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_i_g_f_p(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__i_g_f_p)8 int ostei_i_g_f_p(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__i_g_f_p)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__i_g_f_p);
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__i_s_f_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__i_s_g_s = work + (SIMINT_NSHELL_SIMD * 280);
31 double * const INT__k_s_f_s = work + (SIMINT_NSHELL_SIMD * 700);
32 double * const INT__k_s_g_s = work + (SIMINT_NSHELL_SIMD * 1060);
33 double * const INT__l_s_f_s = work + (SIMINT_NSHELL_SIMD * 1600);
34 double * const INT__l_s_g_s = work + (SIMINT_NSHELL_SIMD * 2050);
35 double * const INT__m_s_f_s = work + (SIMINT_NSHELL_SIMD * 2725);
36 double * const INT__m_s_g_s = work + (SIMINT_NSHELL_SIMD * 3275);
37 double * const INT__n_s_f_s = work + (SIMINT_NSHELL_SIMD * 4100);
38 double * const INT__n_s_g_s = work + (SIMINT_NSHELL_SIMD * 4760);
39 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*5750);
40 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
41 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_s_s = primwork + 15;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_s_s = primwork + 57;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_s_s = primwork + 135;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_p_s = primwork + 255;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_s_s = primwork + 375;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_p_s = primwork + 540;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_d_s = primwork + 720;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_s_s = primwork + 990;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_p_s = primwork + 1200;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_d_s = primwork + 1452;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_f_s = primwork + 1830;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_s_s = primwork + 2250;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_p_s = primwork + 2502;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_d_s = primwork + 2838;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_f_s = primwork + 3342;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_g_s = primwork + 3902;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_s_s = primwork + 4322;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_p_s = primwork + 4610;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_d_s = primwork + 5042;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_f_s = primwork + 5690;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_g_s = primwork + 6410;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_s_s = primwork + 6950;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_p_s = primwork + 7265;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_d_s = primwork + 7805;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_f_s = primwork + 8615;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_g_s = primwork + 9515;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_s_s = primwork + 10190;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_p_s = primwork + 10520;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_d_s = primwork + 11180;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_f_s = primwork + 12170;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_g_s = primwork + 13270;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_s_s = primwork + 14095;
73 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_p_s = primwork + 14425;
74 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_d_s = primwork + 15217;
75 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_f_s = primwork + 16405;
76 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_g_s = primwork + 17725;
77 double * const hrrwork = (double *)(primwork + 18715);
78 double * const HRR_INT__i_p_f_s = hrrwork + 0;
79 double * const HRR_INT__i_p_g_s = hrrwork + 840;
80 double * const HRR_INT__i_d_f_s = hrrwork + 2100;
81 double * const HRR_INT__i_d_g_s = hrrwork + 3780;
82 double * const HRR_INT__i_f_f_s = hrrwork + 6300;
83 double * const HRR_INT__i_f_g_s = hrrwork + 9100;
84 double * const HRR_INT__i_g_f_s = hrrwork + 13300;
85 double * const HRR_INT__i_g_g_s = hrrwork + 17500;
86 double * const HRR_INT__k_p_f_s = hrrwork + 23800;
87 double * const HRR_INT__k_p_g_s = hrrwork + 24880;
88 double * const HRR_INT__k_d_f_s = hrrwork + 26500;
89 double * const HRR_INT__k_d_g_s = hrrwork + 28660;
90 double * const HRR_INT__k_f_f_s = hrrwork + 31900;
