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_p_k_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_p_k_g)8 int ostei_d_p_k_g(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_p_k_g)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__d_p_k_g);
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__d_s_o_s = work + (SIMINT_NSHELL_SIMD * 1212);
34 double * const INT__f_s_k_s = work + (SIMINT_NSHELL_SIMD * 1680);
35 double * const INT__f_s_l_s = work + (SIMINT_NSHELL_SIMD * 2040);
36 double * const INT__f_s_m_s = work + (SIMINT_NSHELL_SIMD * 2490);
37 double * const INT__f_s_n_s = work + (SIMINT_NSHELL_SIMD * 3040);
38 double * const INT__f_s_o_s = work + (SIMINT_NSHELL_SIMD * 3700);
39 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*4480);
40 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
41 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_p_s = primwork + 15;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_d_s = primwork + 57;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_f_s = primwork + 135;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_g_s = primwork + 255;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_h_s = primwork + 420;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_i_s = primwork + 630;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_k_s = primwork + 882;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_l_s = primwork + 1170;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_m_s = primwork + 1485;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_n_s = primwork + 1815;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_o_s = primwork + 2145;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_h_s = primwork + 2457;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_i_s = primwork + 2646;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_k_s = primwork + 2898;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_l_s = primwork + 3222;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_m_s = primwork + 3627;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_n_s = primwork + 4122;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_o_s = primwork + 4716;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_i_s = primwork + 5418;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_k_s = primwork + 5754;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_l_s = primwork + 6186;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_m_s = primwork + 6726;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_n_s = primwork + 7386;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_o_s = primwork + 8178;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_k_s = primwork + 9114;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_l_s = primwork + 9474;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_m_s = primwork + 9924;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_n_s = primwork + 10474;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_o_s = primwork + 11134;
70 double * const hrrwork = (double *)(primwork + 11914);
71 double * const HRR_INT__d_p_k_s = hrrwork + 0;
72 double * const HRR_INT__d_p_k_p = hrrwork + 648;
73 double * const HRR_INT__d_p_k_d = hrrwork + 2592;
74 double * const HRR_INT__d_p_k_f = hrrwork + 6480;
75 double * const HRR_INT__d_p_l_s = hrrwork + 12960;
76 double * const HRR_INT__d_p_l_p = hrrwork + 13770;
77 double * const HRR_INT__d_p_l_d = hrrwork + 16200;
78 double * const HRR_INT__d_p_l_f = hrrwork + 21060;
79 double * const HRR_INT__d_p_m_s = hrrwork + 29160;
80 double * const HRR_INT__d_p_m_p = hrrwork + 30150;
81 double * const HRR_INT__d_p_m_d = hrrwork + 33120;
82 double * const HRR_INT__d_p_n_s = hrrwork + 39060;
83 double * const HRR_INT__d_p_n_p = hrrwork + 40248;
84 double * const HRR_INT__d_p_o_s = hrrwork + 43812;
85
86
87 // Create constants
88 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
89 const SIMINT_DBLTYPE const_10 = SIMINT_DBLSET1(10);
90 const SIMINT_DBLTYPE const_11 = SIMINT_DBLSET1(11);
91 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
92 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
93 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
94 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
95 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
96 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
97 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
98 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
99 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
100
101
102 ////////////////////////////////////////
103 // Loop over shells and primitives
104 ////////////////////////////////////////
105
106 real_abcd = 0;
107 istart = 0;
108 for(ab = 0; ab < P.nshell12_clip; ++ab)
109 {
110 const int iend = istart + P.nprim12[ab];
111
112 cd = 0;
113 jstart = 0;
114
115 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
