR_INT__g_s_k_s,
910 HRR_INT__h_s_k_s,
911 hAB, 36);
912
913 // form INT__f_d_g_s
914 HRR_J_f_d(
915 HRR_INT__f_d_g_s,
916 HRR_INT__f_p_g_s,
917 HRR_INT__g_p_g_s,
918 hAB, 15);
919
920 // form INT__f_d_h_s
921 HRR_J_f_d(
922 HRR_INT__f_d_h_s,
923 HRR_INT__f_p_h_s,
924 HRR_INT__g_p_h_s,
925 hAB, 21);
926
927 // form INT__f_d_i_s
928 HRR_J_f_d(
929 HRR_INT__f_d_i_s,
930 HRR_INT__f_p_i_s,
931 HRR_INT__g_p_i_s,
932 hAB, 28);
933
934 // form INT__f_d_k_s
935 HRR_J_f_d(
936 HRR_INT__f_d_k_s,
937 HRR_INT__f_p_k_s,
938 HRR_INT__g_p_k_s,
939 hAB, 36);
940
941 // form INT__f_d_g_p
942 HRR_L_g_p(
943 HRR_INT__f_d_g_p,
944 HRR_INT__f_d_g_s,
945 HRR_INT__f_d_h_s,
946 hCD, 60);
947
948 // form INT__f_d_h_p
949 ostei_general_hrr_L(3, 2, 5, 1, hCD, HRR_INT__f_d_i_s, HRR_INT__f_d_h_s, HRR_INT__f_d_h_p);
950
951 // form INT__f_d_i_p
952 ostei_general_hrr_L(3, 2, 6, 1, hCD, HRR_INT__f_d_k_s, HRR_INT__f_d_i_s, HRR_INT__f_d_i_p);
953
954 // form INT__f_d_g_d
955 ostei_general_hrr_L(3, 2, 4, 2, hCD, HRR_INT__f_d_h_p, HRR_INT__f_d_g_p, HRR_INT__f_d_g_d);
956
957 // form INT__f_d_h_d
958 ostei_general_hrr_L(3, 2, 5, 2, hCD, HRR_INT__f_d_i_p, HRR_INT__f_d_h_p, HRR_INT__f_d_h_d);
959
960 // form INT__f_d_g_f
961 ostei_general_hrr_L(3, 2, 4, 3, hCD, HRR_INT__f_d_h_d, HRR_INT__f_d_g_d, HRR_INT__f_d_g_f);
962
963
964 } // close HRR loop
965
966
967 } // close loop cdbatch
968
969 istart = iend;
970 } // close loop over ab
971
972 return P.nshell12_clip * Q.nshell12_clip;
973 }
974
ostei_d_f_g_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_f_g_f)975 int ostei_d_f_g_f(struct simint_multi_shellpair const P,
976 struct simint_multi_shellpair const Q,
977 double screen_tol,
978 double * const restrict work,
979 double * const restrict INT__d_f_g_f)
980 {
981 double P_AB[3*P.nshell12];
982 struct simint_multi_shellpair P_tmp = P;
983 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
984 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
985 P_tmp.AB_x = P_AB;
986 P_tmp.AB_y = P_AB + P.nshell12;
987 P_tmp.AB_z = P_AB + 2*P.nshell12;
988
989 for(int i = 0; i < P.nshell12; i++)
990 {
991 P_tmp.AB_x[i] = -P.AB_x[i];
992 P_tmp.AB_y[i] = -P.AB_y[i];
993 P_tmp.AB_z[i] = -P.AB_z[i];
994 }
995
996 int ret = ostei_f_d_g_f(P_tmp, Q, screen_tol, work, INT__d_f_g_f);
997 double buffer[9000] SIMINT_ALIGN_ARRAY_DBL;
998
999 for(int q = 0; q < ret; q++)
1000 {
1001 int idx = 0;
1002 for(int a = 0; a < 6; ++a)
1003 for(int b = 0; b < 10; ++b)
1004 for(int c = 0; c < 15; ++c)
1005 for(int d = 0; d < 10; ++d)
1006 buffer[idx++] = INT__d_f_g_f[q*9000+b*900+a*150+c*10+d];
1007
1008 memcpy(INT__d_f_g_f+q*9000, buffer, 9000*sizeof(double));
1009 }
1010
1011 return ret;
1012 }
1013
ostei_f_d_f_g(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__f_d_f_g)1014 int ostei_f_d_f_g(struct simint_multi_shellpair const P,