91 double * const HRR_INT__k_f_g_s = hrrwork + 35500;
92 double * const HRR_INT__l_p_f_s = hrrwork + 40900;
93 double * const HRR_INT__l_p_g_s = hrrwork + 42250;
94 double * const HRR_INT__l_d_f_s = hrrwork + 44275;
95 double * const HRR_INT__l_d_g_s = hrrwork + 46975;
96 double * const HRR_INT__m_p_f_s = hrrwork + 51025;
97 double * const HRR_INT__m_p_g_s = hrrwork + 52675;
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 * 5750 * 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__i_s_f_s = INT__i_s_f_s + abcd * 280;
155 double * restrict PRIM_PTR_INT__i_s_g_s = INT__i_s_g_s + abcd * 420;
156 double * restrict PRIM_PTR_INT__k_s_f_s = INT__k_s_f_s + abcd * 360;
157 double * restrict PRIM_PTR_INT__k_s_g_s = INT__k_s_g_s + abcd * 540;
158 double * restrict PRIM_PTR_INT__l_s_f_s = INT__l_s_f_s + abcd * 450;
159 double * restrict PRIM_PTR_INT__l_s_g_s = INT__l_s_g_s + abcd * 675;
160 double * restrict PRIM_PTR_INT__m_s_f_s = INT__m_s_f_s + abcd * 550;
161 double * restrict PRIM_PTR_INT__m_s_g_s = INT__m_s_g_s + abcd * 825;
162 double * restrict PRIM_PTR_INT__n_s_f_s = INT__n_s_f_s + abcd * 660;
163 double * restrict PRIM_PTR_INT__n_s_g_s = INT__n_s_g_s + abcd * 990;
164
165
166
167 // Load these one per loop over i
168 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
169 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
170 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
171
172 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
173
174 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
175 {
176 // calculate the shell offsets
177 // these are the offset from the shell pointed to by cd
178 // for each element
179 int shelloffsets[SIMINT_SIMD_LEN] = {0};
180 int lastoffset = 0;
181 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
182
183 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
184 {
185 // Handle if the first element of the vector is a new shell
186 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
187 {
188 nprim_icd += Q.nprim12[cd + (++icd)];
189 PRIM_PTR_INT__i_s_f_s += 280;
190 PRIM_PTR_INT__i_s_g_s += 420;
191 PRIM_PTR_INT__k_s_f_s += 360;
192 PRIM_PTR_INT__k_s_g_s += 540;
193 PRIM_PTR_INT__l_s_f_s += 450;
194 PRIM_PTR_INT__l_s_g_s += 675;
195 PRIM_PTR_INT__m_s_f_s += 550;
196 PRIM_PTR_INT__m_s_g_s += 825;
197 PRIM_PTR_INT__n_s_f_s += 660;
198 PRIM_PTR_INT__n_s_g_s += 990;
199 }
200 iprimcd++;
201 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
202 {
203 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
204 {
205 shelloffsets[n] = shelloffsets[n-1] + 1;
206 lastoffset++;
207 nprim_icd += Q.nprim12[cd + (++icd)];
208 }
209 else
210 shelloffsets[n] = shelloffsets[n-1];
211 iprimcd++;
212 }
213 }
214 else
215 iprimcd += SIMINT_SIMD_LEN;
216
217 // Do we have to compute this vector (or has it been screened out)?
218 // (not_screened != 0 means we have to do this vector)
219 if(check_screen)
220 {
221 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
222 if(vmax < screen_tol)
223 {
224 PRIM_PTR_INT__i_s_f_s += lastoffset*280;
225 PRIM_PTR_INT__i_s_g_s += lastoffset*420;
226 PRIM_PTR_INT__k_s_f_s += lastoffset*360;
227 PRIM_PTR_INT__k_s_g_s += lastoffset*540;
228 PRIM_PTR_INT__l_s_f_s += lastoffset*450;
229 PRIM_PTR_INT__l_s_g_s += lastoffset*675;
230 PRIM_PTR_INT__m_s_f_s += lastoffset*550;
231 PRIM_PTR_INT__m_s_g_s += lastoffset*825;
232 PRIM_PTR_INT__n_s_f_s += lastoffset*660;
233 PRIM_PTR_INT__n_s_g_s += lastoffset*990;
234 continue;
235 }
236 }
237
238 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
239 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
240 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
241 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
242
243
244 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
245 SIMINT_DBLTYPE PQ[3];
246 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
247 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
248 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
249 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
250 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
251 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
252
253 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
254 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
255 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
256 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
257 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
258 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
259 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
260
261 // NOTE: Minus sign!