116 {
117 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
118 int jend = jstart;
119 for(i = 0; i < nshellbatch; i++)
120 jend += Q.nprim12[cd+i];
121
122 // Clear the beginning of the workspace (where we are accumulating integrals)
123 memset(work, 0, SIMINT_NSHELL_SIMD * 4480 * sizeof(double));
124 abcd = 0;
125
126
127 for(i = istart; i < iend; ++i)
128 {
129 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
130
131 if(check_screen)
132 {
133 // Skip this whole thing if always insignificant
134 if((P.screen[i] * Q.screen_max) < screen_tol)
135 continue;
136 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
137 }
138
139 icd = 0;
140 iprimcd = 0;
141 nprim_icd = Q.nprim12[cd];
142 double * restrict PRIM_PTR_INT__d_s_k_s = INT__d_s_k_s + abcd * 216;
143 double * restrict PRIM_PTR_INT__d_s_l_s = INT__d_s_l_s + abcd * 270;
144 double * restrict PRIM_PTR_INT__d_s_m_s = INT__d_s_m_s + abcd * 330;
145 double * restrict PRIM_PTR_INT__d_s_n_s = INT__d_s_n_s + abcd * 396;
146 double * restrict PRIM_PTR_INT__d_s_o_s = INT__d_s_o_s + abcd * 468;
147 double * restrict PRIM_PTR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
148 double * restrict PRIM_PTR_INT__f_s_l_s = INT__f_s_l_s + abcd * 450;
149 double * restrict PRIM_PTR_INT__f_s_m_s = INT__f_s_m_s + abcd * 550;
150 double * restrict PRIM_PTR_INT__f_s_n_s = INT__f_s_n_s + abcd * 660;
151 double * restrict PRIM_PTR_INT__f_s_o_s = INT__f_s_o_s + abcd * 780;
152
153
154
155 // Load these one per loop over i
156 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
157 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
158 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
159
160 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
161
162 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
163 {
164 // calculate the shell offsets
165 // these are the offset from the shell pointed to by cd
166 // for each element
167 int shelloffsets[SIMINT_SIMD_LEN] = {0};
168 int lastoffset = 0;
169 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
170
171 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
172 {
173 // Handle if the first element of the vector is a new shell
174 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
175 {
176 nprim_icd += Q.nprim12[cd + (++icd)];
177 PRIM_PTR_INT__d_s_k_s += 216;
178 PRIM_PTR_INT__d_s_l_s += 270;
179 PRIM_PTR_INT__d_s_m_s += 330;
180 PRIM_PTR_INT__d_s_n_s += 396;
181 PRIM_PTR_INT__d_s_o_s += 468;
182 PRIM_PTR_INT__f_s_k_s += 360;
183 PRIM_PTR_INT__f_s_l_s += 450;
184 PRIM_PTR_INT__f_s_m_s += 550;
185 PRIM_PTR_INT__f_s_n_s += 660;
186 PRIM_PTR_INT__f_s_o_s += 780;
187 }
188 iprimcd++;
189 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
190 {
191 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
192 {
193 shelloffsets[n] = shelloffsets[n-1] + 1;
194 lastoffset++;
195 nprim_icd += Q.nprim12[cd + (++icd)];
196 }
197 else
198 shelloffsets[n] = shelloffsets[n-1];
199 iprimcd++;
200 }
201 }
202 else
203 iprimcd += SIMINT_SIMD_LEN;
204
205 // Do we have to compute this vector (or has it been screened out)?
206 // (not_screened != 0 means we have to do this vector)
207 if(check_screen)
208 {
209 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
210 if(vmax < screen_tol)
211 {
212 PRIM_PTR_INT__d_s_k_s += lastoffset*216;
213 PRIM_PTR_INT__d_s_l_s += lastoffset*270;
214 PRIM_PTR_INT__d_s_m_s += lastoffset*330;
215 PRIM_PTR_INT__d_s_n_s += lastoffset*396;
216 PRIM_PTR_INT__d_s_o_s += lastoffset*468;
217 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
218 PRIM_PTR_INT__f_s_l_s += lastoffset*450;
219 PRIM_PTR_INT__f_s_m_s += lastoffset*550;
220 PRIM_PTR_INT__f_s_n_s += lastoffset*660;
221 PRIM_PTR_INT__f_s_o_s += lastoffset*780;
222 continue;
223 }
224 }
225
226 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
227 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
228 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
229 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
230
231
232 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
233 SIMINT_DBLTYPE PQ[3];
234 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
235 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
236 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
237 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
238 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
239 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
240
241 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
242 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
243 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
244 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
245 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
246 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
247 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
248
249 // NOTE: Minus sign!
