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_d_d_k_f(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_d_k_f)8 int ostei_d_d_k_f(struct simint_multi_shellpair const P,
9 struct simint_multi_shellpair const Q,
10 double screen_tol,
11 double * const restrict work,
12 double * const restrict INT__d_d_k_f)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__d_d_k_f);
17 int ab, cd, abcd;
18 int istart, jstart;
19 int iprimcd, nprim_icd, icd;
20 const int check_screen = (screen_tol > 0.0);
21 int i, j;
22 int n;
23 int not_screened;
24 int real_abcd;
25 int iket;
26 int ibra;
27
28 // partition workspace
29 double * const INT__d_s_k_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__d_s_l_s = work + (SIMINT_NSHELL_SIMD * 216);
31 double * const INT__d_s_m_s = work + (SIMINT_NSHELL_SIMD * 486);
32 double * const INT__d_s_n_s = work + (SIMINT_NSHELL_SIMD * 816);
33 double * const INT__f_s_k_s = work + (SIMINT_NSHELL_SIMD * 1212);
34 double * const INT__f_s_l_s = work + (SIMINT_NSHELL_SIMD * 1572);
35 double * const INT__f_s_m_s = work + (SIMINT_NSHELL_SIMD * 2022);
36 double * const INT__f_s_n_s = work + (SIMINT_NSHELL_SIMD * 2572);
37 double * const INT__g_s_k_s = work + (SIMINT_NSHELL_SIMD * 3232);
38 double * const INT__g_s_l_s = work + (SIMINT_NSHELL_SIMD * 3772);
39 double * const INT__g_s_m_s = work + (SIMINT_NSHELL_SIMD * 4447);
40 double * const INT__g_s_n_s = work + (SIMINT_NSHELL_SIMD * 5272);
41 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*6262);
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 + 15;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_d_s = primwork + 57;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_f_s = primwork + 135;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_g_s = primwork + 255;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_h_s = primwork + 420;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_i_s = primwork + 630;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_k_s = primwork + 882;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_l_s = primwork + 1170;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_m_s = primwork + 1485;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_n_s = primwork + 1815;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_g_s = primwork + 2145;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_h_s = primwork + 2325;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_i_s = primwork + 2577;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_k_s = primwork + 2913;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_l_s = primwork + 3345;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_m_s = primwork + 3885;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_n_s = primwork + 4545;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_h_s = primwork + 5337;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_i_s = primwork + 5715;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_k_s = primwork + 6219;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_l_s = primwork + 6867;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_m_s = primwork + 7677;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_n_s = primwork + 8667;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_i_s = primwork + 9855;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_k_s = primwork + 10415;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_l_s = primwork + 11135;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_m_s = primwork + 12035;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_n_s = primwork + 13135;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_k_s = primwork + 14455;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_l_s = primwork + 14995;
73 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_m_s = primwork + 15670;
74 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_n_s = primwork + 16495;
75 double * const hrrwork = (double *)(primwork + 17485);
76 double * const HRR_INT__d_p_k_s = hrrwork + 0;
77 double * const HRR_INT__d_p_l_s = hrrwork + 648;
78 double * const HRR_INT__d_p_m_s = hrrwork + 1458;
79 double * const HRR_INT__d_p_n_s = hrrwork + 2448;
80 double * const HRR_INT__d_d_k_s = hrrwork + 3636;
81 double * const HRR_INT__d_d_k_p = hrrwork + 4932;
82 double * const HRR_INT__d_d_k_d = hrrwork + 8820;
83 double * const HRR_INT__d_d_l_s = hrrwork + 16596;
84 double * const HRR_INT__d_d_l_p = hrrwork + 18216;
85 double * const HRR_INT__d_d_l_d = hrrwork + 23076;
86 double * const HRR_INT__d_d_m_s = hrrwork + 32796;
87 double * const HRR_INT__d_d_m_p = hrrwork + 34776;
88 double * const HRR_INT__d_d_n_s = hrrwork + 40716;
89 double * const HRR_INT__f_p_k_s = hrrwork + 43092;
90 double * const HRR_INT__f_p_l_s = hrrwork + 44172;
91 double * const HRR_INT__f_p_m_s = hrrwork + 45522;
92 double * const HRR_INT__f_p_n_s = hrrwork + 47172;
93
94
95 // Create constants
96 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
97 const SIMINT_DBLTYPE const_10 = SIMINT_DBLSET1(10);
98 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
99 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
100 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
101 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
102 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
103 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
104 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
105 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
106 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