262 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
263 SIMINT_DBLTYPE aop_PQ[3];
264 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
265 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
266 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
267
268 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
269 SIMINT_DBLTYPE aoq_PQ[3];
270 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
271 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
272 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
273 // Put a minus sign here so we don't have to in RR routines
274 a_over_q = SIMINT_NEG(a_over_q);
275
276
277 //////////////////////////////////////////////
278 // Fjt function section
279 // Maximum v value: 14
280 //////////////////////////////////////////////
281 // The parameter to the Fjt function
282 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
283
284
285 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
286
287
288 boys_F_split(PRIM_INT__s_s_s_s, F_x, 14);
289 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
290 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
291 for(n = 0; n <= 14; n++)
292 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
293
294 //////////////////////////////////////////////
295 // Primitive integrals: Vertical recurrance
296 //////////////////////////////////////////////
297
298 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
299 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
300 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
301 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
302 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
303 const SIMINT_DBLTYPE vrr_const_6_over_2p = SIMINT_MUL(const_6, one_over_2p);
304 const SIMINT_DBLTYPE vrr_const_7_over_2p = SIMINT_MUL(const_7, one_over_2p);
305 const SIMINT_DBLTYPE vrr_const_8_over_2p = SIMINT_MUL(const_8, one_over_2p);
306 const SIMINT_DBLTYPE vrr_const_9_over_2p = SIMINT_MUL(const_9, one_over_2p);
307 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
308 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
309 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
310 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
311 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
312 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
313 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
314 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
315 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
316 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
317 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
318 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
319 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
320
321
322
323 // Forming PRIM_INT__p_s_s_s[14 * 3];
324 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
325 {
326
327 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
328 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]);
329
330 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
331 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]);
332
333 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
334 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]);
335
336 }
337
338
339
340 // Forming PRIM_INT__d_s_s_s[13 * 6];
341 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
342 {
343
344 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_s_s[n * 3 + 0]);
345 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]);
346 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]);
347
348 PRIM_INT__d_s_s_s[n * 6 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 0]);
349 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]);