250 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
251 SIMINT_DBLTYPE aop_PQ[3];
252 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
253 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
254 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
255
256 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
257 SIMINT_DBLTYPE aoq_PQ[3];
258 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
259 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
260 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
261 // Put a minus sign here so we don't have to in RR routines
262 a_over_q = SIMINT_NEG(a_over_q);
263
264
265 //////////////////////////////////////////////
266 // Fjt function section
267 // Maximum v value: 14
268 //////////////////////////////////////////////
269 // The parameter to the Fjt function
270 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
271
272
273 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
274
275
276 boys_F_split(PRIM_INT__s_s_s_s, F_x, 14);
277 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
278 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
279 for(n = 0; n <= 14; n++)
280 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
281
282 //////////////////////////////////////////////
283 // Primitive integrals: Vertical recurrance
284 //////////////////////////////////////////////
285
286 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
287 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
288 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
289 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
290 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
291 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
292 const SIMINT_DBLTYPE vrr_const_5_over_2q = SIMINT_MUL(const_5, one_over_2q);
293 const SIMINT_DBLTYPE vrr_const_6_over_2q = SIMINT_MUL(const_6, one_over_2q);
294 const SIMINT_DBLTYPE vrr_const_7_over_2q = SIMINT_MUL(const_7, one_over_2q);
295 const SIMINT_DBLTYPE vrr_const_8_over_2q = SIMINT_MUL(const_8, one_over_2q);
296 const SIMINT_DBLTYPE vrr_const_9_over_2q = SIMINT_MUL(const_9, one_over_2q);
297 const SIMINT_DBLTYPE vrr_const_10_over_2q = SIMINT_MUL(const_10, one_over_2q);
298 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
299 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
300 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
301 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
302 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
303 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
304 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
305 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
306 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
307 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
308 const SIMINT_DBLTYPE vrr_const_11_over_2pq = SIMINT_MUL(const_11, one_over_2pq);
309
310
311
312 // Forming PRIM_INT__s_s_p_s[14 * 3];
313 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
314 {
315
316 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
317 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]);
318
319 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
320 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]);
321
322 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
323 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]);
324
325 }
326
327
328
329 // Forming PRIM_INT__s_s_d_s[13 * 6];
330 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
331 {
332
333 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
334 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]);
335 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]);
336
337 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 1]);
338 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]);
339 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]);
340
341 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
342 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]);
343 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]);
344
345 }
346
347
348
349 // Forming PRIM_INT__s_s_f_s[12 * 10];
350 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
351 {
352