107
108
109 ////////////////////////////////////////
110 // Loop over shells and primitives
111 ////////////////////////////////////////
112
113 real_abcd = 0;
114 istart = 0;
115 for(ab = 0; ab < P.nshell12_clip; ++ab)
116 {
117 const int iend = istart + P.nprim12[ab];
118
119 cd = 0;
120 jstart = 0;
121
122 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
123 {
124 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
125 int jend = jstart;
126 for(i = 0; i < nshellbatch; i++)
127 jend += Q.nprim12[cd+i];
128
129 // Clear the beginning of the workspace (where we are accumulating integrals)
130 memset(work, 0, SIMINT_NSHELL_SIMD * 6262 * sizeof(double));
131 abcd = 0;
132
133
134 for(i = istart; i < iend; ++i)
135 {
136 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
137
138 if(check_screen)
139 {
140 // Skip this whole thing if always insignificant
141 if((P.screen[i] * Q.screen_max) < screen_tol)
142 continue;
143 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
144 }
145
146 icd = 0;
147 iprimcd = 0;
148 nprim_icd = Q.nprim12[cd];
149 double * restrict PRIM_PTR_INT__d_s_k_s = INT__d_s_k_s + abcd * 216;
150 double * restrict PRIM_PTR_INT__d_s_l_s = INT__d_s_l_s + abcd * 270;
151 double * restrict PRIM_PTR_INT__d_s_m_s = INT__d_s_m_s + abcd * 330;
152 double * restrict PRIM_PTR_INT__d_s_n_s = INT__d_s_n_s + abcd * 396;
153 double * restrict PRIM_PTR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
154 double * restrict PRIM_PTR_INT__f_s_l_s = INT__f_s_l_s + abcd * 450;
155 double * restrict PRIM_PTR_INT__f_s_m_s = INT__f_s_m_s + abcd * 550;
156 double * restrict PRIM_PTR_INT__f_s_n_s = INT__f_s_n_s + abcd * 660;
157 double * restrict PRIM_PTR_INT__g_s_k_s = INT__g_s_k_s + abcd * 540;
158 double * restrict PRIM_PTR_INT__g_s_l_s = INT__g_s_l_s + abcd * 675;
159 double * restrict PRIM_PTR_INT__g_s_m_s = INT__g_s_m_s + abcd * 825;
160 double * restrict PRIM_PTR_INT__g_s_n_s = INT__g_s_n_s + abcd * 990;
161
162
163
164 // Load these one per loop over i
165 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
166 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
167 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
168
169 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
170
171 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
172 {
173 // calculate the shell offsets
174 // these are the offset from the shell pointed to by cd
175 // for each element
176 int shelloffsets[SIMINT_SIMD_LEN] = {0};
177 int lastoffset = 0;
178 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
179
180 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
181 {
182 // Handle if the first element of the vector is a new shell
183 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
184 {
185 nprim_icd += Q.nprim12[cd + (++icd)];
186 PRIM_PTR_INT__d_s_k_s += 216;
187 PRIM_PTR_INT__d_s_l_s += 270;
188 PRIM_PTR_INT__d_s_m_s += 330;
189 PRIM_PTR_INT__d_s_n_s += 396;
190 PRIM_PTR_INT__f_s_k_s += 360;
191 PRIM_PTR_INT__f_s_l_s += 450;
192 PRIM_PTR_INT__f_s_m_s += 550;
193 PRIM_PTR_INT__f_s_n_s += 660;
194 PRIM_PTR_INT__g_s_k_s += 540;
195 PRIM_PTR_INT__g_s_l_s += 675;
196 PRIM_PTR_INT__g_s_m_s += 825;
197 PRIM_PTR_INT__g_s_n_s += 990;
198 }
199 iprimcd++;
200 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
201 {
202 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
203 {
204 shelloffsets[n] = shelloffsets[n-1] + 1;
205 lastoffset++;
206 nprim_icd += Q.nprim12[cd + (++icd)];
207 }
208 else
209 shelloffsets[n] = shelloffsets[n-1];
210 iprimcd++;
211 }
212 }
213 else
214 iprimcd += SIMINT_SIMD_LEN;
215
216 // Do we have to compute this vector (or has it been screened out)?
217 // (not_screened != 0 means we have to do this vector)
218 if(check_screen)
219 {
220 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
221 if(vmax < screen_tol)
222 {
223 PRIM_PTR_INT__d_s_k_s += lastoffset*216;
224 PRIM_PTR_INT__d_s_l_s += lastoffset*270;
225 PRIM_PTR_INT__d_s_m_s += lastoffset*330;
226 PRIM_PTR_INT__d_s_n_s += lastoffset*396;
227 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
228 PRIM_PTR_INT__f_s_l_s += lastoffset*450;
229 PRIM_PTR_INT__f_s_m_s += lastoffset*550;
230 PRIM_PTR_INT__f_s_n_s += lastoffset*660;
231 PRIM_PTR_INT__g_s_k_s += lastoffset*540;
232 PRIM_PTR_INT__g_s_l_s += lastoffset*675;
233 PRIM_PTR_INT__g_s_m_s += lastoffset*825;
234 PRIM_PTR_INT__g_s_n_s += lastoffset*990;
235 continue;
236 }
237 }
238
239 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
240 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
241 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
242 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
243
244
245 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
246 SIMINT_DBLTYPE PQ[3];
247 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
248 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
249 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
250 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
251 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
252 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
253
254 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
255 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
256 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
257 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
258 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
259 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
260 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
261
262 // NOTE: Minus sign!