350
351 PRIM_INT__d_s_s_s[n * 6 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 0]);
352 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]);
353
354 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 1]);
355 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]);
356 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]);
357
358 PRIM_INT__d_s_s_s[n * 6 + 4] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 1]);
359 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]);
360
361 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 2]);
362 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]);
363 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]);
364
365 }
366
367
368
369 // Forming PRIM_INT__f_s_s_s[12 * 10];
370 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
371 {
372
373 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 0]);
374 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]);
375 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]);
376
377 PRIM_INT__f_s_s_s[n * 10 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 0]);
378 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]);
379
380 PRIM_INT__f_s_s_s[n * 10 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 0]);
381 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]);
382
383 PRIM_INT__f_s_s_s[n * 10 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 3]);
384 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]);
385
386 PRIM_INT__f_s_s_s[n * 10 + 4] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 1]);
387 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]);
388
389 PRIM_INT__f_s_s_s[n * 10 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 5]);
390 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]);
391
392 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 3]);
393 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]);
394 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]);
395
396 PRIM_INT__f_s_s_s[n * 10 + 7] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 3]);
397 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]);
398
399 PRIM_INT__f_s_s_s[n * 10 + 8] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 5]);
400 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]);
401
402 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 5]);
403 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]);
404 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]);
405
406 }
407
408
409 VRR_I_g_s_s_s(
410 PRIM_INT__g_s_s_s,
411 PRIM_INT__f_s_s_s,
412 PRIM_INT__d_s_s_s,
413 P_PA,
414 a_over_p,
415 aop_PQ,
416 one_over_2p,
417 11);
418
419
420 VRR_I_h_s_s_s(
421 PRIM_INT__h_s_s_s,
422 PRIM_INT__g_s_s_s,
423 PRIM_INT__f_s_s_s,
424 P_PA,
425 a_over_p,
426 aop_PQ,
427 one_over_2p,
428 10);
429
430
431 ostei_general_vrr1_I(6, 9,
432 one_over_2p, a_over_p, aop_PQ, P_PA,
433 PRIM_INT__h_s_s_s, PRIM_INT__g_s_s_s, PRIM_INT__i_s_s_s);
434
435
436 ostei_general_vrr_K(6, 0, 1, 0, 4,
437 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
438 PRIM_INT__i_s_s_s, NULL, NULL, PRIM_INT__h_s_s_s, NULL, PRIM_INT__i_s_p_s);
439
440
441 ostei_general_vrr_K(5, 0, 1, 0, 4,
442 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
443 PRIM_INT__h_s_s_s, NULL, NULL, PRIM_INT__g_s_s_s, NULL, PRIM_INT__h_s_p_s);
444
445
446 ostei_general_vrr_K(6, 0, 2, 0, 3,
447 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
448 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);
449
450
451 VRR_K_g_s_p_s(
452 PRIM_INT__g_s_p_s,
453 PRIM_INT__g_s_s_s,
454 PRIM_INT__f_s_s_s,
455 Q_PA,
456 aoq_PQ,
457 one_over_2pq,
458 4);
459
460
461 ostei_general_vrr_K(5, 0, 2, 0, 3,