353 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
354 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]);
355 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]);
356
357 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
358 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]);
359
360 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
361 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]);
362
363 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
364 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]);
365
366 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
367 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]);
368
369 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
370 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]);
371 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]);
372
373 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
374 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]);
375
376 PRIM_INT__s_s_f_s[n * 10 + 9] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
377 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]);
378 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]);
379
380 }
381
382
383 VRR_K_s_s_g_s(
384 PRIM_INT__s_s_g_s,
385 PRIM_INT__s_s_f_s,
386 PRIM_INT__s_s_d_s,
387 Q_PA,
388 a_over_q,
389 aoq_PQ,
390 one_over_2q,
391 11);
392
393
394 VRR_K_s_s_h_s(
395 PRIM_INT__s_s_h_s,
396 PRIM_INT__s_s_g_s,
397 PRIM_INT__s_s_f_s,
398 Q_PA,
399 a_over_q,
400 aoq_PQ,
401 one_over_2q,
402 10);
403
404
405 ostei_general_vrr1_K(6, 9,
406 one_over_2q, a_over_q, aoq_PQ, Q_PA,
407 PRIM_INT__s_s_h_s, PRIM_INT__s_s_g_s, PRIM_INT__s_s_i_s);
408
409
410 ostei_general_vrr1_K(7, 8,
411 one_over_2q, a_over_q, aoq_PQ, Q_PA,
412 PRIM_INT__s_s_i_s, PRIM_INT__s_s_h_s, PRIM_INT__s_s_k_s);
413
414
415 ostei_general_vrr_I(1, 0, 7, 0, 3,
416 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
417 PRIM_INT__s_s_k_s, NULL, NULL, PRIM_INT__s_s_i_s, NULL, PRIM_INT__p_s_k_s);
418
419
420 ostei_general_vrr_I(1, 0, 6, 0, 3,
421 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
422 PRIM_INT__s_s_i_s, NULL, NULL, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_i_s);
423
424
425 ostei_general_vrr_I(2, 0, 7, 0, 2,
426 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
427 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);
428
429
430 ostei_general_vrr1_K(8, 7,
431 one_over_2q, a_over_q, aoq_PQ, Q_PA,
432 PRIM_INT__s_s_k_s, PRIM_INT__s_s_i_s, PRIM_INT__s_s_l_s);
433
434
435 ostei_general_vrr_I(1, 0, 8, 0, 3,
436 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
437 PRIM_INT__s_s_l_s, NULL, NULL, PRIM_INT__s_s_k_s, NULL, PRIM_INT__p_s_l_s);
438
439
440 ostei_general_vrr_I(2, 0, 8, 0, 2,
441 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
442 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);
443
444
445 ostei_general_vrr_I(1, 0, 5, 0, 3,
446 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
447 PRIM_INT__s_s_h_s, NULL, NULL, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_h_s);
448
449
450 ostei_general_vrr_I(2, 0, 6, 0, 2,
451 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
452 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);
453
454
455 ostei_general_vrr_I(3, 0, 7, 0, 1,
456 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
457 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);
458
459
460 ostei_general_vrr1_K(9, 6,
461 one_over_2q, a_over_q, aoq_PQ, Q_PA,
462 PRIM_INT__s_s_l_s, PRIM_INT__s_s_k_s, PRIM_INT__s_s_m_s);
463
464
465 ostei_general_vrr_I(1, 0, 9, 0, 3,
466 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
467 PRIM_INT__s_s_m_s, NULL, NULL, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_m_s);
468
469
470 ostei_general_vrr_I(2, 0, 9, 0, 2,
471 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
472 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);
473
474
475 ostei_general_vrr_I(3, 0, 8, 0, 1,
476 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
477 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);
478
479
480 ostei_general_vrr1_K(10, 5,
481 one_over_2q, a_over_q, aoq_PQ, Q_PA,
482 PRIM_INT__s_s_m_s, PRIM_INT__s_s_l_s, PRIM_INT__s_s_n_s);
483
484
485 ostei_general_vrr_I(1, 0, 10, 0, 3,