263 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
264 SIMINT_DBLTYPE aop_PQ[3];
265 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
266 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
267 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
268
269 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
270 SIMINT_DBLTYPE aoq_PQ[3];
271 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
272 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
273 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
274 // Put a minus sign here so we don't have to in RR routines
275 a_over_q = SIMINT_NEG(a_over_q);
276
277
278 //////////////////////////////////////////////
279 // Fjt function section
280 // Maximum v value: 14
281 //////////////////////////////////////////////
282 // The parameter to the Fjt function
283 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
284
285
286 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
287
288
289 boys_F_split(PRIM_INT__s_s_s_s, F_x, 14);
290 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
291 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
292 for(n = 0; n <= 14; n++)
293 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
294
295 //////////////////////////////////////////////
296 // Primitive integrals: Vertical recurrance
297 //////////////////////////////////////////////
298
299 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
300 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
301 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
302 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
303 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
304 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
305 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
306 const SIMINT_DBLTYPE vrr_const_5_over_2q = SIMINT_MUL(const_5, one_over_2q);
307 const SIMINT_DBLTYPE vrr_const_6_over_2q = SIMINT_MUL(const_6, one_over_2q);
308 const SIMINT_DBLTYPE vrr_const_7_over_2q = SIMINT_MUL(const_7, one_over_2q);
309 const SIMINT_DBLTYPE vrr_const_8_over_2q = SIMINT_MUL(const_8, one_over_2q);
310 const SIMINT_DBLTYPE vrr_const_9_over_2q = SIMINT_MUL(const_9, 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 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
319 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
320 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
321
322
323
324 // Forming PRIM_INT__s_s_p_s[14 * 3];
325 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
326 {
327
328 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
329 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]);
330
331 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
332 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]);
333
334 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
335 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]);
336
337 }
338
339
340
341 // Forming PRIM_INT__s_s_d_s[13 * 6];
342 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
343 {
344
345 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
346 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]);
347 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]);
348
349 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 0]);
350 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]);
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 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
357 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]);
358 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]);
359
360 }
361
362
363
364 // Forming PRIM_INT__s_s_f_s[12 * 10];
365 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
366 {
367
368 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
369 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]);
370 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]);
371
372 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
373 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]);
374
375 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
376 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]);
377
378 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
379 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]);
380
381 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
382 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]);
383
384 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
385 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]);
386
387 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