462 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
463 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);
464
465
466 ostei_general_vrr_K(6, 0, 3, 0, 2,
467 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
468 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);
469
470
471 VRR_K_f_s_p_s(
472 PRIM_INT__f_s_p_s,
473 PRIM_INT__f_s_s_s,
474 PRIM_INT__d_s_s_s,
475 Q_PA,
476 aoq_PQ,
477 one_over_2pq,
478 4);
479
480
481 ostei_general_vrr_K(4, 0, 2, 0, 3,
482 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
483 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);
484
485
486 ostei_general_vrr_K(5, 0, 3, 0, 2,
487 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
488 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);
489
490
491 ostei_general_vrr_K(6, 0, 4, 0, 1,
492 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
493 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);
494
495
496 ostei_general_vrr1_I(7, 8,
497 one_over_2p, a_over_p, aop_PQ, P_PA,
498 PRIM_INT__i_s_s_s, PRIM_INT__h_s_s_s, PRIM_INT__k_s_s_s);
499
500
501 ostei_general_vrr_K(7, 0, 1, 0, 4,
502 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
503 PRIM_INT__k_s_s_s, NULL, NULL, PRIM_INT__i_s_s_s, NULL, PRIM_INT__k_s_p_s);
504
505
506 ostei_general_vrr_K(7, 0, 2, 0, 3,
507 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
508 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);
509
510
511 ostei_general_vrr_K(7, 0, 3, 0, 2,
512 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
513 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);
514
515
516 ostei_general_vrr_K(7, 0, 4, 0, 1,
517 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
518 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);
519
520
521 ostei_general_vrr1_I(8, 7,
522 one_over_2p, a_over_p, aop_PQ, P_PA,
523 PRIM_INT__k_s_s_s, PRIM_INT__i_s_s_s, PRIM_INT__l_s_s_s);
524
525
526 ostei_general_vrr_K(8, 0, 1, 0, 4,
527 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
528 PRIM_INT__l_s_s_s, NULL, NULL, PRIM_INT__k_s_s_s, NULL, PRIM_INT__l_s_p_s);
529
530
531 ostei_general_vrr_K(8, 0, 2, 0, 3,
532 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
533 PRIM_INT__l_s_p_s, PRIM_INT__l_s_s_s, NULL, PRIM_INT__k_s_p_s, NULL, PRIM_INT__l_s_d_s);
534
535
536 ostei_general_vrr_K(8, 0, 3, 0, 2,
537 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
538 PRIM_INT__l_s_d_s, PRIM_INT__l_s_p_s, NULL, PRIM_INT__k_s_d_s, NULL, PRIM_INT__l_s_f_s);
539
540
541 ostei_general_vrr_K(8, 0, 4, 0, 1,
542 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
543 PRIM_INT__l_s_f_s, PRIM_INT__l_s_d_s, NULL, PRIM_INT__k_s_f_s, NULL, PRIM_INT__l_s_g_s);
544
545
546 ostei_general_vrr1_I(9, 6,
547 one_over_2p, a_over_p, aop_PQ, P_PA,
548 PRIM_INT__l_s_s_s, PRIM_INT__k_s_s_s, PRIM_INT__m_s_s_s);
549
550
551 ostei_general_vrr_K(9, 0, 1, 0, 4,
552 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
553 PRIM_INT__m_s_s_s, NULL, NULL, PRIM_INT__l_s_s_s, NULL, PRIM_INT__m_s_p_s);
554
555
556 ostei_general_vrr_K(9, 0, 2, 0, 3,
557 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
558 PRIM_INT__m_s_p_s, PRIM_INT__m_s_s_s, NULL, PRIM_INT__l_s_p_s, NULL, PRIM_INT__m_s_d_s);
559
560
561 ostei_general_vrr_K(9, 0, 3, 0, 2,
562 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
563 PRIM_INT__m_s_d_s, PRIM_INT__m_s_p_s, NULL, PRIM_INT__l_s_d_s, NULL, PRIM_INT__m_s_f_s);
564
565
566 ostei_general_vrr_K(9, 0, 4, 0, 1,
567 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
568 PRIM_INT__m_s_f_s, PRIM_INT__m_s_d_s, NULL, PRIM_INT__l_s_f_s, NULL, PRIM_INT__m_s_g_s);
569
570
571 ostei_general_vrr1_I(10, 5,
572 one_over_2p, a_over_p, aop_PQ, P_PA,
573 PRIM_INT__m_s_s_s, PRIM_INT__l_s_s_s, PRIM_INT__n_s_s_s);