1015 struct simint_multi_shellpair const Q,
1016 double screen_tol,
1017 double * const restrict work,
1018 double * const restrict INT__f_d_f_g)
1019 {
1020 double Q_AB[3*Q.nshell12];
1021 struct simint_multi_shellpair Q_tmp = Q;
1022 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
1023 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
1024 Q_tmp.AB_x = Q_AB;
1025 Q_tmp.AB_y = Q_AB + Q.nshell12;
1026 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
1027
1028 for(int i = 0; i < Q.nshell12; i++)
1029 {
1030 Q_tmp.AB_x[i] = -Q.AB_x[i];
1031 Q_tmp.AB_y[i] = -Q.AB_y[i];
1032 Q_tmp.AB_z[i] = -Q.AB_z[i];
1033 }
1034
1035 int ret = ostei_f_d_g_f(P, Q_tmp, screen_tol, work, INT__f_d_f_g);
1036 double buffer[9000] SIMINT_ALIGN_ARRAY_DBL;
1037
1038 for(int q = 0; q < ret; q++)
1039 {
1040 int idx = 0;
1041 for(int a = 0; a < 10; ++a)
1042 for(int b = 0; b < 6; ++b)
1043 for(int c = 0; c < 10; ++c)
1044 for(int d = 0; d < 15; ++d)
1045 buffer[idx++] = INT__f_d_f_g[q*9000+a*900+b*150+d*10+c];
1046
1047 memcpy(INT__f_d_f_g+q*9000, buffer, 9000*sizeof(double));
1048 }
1049
1050 return ret;
1051 }
1052
ostei_d_f_f_g(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_f_f_g)1053 int ostei_d_f_f_g(struct simint_multi_shellpair const P,
1054 struct simint_multi_shellpair const Q,
1055 double screen_tol,
1056 double * const restrict work,
1057 double * const restrict INT__d_f_f_g)
1058 {
1059 double P_AB[3*P.nshell12];
1060 struct simint_multi_shellpair P_tmp = P;
1061 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
1062 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
1063 P_tmp.AB_x = P_AB;
1064 P_tmp.AB_y = P_AB + P.nshell12;
1065 P_tmp.AB_z = P_AB + 2*P.nshell12;
1066
1067 for(int i = 0; i < P.nshell12; i++)
1068 {
1069 P_tmp.AB_x[i] = -P.AB_x[i];
1070 P_tmp.AB_y[i] = -P.AB_y[i];
1071 P_tmp.AB_z[i] = -P.AB_z[i];
1072 }
1073
1074 double Q_AB[3*Q.nshell12];
1075 struct simint_multi_shellpair Q_tmp = Q;
1076 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
1077 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
1078 Q_tmp.AB_x = Q_AB;
1079 Q_tmp.AB_y = Q_AB + Q.nshell12;
1080 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
1081
1082 for(int i = 0; i < Q.nshell12; i++)
1083 {
1084 Q_tmp.AB_x[i] = -Q.AB_x[i];
1085 Q_tmp.AB_y[i] = -Q.AB_y[i];
1086 Q_tmp.AB_z[i] = -Q.AB_z[i];
1087 }
1088
1089 int ret = ostei_f_d_g_f(P_tmp, Q_tmp, screen_tol, work, INT__d_f_f_g);
1090 double buffer[9000] SIMINT_ALIGN_ARRAY_DBL;
1091
1092 for(int q = 0; q < ret; q++)
1093 {
1094 int idx = 0;
1095 for(int a = 0; a < 6; ++a)
1096 for(int b = 0; b < 10; ++b)
1097 for(int c = 0; c < 10; ++c)
1098 for(int d = 0; d < 15; ++d)
1099 buffer[idx++] = INT__d_f_f_g[q*9000+b*900+a*150+d*10+c];
1100
1101 memcpy(INT__d_f_f_g+q*9000, buffer, 9000*sizeof(double));
1102 }
1103
1104 return ret;
1105 }
1106
1107