388 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]);
389 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]);
390
391 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
392 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]);
393
394 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
395 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]);
396
397 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
398 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]);
399 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]);
400
401 }
402
403
404 VRR_K_s_s_g_s(
405 PRIM_INT__s_s_g_s,
406 PRIM_INT__s_s_f_s,
407 PRIM_INT__s_s_d_s,
408 Q_PA,
409 a_over_q,
410 aoq_PQ,
411 one_over_2q,
412 11);
413
414
415 VRR_K_s_s_h_s(
416 PRIM_INT__s_s_h_s,
417 PRIM_INT__s_s_g_s,
418 PRIM_INT__s_s_f_s,
419 Q_PA,
420 a_over_q,
421 aoq_PQ,
422 one_over_2q,
423 10);
424
425
426 ostei_general_vrr1_K(6, 9,
427 one_over_2q, a_over_q, aoq_PQ, Q_PA,
428 PRIM_INT__s_s_h_s, PRIM_INT__s_s_g_s, PRIM_INT__s_s_i_s);
429
430
431 ostei_general_vrr1_K(7, 8,
432 one_over_2q, a_over_q, aoq_PQ, Q_PA,
433 PRIM_INT__s_s_i_s, PRIM_INT__s_s_h_s, PRIM_INT__s_s_k_s);
434
435
436 ostei_general_vrr_I(1, 0, 7, 0, 4,
437 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
438 PRIM_INT__s_s_k_s, NULL, NULL, PRIM_INT__s_s_i_s, NULL, PRIM_INT__p_s_k_s);
439
440
441 ostei_general_vrr_I(1, 0, 6, 0, 4,
442 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
443 PRIM_INT__s_s_i_s, NULL, NULL, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_i_s);
444
445
446 ostei_general_vrr_I(2, 0, 7, 0, 3,
447 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
448 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);
449
450
451 ostei_general_vrr1_K(8, 7,
452 one_over_2q, a_over_q, aoq_PQ, Q_PA,
453 PRIM_INT__s_s_k_s, PRIM_INT__s_s_i_s, PRIM_INT__s_s_l_s);
454
455
456 ostei_general_vrr_I(1, 0, 8, 0, 4,
457 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
458 PRIM_INT__s_s_l_s, NULL, NULL, PRIM_INT__s_s_k_s, NULL, PRIM_INT__p_s_l_s);
459
460
461 ostei_general_vrr_I(2, 0, 8, 0, 3,
462 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
463 PRIM_INT__p_s_l_s, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_k_s, NULL, PRIM_INT__d_s_l_s);
464
465
466 ostei_general_vrr_I(1, 0, 5, 0, 4,
467 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
468 PRIM_INT__s_s_h_s, NULL, NULL, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_h_s);
469
470
471 ostei_general_vrr_I(2, 0, 6, 0, 3,
472 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
473 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);
474
475
476 ostei_general_vrr_I(3, 0, 7, 0, 2,
477 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
478 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);
479
480
481 ostei_general_vrr1_K(9, 6,
482 one_over_2q, a_over_q, aoq_PQ, Q_PA,
483 PRIM_INT__s_s_l_s, PRIM_INT__s_s_k_s, PRIM_INT__s_s_m_s);
484
485
486 ostei_general_vrr_I(1, 0, 9, 0, 4,
487 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
488 PRIM_INT__s_s_m_s, NULL, NULL, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_m_s);
489
490
491 ostei_general_vrr_I(2, 0, 9, 0, 3,
492 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
493 PRIM_INT__p_s_m_s, PRIM_INT__s_s_m_s, NULL, PRIM_INT__p_s_l_s, NULL, PRIM_INT__d_s_m_s);
494
495
496 ostei_general_vrr_I(3, 0, 8, 0, 2,
497 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
498 PRIM_INT__d_s_l_s, PRIM_INT__p_s_l_s, NULL, PRIM_INT__d_s_k_s, NULL, PRIM_INT__f_s_l_s);
499
500
501 VRR_I_p_s_g_s(
502 PRIM_INT__p_s_g_s,
503 PRIM_INT__s_s_g_s,
504 PRIM_INT__s_s_f_s,
505 P_PA,
506 aop_PQ,
507 one_over_2pq,
508 4);
509
510
511 ostei_general_vrr_I(2, 0, 5, 0, 3,
512 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
513 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);
514
515
516 ostei_general_vrr_I(3, 0, 6, 0, 2,
517 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
518 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);
519
520
521 ostei_general_vrr_I(4, 0, 7, 0, 1,
522 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
523 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);
524
525
526 ostei_general_vrr1_K(10, 5,
527 one_over_2q, a_over_q, aoq_PQ, Q_PA,
528 PRIM_INT__s_s_m_s, PRIM_INT__s_s_l_s, PRIM_INT__s_s_n_s);
529
530
531 ostei_general_vrr_I(1, 0, 10, 0, 4,
532 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
533 PRIM_INT__s_s_n_s, NULL, NULL, PRIM_INT__s_s_m_s, NULL, PRIM_INT__p_s_n_s);
534
535