574
575
576 ostei_general_vrr_K(10, 0, 1, 0, 4,
577 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
578 PRIM_INT__n_s_s_s, NULL, NULL, PRIM_INT__m_s_s_s, NULL, PRIM_INT__n_s_p_s);
579
580
581 ostei_general_vrr_K(10, 0, 2, 0, 3,
582 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
583 PRIM_INT__n_s_p_s, PRIM_INT__n_s_s_s, NULL, PRIM_INT__m_s_p_s, NULL, PRIM_INT__n_s_d_s);
584
585
586 ostei_general_vrr_K(10, 0, 3, 0, 2,
587 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
588 PRIM_INT__n_s_d_s, PRIM_INT__n_s_p_s, NULL, PRIM_INT__m_s_d_s, NULL, PRIM_INT__n_s_f_s);
589
590
591 ostei_general_vrr_K(10, 0, 4, 0, 1,
592 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
593 PRIM_INT__n_s_f_s, PRIM_INT__n_s_d_s, NULL, PRIM_INT__m_s_f_s, NULL, PRIM_INT__n_s_g_s);
594
595
596
597
598 ////////////////////////////////////
599 // Accumulate contracted integrals
600 ////////////////////////////////////
601 if(lastoffset == 0)
602 {
603 contract_all(280, PRIM_INT__i_s_f_s, PRIM_PTR_INT__i_s_f_s);
604 contract_all(420, PRIM_INT__i_s_g_s, PRIM_PTR_INT__i_s_g_s);
605 contract_all(360, PRIM_INT__k_s_f_s, PRIM_PTR_INT__k_s_f_s);
606 contract_all(540, PRIM_INT__k_s_g_s, PRIM_PTR_INT__k_s_g_s);
607 contract_all(450, PRIM_INT__l_s_f_s, PRIM_PTR_INT__l_s_f_s);
608 contract_all(675, PRIM_INT__l_s_g_s, PRIM_PTR_INT__l_s_g_s);
609 contract_all(550, PRIM_INT__m_s_f_s, PRIM_PTR_INT__m_s_f_s);
610 contract_all(825, PRIM_INT__m_s_g_s, PRIM_PTR_INT__m_s_g_s);
611 contract_all(660, PRIM_INT__n_s_f_s, PRIM_PTR_INT__n_s_f_s);
612 contract_all(990, PRIM_INT__n_s_g_s, PRIM_PTR_INT__n_s_g_s);
613 }
614 else
615 {
616 contract(280, shelloffsets, PRIM_INT__i_s_f_s, PRIM_PTR_INT__i_s_f_s);
617 contract(420, shelloffsets, PRIM_INT__i_s_g_s, PRIM_PTR_INT__i_s_g_s);
618 contract(360, shelloffsets, PRIM_INT__k_s_f_s, PRIM_PTR_INT__k_s_f_s);
619 contract(540, shelloffsets, PRIM_INT__k_s_g_s, PRIM_PTR_INT__k_s_g_s);
620 contract(450, shelloffsets, PRIM_INT__l_s_f_s, PRIM_PTR_INT__l_s_f_s);
621 contract(675, shelloffsets, PRIM_INT__l_s_g_s, PRIM_PTR_INT__l_s_g_s);
622 contract(550, shelloffsets, PRIM_INT__m_s_f_s, PRIM_PTR_INT__m_s_f_s);
623 contract(825, shelloffsets, PRIM_INT__m_s_g_s, PRIM_PTR_INT__m_s_g_s);
624 contract(660, shelloffsets, PRIM_INT__n_s_f_s, PRIM_PTR_INT__n_s_f_s);
625 contract(990, shelloffsets, PRIM_INT__n_s_g_s, PRIM_PTR_INT__n_s_g_s);
626 PRIM_PTR_INT__i_s_f_s += lastoffset*280;
627 PRIM_PTR_INT__i_s_g_s += lastoffset*420;
628 PRIM_PTR_INT__k_s_f_s += lastoffset*360;
629 PRIM_PTR_INT__k_s_g_s += lastoffset*540;
630 PRIM_PTR_INT__l_s_f_s += lastoffset*450;
631 PRIM_PTR_INT__l_s_g_s += lastoffset*675;
632 PRIM_PTR_INT__m_s_f_s += lastoffset*550;
633 PRIM_PTR_INT__m_s_g_s += lastoffset*825;
634 PRIM_PTR_INT__n_s_f_s += lastoffset*660;
635 PRIM_PTR_INT__n_s_g_s += lastoffset*990;
636 }
637
638 } // close loop over j
639 } // close loop over i
640
641 //Advance to the next batch
642 jstart = SIMINT_SIMD_ROUND(jend);
643
644 //////////////////////////////////////////////
645 // Contracted integrals: Horizontal recurrance
646 //////////////////////////////////////////////
647
648
649 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
650
651
652 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
653 {
654 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
655
656 // set up HRR pointers
657 double const * restrict HRR_INT__i_s_f_s = INT__i_s_f_s + abcd * 280;
658 double const * restrict HRR_INT__i_s_g_s = INT__i_s_g_s + abcd * 420;
659 double const * restrict HRR_INT__k_s_f_s = INT__k_s_f_s + abcd * 360;
660 double const * restrict HRR_INT__k_s_g_s = INT__k_s_g_s + abcd * 540;
661 double const * restrict HRR_INT__l_s_f_s = INT__l_s_f_s + abcd * 450;
662 double const * restrict HRR_INT__l_s_g_s = INT__l_s_g_s + abcd * 675;
663 double const * restrict HRR_INT__m_s_f_s = INT__m_s_f_s + abcd * 550;
664 double const * restrict HRR_INT__m_s_g_s = INT__m_s_g_s + abcd * 825;
665 double const * restrict HRR_INT__n_s_f_s = INT__n_s_f_s + abcd * 660;
666 double const * restrict HRR_INT__n_s_g_s = INT__n_s_g_s + abcd * 990;
667 double * restrict HRR_INT__i_g_f_p = INT__i_g_f_p + real_abcd * 12600;
668
669 // form INT__i_p_f_s
670 ostei_general_hrr_J(6, 1, 3, 0, hAB, HRR_INT__k_s_f_s, HRR_INT__i_s_f_s, HRR_INT__i_p_f_s);
671
672 // form INT__i_p_g_s
673 ostei_general_hrr_J(6, 1, 4, 0, hAB, HRR_INT__k_s_g_s, HRR_INT__i_s_g_s, HRR_INT__i_p_g_s);
674
675 // form INT__k_p_f_s
676 ostei_general_hrr_J(7, 1, 3, 0, hAB, HRR_INT__l_s_f_s, HRR_INT__k_s_f_s, HRR_INT__k_p_f_s);
677
678 // form INT__k_p_g_s
679 ostei_general_hrr_J(7, 1, 4, 0, hAB, HRR_INT__l_s_g_s, HRR_INT__k_s_g_s, HRR_INT__k_p_g_s);
680
681 // form INT__l_p_f_s
682 ostei_general_hrr_J(8, 1, 3, 0, hAB, HRR_INT__m_s_f_s, HRR_INT__l_s_f_s, HRR_INT__l_p_f_s);
683
684 // form INT__l_p_g_s
685 ostei_general_hrr_J(8, 1, 4, 0, hAB, HRR_INT__m_s_g_s, HRR_INT__l_s_g_s, HRR_INT__l_p_g_s);
686
687 // form INT__m_p_f_s
688 ostei_general_hrr_J(9, 1, 3, 0, hAB, HRR_INT__n_s_f_s, HRR_INT__m_s_f_s, HRR_INT__m_p_f_s);
689
690 // form INT__m_p_g_s
691 ostei_general_hrr_J(9, 1, 4, 0, hAB, HRR_INT__n_s_g_s, HRR_INT__m_s_g_s, HRR_INT__m_p_g_s);
692
693 // form INT__i_d_f_s
694 ostei_general_hrr_J(6, 2, 3, 0, hAB, HRR_INT__k_p_f_s, HRR_INT__i_p_f_s, HRR_INT__i_d_f_s);
695
696 // form INT__i_d_g_s
697 ostei_general_hrr_J(6, 2, 4, 0, hAB, HRR_INT__k_p_g_s, HRR_INT__i_p_g_s, HRR_INT__i_d_g_s);
698
699 // form INT__k_d_f_s
700 ostei_general_hrr_J(7, 2, 3, 0, hAB, HRR_INT__l_p_f_s, HRR_INT__k_p_f_s, HRR_INT__k_d_f_s);
701
702 // form INT__k_d_g_s
703 ostei_general_hrr_J(7, 2, 4, 0, hAB, HRR_INT__l_p_g_s, HRR_INT__k_p_g_s, HRR_INT__k_d_g_s);
704
705 // form INT__l_d_f_s
706 ostei_general_hrr_J(8, 2, 3, 0, hAB, HRR_INT__m_p_f_s, HRR_INT__l_p_f_s, HRR_INT__l_d_f_s);
707
708 // form INT__l_d_g_s
709 ostei_general_hrr_J(8, 2, 4, 0, hAB, HRR_INT__m_p_g_s, HRR_INT__l_p_g_s, HRR_INT__l_d_g_s);
710
711 // form INT__i_f_f_s
712 ostei_general_hrr_J(6, 3, 3, 0, hAB, HRR_INT__k_d_f_s, HRR_INT__i_d_f_s, HRR_INT__i_f_f_s);
713
714 // form INT__i_f_g_s
715 ostei_general_hrr_J(6, 3, 4, 0, hAB, HRR_INT__k_d_g_s, HRR_INT__i_d_g_s, HRR_INT__i_f_g_s);
716
717 // form INT__k_f_f_s
718 ostei_general_hrr_J(7, 3, 3, 0, hAB, HRR_INT__l_d_f_s, HRR_INT__k_d_f_s, HRR_INT__k_f_f_s);
719
720 // form INT__k_f_g_s
721 ostei_general_hrr_J(7, 3, 4, 0, hAB, HRR_INT__l_d_g_s, HRR_INT__k_d_g_s, HRR_INT__k_f_g_s);
722
723 // form INT__i_g_f_s
724 ostei_general_hrr_J(6, 4, 3, 0, hAB, HRR_INT__k_f_f_s, HRR_INT__i_f_f_s, HRR_INT__i_g_f_s);
725
726 // form INT__i_g_g_s
727 ostei_general_hrr_J(6, 4, 4, 0, hAB, HRR_INT__k_f_g_s, HRR_INT__i_f_g_s, HRR_INT__i_g_g_s);
728
729 // form INT__i_g_f_p
730 HRR_L_f_p(
731 HRR_INT__i_g_f_p,
732 HRR_INT__i_g_f_s,
733 HRR_INT__i_g_g_s,
734 hCD, 420);
735
736
737 } // close HRR loop
738
739
740 } // close loop cdbatch
741
742 istart = iend;
743 } // close loop over ab
744
745 return P.nshell12_clip * Q.nshell12_clip;
746 }
747
ostei_g_i_f_p(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__g_i_f_p)748 int ostei_g_i_f_p(struct simint_multi_shellpair const P,
749 struct simint_multi_shellpair const Q,
750 double screen_tol,
751 double * const restrict work,
752 double * const restrict INT__g_i_f_p)
753 {
754 double P_AB[3*P.nshell12];
755 struct simint_multi_shellpair P_tmp = P;
756 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