536 ostei_general_vrr_I(2, 0, 10, 0, 3,
537 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
538 PRIM_INT__p_s_n_s, PRIM_INT__s_s_n_s, NULL, PRIM_INT__p_s_m_s, NULL, PRIM_INT__d_s_n_s);
539
540
541 ostei_general_vrr_I(3, 0, 9, 0, 2,
542 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
543 PRIM_INT__d_s_m_s, PRIM_INT__p_s_m_s, NULL, PRIM_INT__d_s_l_s, NULL, PRIM_INT__f_s_m_s);
544
545
546 ostei_general_vrr_I(4, 0, 8, 0, 1,
547 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
548 PRIM_INT__f_s_l_s, PRIM_INT__d_s_l_s, NULL, PRIM_INT__f_s_k_s, NULL, PRIM_INT__g_s_l_s);
549
550
551 ostei_general_vrr_I(3, 0, 10, 0, 2,
552 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
553 PRIM_INT__d_s_n_s, PRIM_INT__p_s_n_s, NULL, PRIM_INT__d_s_m_s, NULL, PRIM_INT__f_s_n_s);
554
555
556 ostei_general_vrr_I(4, 0, 9, 0, 1,
557 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
558 PRIM_INT__f_s_m_s, PRIM_INT__d_s_m_s, NULL, PRIM_INT__f_s_l_s, NULL, PRIM_INT__g_s_m_s);
559
560
561 ostei_general_vrr_I(4, 0, 10, 0, 1,
562 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
563 PRIM_INT__f_s_n_s, PRIM_INT__d_s_n_s, NULL, PRIM_INT__f_s_m_s, NULL, PRIM_INT__g_s_n_s);
564
565
566
567
568 ////////////////////////////////////
569 // Accumulate contracted integrals
570 ////////////////////////////////////
571 if(lastoffset == 0)
572 {
573 contract_all(216, PRIM_INT__d_s_k_s, PRIM_PTR_INT__d_s_k_s);
574 contract_all(270, PRIM_INT__d_s_l_s, PRIM_PTR_INT__d_s_l_s);
575 contract_all(330, PRIM_INT__d_s_m_s, PRIM_PTR_INT__d_s_m_s);
576 contract_all(396, PRIM_INT__d_s_n_s, PRIM_PTR_INT__d_s_n_s);
577 contract_all(360, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
578 contract_all(450, PRIM_INT__f_s_l_s, PRIM_PTR_INT__f_s_l_s);
579 contract_all(550, PRIM_INT__f_s_m_s, PRIM_PTR_INT__f_s_m_s);
580 contract_all(660, PRIM_INT__f_s_n_s, PRIM_PTR_INT__f_s_n_s);
581 contract_all(540, PRIM_INT__g_s_k_s, PRIM_PTR_INT__g_s_k_s);
582 contract_all(675, PRIM_INT__g_s_l_s, PRIM_PTR_INT__g_s_l_s);
583 contract_all(825, PRIM_INT__g_s_m_s, PRIM_PTR_INT__g_s_m_s);
584 contract_all(990, PRIM_INT__g_s_n_s, PRIM_PTR_INT__g_s_n_s);
585 }
586 else
587 {
588 contract(216, shelloffsets, PRIM_INT__d_s_k_s, PRIM_PTR_INT__d_s_k_s);
589 contract(270, shelloffsets, PRIM_INT__d_s_l_s, PRIM_PTR_INT__d_s_l_s);
590 contract(330, shelloffsets, PRIM_INT__d_s_m_s, PRIM_PTR_INT__d_s_m_s);
591 contract(396, shelloffsets, PRIM_INT__d_s_n_s, PRIM_PTR_INT__d_s_n_s);
592 contract(360, shelloffsets, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
593 contract(450, shelloffsets, PRIM_INT__f_s_l_s, PRIM_PTR_INT__f_s_l_s);
594 contract(550, shelloffsets, PRIM_INT__f_s_m_s, PRIM_PTR_INT__f_s_m_s);
595 contract(660, shelloffsets, PRIM_INT__f_s_n_s, PRIM_PTR_INT__f_s_n_s);
596 contract(540, shelloffsets, PRIM_INT__g_s_k_s, PRIM_PTR_INT__g_s_k_s);
597 contract(675, shelloffsets, PRIM_INT__g_s_l_s, PRIM_PTR_INT__g_s_l_s);
598 contract(825, shelloffsets, PRIM_INT__g_s_m_s, PRIM_PTR_INT__g_s_m_s);
599 contract(990, shelloffsets, PRIM_INT__g_s_n_s, PRIM_PTR_INT__g_s_n_s);
600 PRIM_PTR_INT__d_s_k_s += lastoffset*216;
601 PRIM_PTR_INT__d_s_l_s += lastoffset*270;
602 PRIM_PTR_INT__d_s_m_s += lastoffset*330;
603 PRIM_PTR_INT__d_s_n_s += lastoffset*396;
604 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
605 PRIM_PTR_INT__f_s_l_s += lastoffset*450;
606 PRIM_PTR_INT__f_s_m_s += lastoffset*550;
607 PRIM_PTR_INT__f_s_n_s += lastoffset*660;
608 PRIM_PTR_INT__g_s_k_s += lastoffset*540;
609 PRIM_PTR_INT__g_s_l_s += lastoffset*675;
610 PRIM_PTR_INT__g_s_m_s += lastoffset*825;
611 PRIM_PTR_INT__g_s_n_s += lastoffset*990;
612 }
613
614 } // close loop over j
615 } // close loop over i
616
617 //Advance to the next batch
618 jstart = SIMINT_SIMD_ROUND(jend);
619
620 //////////////////////////////////////////////
621 // Contracted integrals: Horizontal recurrance
622 //////////////////////////////////////////////
623
624
625 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
626
627
628 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
629 {
630 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
631
632 // set up HRR pointers
633 double const * restrict HRR_INT__d_s_k_s = INT__d_s_k_s + abcd * 216;
634 double const * restrict HRR_INT__d_s_l_s = INT__d_s_l_s + abcd * 270;
635 double const * restrict HRR_INT__d_s_m_s = INT__d_s_m_s + abcd * 330;
636 double const * restrict HRR_INT__d_s_n_s = INT__d_s_n_s + abcd * 396;
637 double const * restrict HRR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
638 double const * restrict HRR_INT__f_s_l_s = INT__f_s_l_s + abcd * 450;
639 double const * restrict HRR_INT__f_s_m_s = INT__f_s_m_s + abcd * 550;
640 double const * restrict HRR_INT__f_s_n_s = INT__f_s_n_s + abcd * 660;
641 double const * restrict HRR_INT__g_s_k_s = INT__g_s_k_s + abcd * 540;
642 double const * restrict HRR_INT__g_s_l_s = INT__g_s_l_s + abcd * 675;
643 double const * restrict HRR_INT__g_s_m_s = INT__g_s_m_s + abcd * 825;
644 double const * restrict HRR_INT__g_s_n_s = INT__g_s_n_s + abcd * 990;
645 double * restrict HRR_INT__d_d_k_f = INT__d_d_k_f + real_abcd * 12960;
646
647 // form INT__d_p_k_s
648 for(iket = 0; iket < 36; ++iket)
649 {
650 HRR_INT__d_p_k_s[0 * 36 + iket] = HRR_INT__f_s_k_s[0 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[0 * 36 + iket] );
651
652 HRR_INT__d_p_k_s[1 * 36 + iket] = HRR_INT__f_s_k_s[1 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[0 * 36 + iket] );
653
654 HRR_INT__d_p_k_s[2 * 36 + iket] = HRR_INT__f_s_k_s[2 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[0 * 36 + iket] );
655
656 HRR_INT__d_p_k_s[3 * 36 + iket] = HRR_INT__f_s_k_s[1 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[1 * 36 + iket] );
657
658 HRR_INT__d_p_k_s[4 * 36 + iket] = HRR_INT__f_s_k_s[3 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[1 * 36 + iket] );
659
660 HRR_INT__d_p_k_s[5 * 36 + iket] = HRR_INT__f_s_k_s[4 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[1 * 36 + iket] );
661
662 HRR_INT__d_p_k_s[6 * 36 + iket] = HRR_INT__f_s_k_s[2 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[2 * 36 + iket] );
663
664 HRR_INT__d_p_k_s[7 * 36 + iket] = HRR_INT__f_s_k_s[4 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[2 * 36 + iket] );
665
666 HRR_INT__d_p_k_s[8 * 36 + iket] = HRR_INT__f_s_k_s[5 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[2 * 36 + iket] );
667
668 HRR_INT__d_p_k_s[9 * 36 + iket] = HRR_INT__f_s_k_s[3 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[3 * 36 + iket] );
669
670 HRR_INT__d_p_k_s[10 * 36 + iket] = HRR_INT__f_s_k_s[6 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[3 * 36 + iket] );
671
672 HRR_INT__d_p_k_s[11 * 36 + iket] = HRR_INT__f_s_k_s[7 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[3 * 36 + iket] );
673
674 HRR_INT__d_p_k_s[12 * 36 + iket] = HRR_INT__f_s_k_s[4 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[4 * 36 + iket] );
675
676 HRR_INT__d_p_k_s[13 * 36 + iket] = HRR_INT__f_s_k_s[7 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[4 * 36 + iket] );
677
678 HRR_INT__d_p_k_s[14 * 36 + iket] = HRR_INT__f_s_k_s[8 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[4 * 36 + iket] );
679
680 HRR_INT__d_p_k_s[15 * 36 + iket] = HRR_INT__f_s_k_s[5 * 36 + iket] + ( hAB[0] * HRR_INT__d_s_k_s[5 * 36 + iket] );
681
682 HRR_INT__d_p_k_s[16 * 36 + iket] = HRR_INT__f_s_k_s[8 * 36 + iket] + ( hAB[1] * HRR_INT__d_s_k_s[5 * 36 + iket] );
683
684 HRR_INT__d_p_k_s[17 * 36 + iket] = HRR_INT__f_s_k_s[9 * 36 + iket] + ( hAB[2] * HRR_INT__d_s_k_s[5 * 36 + iket] );
685
686 }
687
688
689 // form INT__d_p_l_s
690 for(iket = 0; iket < 45; ++iket)
691 {
692 HRR_INT__d_p_l_s[0 * 45 + iket] = HRR_INT__f_s_l_s[0 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[0 * 45 + iket] );
693
694 HRR_INT__d_p_l_s[1 * 45 + iket] = HRR_INT__f_s_l_s[1 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[0 * 45 + iket] );
695
696 HRR_INT__d_p_l_s[2 * 45 + iket] = HRR_INT__f_s_l_s[2 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[0 * 45 + iket] );
697
698 HRR_INT__d_p_l_s[3 * 45 + iket] = HRR_INT__f_s_l_s[1 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[1 * 45 + iket] );
699
700 HRR_INT__d_p_l_s[4 * 45 + iket] = HRR_INT__f_s_l_s[3 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[1 * 45 + iket] );
701
702 HRR_INT__d_p_l_s[5 * 45 + iket] = HRR_INT__f_s_l_s[4 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[1 * 45 + iket] );
703
704 HRR_INT__d_p_l_s[6 * 45 + iket] = HRR_INT__f_s_l_s[2 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[2 * 45 + iket] );
705
706 HRR_INT__d_p_l_s[7 * 45 + iket] = HRR_INT__f_s_l_s[4 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[2 * 45 + iket] );
707
708 HRR_INT__d_p_l_s[8 * 45 + iket] = HRR_INT__f_s_l_s[5 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[2 * 45 + iket] );
709
710 HRR_INT__d_p_l_s[9 * 45 + iket] = HRR_INT__f_s_l_s[3 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[3 * 45 + iket] );
711