757 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
758 P_tmp.AB_x = P_AB;
759 P_tmp.AB_y = P_AB + P.nshell12;
760 P_tmp.AB_z = P_AB + 2*P.nshell12;
761
762 for(int i = 0; i < P.nshell12; i++)
763 {
764 P_tmp.AB_x[i] = -P.AB_x[i];
765 P_tmp.AB_y[i] = -P.AB_y[i];
766 P_tmp.AB_z[i] = -P.AB_z[i];
767 }
768
769 int ret = ostei_i_g_f_p(P_tmp, Q, screen_tol, work, INT__g_i_f_p);
770 double buffer[12600] SIMINT_ALIGN_ARRAY_DBL;
771
772 for(int q = 0; q < ret; q++)
773 {
774 int idx = 0;
775 for(int a = 0; a < 15; ++a)
776 for(int b = 0; b < 28; ++b)
777 for(int c = 0; c < 10; ++c)
778 for(int d = 0; d < 3; ++d)
779 buffer[idx++] = INT__g_i_f_p[q*12600+b*450+a*30+c*3+d];
780
781 memcpy(INT__g_i_f_p+q*12600, buffer, 12600*sizeof(double));
782 }
783
784 return ret;
785 }
786
ostei_i_g_p_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__i_g_p_f)787 int ostei_i_g_p_f(struct simint_multi_shellpair const P,
788 struct simint_multi_shellpair const Q,
789 double screen_tol,
790 double * const restrict work,
791 double * const restrict INT__i_g_p_f)
792 {
793 double Q_AB[3*Q.nshell12];
794 struct simint_multi_shellpair Q_tmp = Q;
795 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
796 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
797 Q_tmp.AB_x = Q_AB;
798 Q_tmp.AB_y = Q_AB + Q.nshell12;
799 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
800
801 for(int i = 0; i < Q.nshell12; i++)
802 {
803 Q_tmp.AB_x[i] = -Q.AB_x[i];
804 Q_tmp.AB_y[i] = -Q.AB_y[i];
805 Q_tmp.AB_z[i] = -Q.AB_z[i];
806 }
807
808 int ret = ostei_i_g_f_p(P, Q_tmp, screen_tol, work, INT__i_g_p_f);
809 double buffer[12600] SIMINT_ALIGN_ARRAY_DBL;
810
811 for(int q = 0; q < ret; q++)
812 {
813 int idx = 0;
814 for(int a = 0; a < 28; ++a)
815 for(int b = 0; b < 15; ++b)
816 for(int c = 0; c < 3; ++c)
817 for(int d = 0; d < 10; ++d)
818 buffer[idx++] = INT__i_g_p_f[q*12600+a*450+b*30+d*3+c];
819
820 memcpy(INT__i_g_p_f+q*12600, buffer, 12600*sizeof(double));
821 }
822
823 return ret;
824 }
825
ostei_g_i_p_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__g_i_p_f)826 int ostei_g_i_p_f(struct simint_multi_shellpair const P,
827 struct simint_multi_shellpair const Q,
828 double screen_tol,
829 double * const restrict work,
830 double * const restrict INT__g_i_p_f)
831 {
832 double P_AB[3*P.nshell12];
833 struct simint_multi_shellpair P_tmp = P;
834 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
835 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
836 P_tmp.AB_x = P_AB;
837 P_tmp.AB_y = P_AB + P.nshell12;
838 P_tmp.AB_z = P_AB + 2*P.nshell12;
839
840 for(int i = 0; i < P.nshell12; i++)
841 {
842 P_tmp.AB_x[i] = -P.AB_x[i];
843 P_tmp.AB_y[i] = -P.AB_y[i];
844 P_tmp.AB_z[i] = -P.AB_z[i];
845 }
846
847 double Q_AB[3*Q.nshell12];
848 struct simint_multi_shellpair Q_tmp = Q;
849 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
850 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
851 Q_tmp.AB_x = Q_AB;
852 Q_tmp.AB_y = Q_AB + Q.nshell12;
853 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
854
855 for(int i = 0; i < Q.nshell12; i++)
856 {
857 Q_tmp.AB_x[i] = -Q.AB_x[i];
858 Q_tmp.AB_y[i] = -Q.AB_y[i];
859 Q_tmp.AB_z[i] = -Q.AB_z[i];
860 }
861
862 int ret = ostei_i_g_f_p(P_tmp, Q_tmp, screen_tol, work, INT__g_i_p_f);
863 double buffer[12600] SIMINT_ALIGN_ARRAY_DBL;
864
865 for(int q = 0; q < ret; q++)
866 {
867 int idx = 0;
868 for(int a = 0; a < 15; ++a)
869 for(int b = 0; b < 28; ++b)
870 for(int c = 0; c < 3; ++c)
871 for(int d = 0; d < 10; ++d)
872 buffer[idx++] = INT__g_i_p_f[q*12600+b*450+a*30+d*3+c];
873
874 memcpy(INT__g_i_p_f+q*12600, buffer, 12600*sizeof(double));
875 }
876
877 return ret;
878 }
879
880