712 HRR_INT__d_p_l_s[10 * 45 + iket] = HRR_INT__f_s_l_s[6 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[3 * 45 + iket] );
713
714 HRR_INT__d_p_l_s[11 * 45 + iket] = HRR_INT__f_s_l_s[7 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[3 * 45 + iket] );
715
716 HRR_INT__d_p_l_s[12 * 45 + iket] = HRR_INT__f_s_l_s[4 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[4 * 45 + iket] );
717
718 HRR_INT__d_p_l_s[13 * 45 + iket] = HRR_INT__f_s_l_s[7 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[4 * 45 + iket] );
719
720 HRR_INT__d_p_l_s[14 * 45 + iket] = HRR_INT__f_s_l_s[8 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[4 * 45 + iket] );
721
722 HRR_INT__d_p_l_s[15 * 45 + iket] = HRR_INT__f_s_l_s[5 * 45 + iket] + ( hAB[0] * HRR_INT__d_s_l_s[5 * 45 + iket] );
723
724 HRR_INT__d_p_l_s[16 * 45 + iket] = HRR_INT__f_s_l_s[8 * 45 + iket] + ( hAB[1] * HRR_INT__d_s_l_s[5 * 45 + iket] );
725
726 HRR_INT__d_p_l_s[17 * 45 + iket] = HRR_INT__f_s_l_s[9 * 45 + iket] + ( hAB[2] * HRR_INT__d_s_l_s[5 * 45 + iket] );
727
728 }
729
730
731 // form INT__d_p_m_s
732 for(iket = 0; iket < 55; ++iket)
733 {
734 HRR_INT__d_p_m_s[0 * 55 + iket] = HRR_INT__f_s_m_s[0 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[0 * 55 + iket] );
735
736 HRR_INT__d_p_m_s[1 * 55 + iket] = HRR_INT__f_s_m_s[1 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[0 * 55 + iket] );
737
738 HRR_INT__d_p_m_s[2 * 55 + iket] = HRR_INT__f_s_m_s[2 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[0 * 55 + iket] );
739
740 HRR_INT__d_p_m_s[3 * 55 + iket] = HRR_INT__f_s_m_s[1 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[1 * 55 + iket] );
741
742 HRR_INT__d_p_m_s[4 * 55 + iket] = HRR_INT__f_s_m_s[3 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[1 * 55 + iket] );
743
744 HRR_INT__d_p_m_s[5 * 55 + iket] = HRR_INT__f_s_m_s[4 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[1 * 55 + iket] );
745
746 HRR_INT__d_p_m_s[6 * 55 + iket] = HRR_INT__f_s_m_s[2 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[2 * 55 + iket] );
747
748 HRR_INT__d_p_m_s[7 * 55 + iket] = HRR_INT__f_s_m_s[4 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[2 * 55 + iket] );
749
750 HRR_INT__d_p_m_s[8 * 55 + iket] = HRR_INT__f_s_m_s[5 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[2 * 55 + iket] );
751
752 HRR_INT__d_p_m_s[9 * 55 + iket] = HRR_INT__f_s_m_s[3 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[3 * 55 + iket] );
753
754 HRR_INT__d_p_m_s[10 * 55 + iket] = HRR_INT__f_s_m_s[6 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[3 * 55 + iket] );
755
756 HRR_INT__d_p_m_s[11 * 55 + iket] = HRR_INT__f_s_m_s[7 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[3 * 55 + iket] );
757
758 HRR_INT__d_p_m_s[12 * 55 + iket] = HRR_INT__f_s_m_s[4 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[4 * 55 + iket] );
759
760 HRR_INT__d_p_m_s[13 * 55 + iket] = HRR_INT__f_s_m_s[7 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[4 * 55 + iket] );
761
762 HRR_INT__d_p_m_s[14 * 55 + iket] = HRR_INT__f_s_m_s[8 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[4 * 55 + iket] );
763
764 HRR_INT__d_p_m_s[15 * 55 + iket] = HRR_INT__f_s_m_s[5 * 55 + iket] + ( hAB[0] * HRR_INT__d_s_m_s[5 * 55 + iket] );
765
766 HRR_INT__d_p_m_s[16 * 55 + iket] = HRR_INT__f_s_m_s[8 * 55 + iket] + ( hAB[1] * HRR_INT__d_s_m_s[5 * 55 + iket] );
767
768 HRR_INT__d_p_m_s[17 * 55 + iket] = HRR_INT__f_s_m_s[9 * 55 + iket] + ( hAB[2] * HRR_INT__d_s_m_s[5 * 55 + iket] );
769
770 }
771
772
773 // form INT__d_p_n_s
774 for(iket = 0; iket < 66; ++iket)
775 {
776 HRR_INT__d_p_n_s[0 * 66 + iket] = HRR_INT__f_s_n_s[0 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[0 * 66 + iket] );
777
778 HRR_INT__d_p_n_s[1 * 66 + iket] = HRR_INT__f_s_n_s[1 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[0 * 66 + iket] );
779
780 HRR_INT__d_p_n_s[2 * 66 + iket] = HRR_INT__f_s_n_s[2 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[0 * 66 + iket] );
781
782 HRR_INT__d_p_n_s[3 * 66 + iket] = HRR_INT__f_s_n_s[1 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[1 * 66 + iket] );
783
784 HRR_INT__d_p_n_s[4 * 66 + iket] = HRR_INT__f_s_n_s[3 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[1 * 66 + iket] );
785
786 HRR_INT__d_p_n_s[5 * 66 + iket] = HRR_INT__f_s_n_s[4 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[1 * 66 + iket] );
787
788 HRR_INT__d_p_n_s[6 * 66 + iket] = HRR_INT__f_s_n_s[2 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[2 * 66 + iket] );
789
790 HRR_INT__d_p_n_s[7 * 66 + iket] = HRR_INT__f_s_n_s[4 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[2 * 66 + iket] );
791
792 HRR_INT__d_p_n_s[8 * 66 + iket] = HRR_INT__f_s_n_s[5 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[2 * 66 + iket] );
793
794 HRR_INT__d_p_n_s[9 * 66 + iket] = HRR_INT__f_s_n_s[3 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[3 * 66 + iket] );
795
796 HRR_INT__d_p_n_s[10 * 66 + iket] = HRR_INT__f_s_n_s[6 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[3 * 66 + iket] );
797
798 HRR_INT__d_p_n_s[11 * 66 + iket] = HRR_INT__f_s_n_s[7 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[3 * 66 + iket] );
799
800 HRR_INT__d_p_n_s[12 * 66 + iket] = HRR_INT__f_s_n_s[4 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[4 * 66 + iket] );
801
802 HRR_INT__d_p_n_s[13 * 66 + iket] = HRR_INT__f_s_n_s[7 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[4 * 66 + iket] );
803
804 HRR_INT__d_p_n_s[14 * 66 + iket] = HRR_INT__f_s_n_s[8 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[4 * 66 + iket] );
805
806 HRR_INT__d_p_n_s[15 * 66 + iket] = HRR_INT__f_s_n_s[5 * 66 + iket] + ( hAB[0] * HRR_INT__d_s_n_s[5 * 66 + iket] );
807
808 HRR_INT__d_p_n_s[16 * 66 + iket] = HRR_INT__f_s_n_s[8 * 66 + iket] + ( hAB[1] * HRR_INT__d_s_n_s[5 * 66 + iket] );
809
810 HRR_INT__d_p_n_s[17 * 66 + iket] = HRR_INT__f_s_n_s[9 * 66 + iket] + ( hAB[2] * HRR_INT__d_s_n_s[5 * 66 + iket] );
811
812 }
813
814
815 // form INT__f_p_k_s
816 HRR_J_f_p(
817 HRR_INT__f_p_k_s,
818 HRR_INT__f_s_k_s,
819 HRR_INT__g_s_k_s,
820 hAB, 36);
821
822 // form INT__f_p_l_s
823 HRR_J_f_p(
824 HRR_INT__f_p_l_s,
825 HRR_INT__f_s_l_s,
826 HRR_INT__g_s_l_s,
827 hAB, 45);
828
829 // form INT__f_p_m_s
830 HRR_J_f_p(
831 HRR_INT__f_p_m_s,
832 HRR_INT__f_s_m_s,
833 HRR_INT__g_s_m_s,
834 hAB, 55);
835
836 // form INT__f_p_n_s
837 HRR_J_f_p(
838 HRR_INT__f_p_n_s,
839 HRR_INT__f_s_n_s,
840 HRR_INT__g_s_n_s,
841 hAB, 66);
842
843 // form INT__d_d_k_s
844 HRR_J_d_d(
845 HRR_INT__d_d_k_s,
846 HRR_INT__d_p_k_s,
847 HRR_INT__f_p_k_s,
848 hAB, 36);
849
850 // form INT__d_d_l_s
851 HRR_J_d_d(
852 HRR_INT__d_d_l_s,
853 HRR_INT__d_p_l_s,
854 HRR_INT__f_p_l_s,
855 hAB, 45);
856
857 // form INT__d_d_m_s
858 HRR_J_d_d(
859 HRR_INT__d_d_m_s,
860 HRR_INT__d_p_m_s,
861 HRR_INT__f_p_m_s,
862 hAB, 55);
863
864 // form INT__d_d_n_s
865 HRR_J_d_d(
866 HRR_INT__d_d_n_s,
867 HRR_INT__d_p_n_s,
868 HRR_INT__f_p_n_s,
869 hAB, 66);
870
871 // form INT__d_d_k_p
872 ostei_general_hrr_L(2, 2, 7, 1, hCD, HRR_INT__d_d_l_s, HRR_INT__d_d_k_s, HRR_INT__d_d_k_p);
873
874 // form INT__d_d_l_p
875 ostei_general_hrr_L(2, 2, 8, 1, hCD, HRR_INT__d_d_m_s, HRR_INT__d_d_l_s, HRR_INT__d_d_l_p);
876
877 // form INT__d_d_m_p
878 ostei_general_hrr_L(2, 2, 9, 1, hCD, HRR_INT__d_d_n_s, HRR_INT__d_d_m_s, HRR_INT__d_d_m_p);
879
880 // form INT__d_d_k_d
881 ostei_general_hrr_L(2, 2, 7, 2, hCD, HRR_INT__d_d_l_p, HRR_INT__d_d_k_p, HRR_INT__d_d_k_d);
882
883 // form INT__d_d_l_d
884 ostei_general_hrr_L(2, 2, 8, 2, hCD, HRR_INT__d_d_m_p, HRR_INT__d_d_l_p, HRR_INT__d_d_l_d);
885
886 // form INT__d_d_k_f
887 ostei_general_hrr_L(2, 2, 7, 3, hCD, HRR_INT__d_d_l_d, HRR_INT__d_d_k_d, HRR_INT__d_d_k_f);
888
889
890 } // close HRR loop
891
892
893 } // close loop cdbatch
894
895 istart = iend;
896 } // close loop over ab
897
898 return P.nshell12_clip * Q.nshell12_clip;
899 }
900
ostei_d_d_f_k(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__d_d_f_k)901 int ostei_d_d_f_k(struct simint_multi_shellpair const P,
902 struct simint_multi_shellpair const Q,
903 double screen_tol,
904 double * const restrict work,
905 double * const restrict INT__d_d_f_k)
906 {
907 double Q_AB[3*Q.nshell12];
908 struct simint_multi_shellpair Q_tmp = Q;
909 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
910 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
911 Q_tmp.AB_x = Q_AB;
912 Q_tmp.AB_y = Q_AB + Q.nshell12;
913 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
914
915 for(int i = 0; i < Q.nshell12; i++)
916 {
917 Q_tmp.AB_x[i] = -Q.AB_x[i];
918 Q_tmp.AB_y[i] = -Q.AB_y[i];
919 Q_tmp.AB_z[i] = -Q.AB_z[i];
920 }
921
922 int ret = ostei_d_d_k_f(P, Q_tmp, screen_tol, work, INT__d_d_f_k);
923 double buffer[12960] SIMINT_ALIGN_ARRAY_DBL;
924
925 for(int q = 0; q < ret; q++)
926 {
927 int idx = 0;
928 for(int a = 0; a < 6; ++a)
929 for(int b = 0; b < 6; ++b)
930 for(int c = 0; c < 10; ++c)
931 for(int d = 0; d < 36; ++d)
932 buffer[idx++] = INT__d_d_f_k[q*12960+a*2160+b*360+d*10+c];
933
934 memcpy(INT__d_d_f_k+q*12960, buffer, 12960*sizeof(double));
935 }
936
937 return ret;
938 }
939
940