486 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
487 PRIM_INT__s_s_n_s, NULL, NULL, PRIM_INT__s_s_m_s, NULL, PRIM_INT__p_s_n_s);
488
489
490 ostei_general_vrr_I(2, 0, 10, 0, 2,
491 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
492 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);
493
494
495 ostei_general_vrr_I(3, 0, 9, 0, 1,
496 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
497 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);
498
499
500 ostei_general_vrr1_K(11, 4,
501 one_over_2q, a_over_q, aoq_PQ, Q_PA,
502 PRIM_INT__s_s_n_s, PRIM_INT__s_s_m_s, PRIM_INT__s_s_o_s);
503
504
505 ostei_general_vrr_I(1, 0, 11, 0, 3,
506 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
507 PRIM_INT__s_s_o_s, NULL, NULL, PRIM_INT__s_s_n_s, NULL, PRIM_INT__p_s_o_s);
508
509
510 ostei_general_vrr_I(2, 0, 11, 0, 2,
511 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
512 PRIM_INT__p_s_o_s, PRIM_INT__s_s_o_s, NULL, PRIM_INT__p_s_n_s, NULL, PRIM_INT__d_s_o_s);
513
514
515 ostei_general_vrr_I(3, 0, 10, 0, 1,
516 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
517 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);
518
519
520 ostei_general_vrr_I(3, 0, 11, 0, 1,
521 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
522 PRIM_INT__d_s_o_s, PRIM_INT__p_s_o_s, NULL, PRIM_INT__d_s_n_s, NULL, PRIM_INT__f_s_o_s);
523
524
525
526
527 ////////////////////////////////////
528 // Accumulate contracted integrals
529 ////////////////////////////////////
530 if(lastoffset == 0)
531 {
532 contract_all(216, PRIM_INT__d_s_k_s, PRIM_PTR_INT__d_s_k_s);
533 contract_all(270, PRIM_INT__d_s_l_s, PRIM_PTR_INT__d_s_l_s);
534 contract_all(330, PRIM_INT__d_s_m_s, PRIM_PTR_INT__d_s_m_s);
535 contract_all(396, PRIM_INT__d_s_n_s, PRIM_PTR_INT__d_s_n_s);
536 contract_all(468, PRIM_INT__d_s_o_s, PRIM_PTR_INT__d_s_o_s);
537 contract_all(360, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
538 contract_all(450, PRIM_INT__f_s_l_s, PRIM_PTR_INT__f_s_l_s);
539 contract_all(550, PRIM_INT__f_s_m_s, PRIM_PTR_INT__f_s_m_s);
540 contract_all(660, PRIM_INT__f_s_n_s, PRIM_PTR_INT__f_s_n_s);
541 contract_all(780, PRIM_INT__f_s_o_s, PRIM_PTR_INT__f_s_o_s);
542 }
543 else
544 {
545 contract(216, shelloffsets, PRIM_INT__d_s_k_s, PRIM_PTR_INT__d_s_k_s);
546 contract(270, shelloffsets, PRIM_INT__d_s_l_s, PRIM_PTR_INT__d_s_l_s);
547 contract(330, shelloffsets, PRIM_INT__d_s_m_s, PRIM_PTR_INT__d_s_m_s);
548 contract(396, shelloffsets, PRIM_INT__d_s_n_s, PRIM_PTR_INT__d_s_n_s);
549 contract(468, shelloffsets, PRIM_INT__d_s_o_s, PRIM_PTR_INT__d_s_o_s);
550 contract(360, shelloffsets, PRIM_INT__f_s_k_s, PRIM_PTR_INT__f_s_k_s);
551 contract(450, shelloffsets, PRIM_INT__f_s_l_s, PRIM_PTR_INT__f_s_l_s);
552 contract(550, shelloffsets, PRIM_INT__f_s_m_s, PRIM_PTR_INT__f_s_m_s);
553 contract(660, shelloffsets, PRIM_INT__f_s_n_s, PRIM_PTR_INT__f_s_n_s);
554 contract(780, shelloffsets, PRIM_INT__f_s_o_s, PRIM_PTR_INT__f_s_o_s);
555 PRIM_PTR_INT__d_s_k_s += lastoffset*216;
556 PRIM_PTR_INT__d_s_l_s += lastoffset*270;
557 PRIM_PTR_INT__d_s_m_s += lastoffset*330;
558 PRIM_PTR_INT__d_s_n_s += lastoffset*396;
559 PRIM_PTR_INT__d_s_o_s += lastoffset*468;
560 PRIM_PTR_INT__f_s_k_s += lastoffset*360;
561 PRIM_PTR_INT__f_s_l_s += lastoffset*450;
562 PRIM_PTR_INT__f_s_m_s += lastoffset*550;
563 PRIM_PTR_INT__f_s_n_s += lastoffset*660;
564 PRIM_PTR_INT__f_s_o_s += lastoffset*780;
565 }
566
567 } // close loop over j
568 } // close loop over i
569
570 //Advance to the next batch
571 jstart = SIMINT_SIMD_ROUND(jend);
572
573 //////////////////////////////////////////////
574 // Contracted integrals: Horizontal recurrance
575 //////////////////////////////////////////////
576
577
578 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
579
580
581 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
582 {
583 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
584
585 // set up HRR pointers
586 double const * restrict HRR_INT__d_s_k_s = INT__d_s_k_s + abcd * 216;
587 double const * restrict HRR_INT__d_s_l_s = INT__d_s_l_s + abcd * 270;
588 double const * restrict HRR_INT__d_s_m_s = INT__d_s_m_s + abcd * 330;
589 double const * restrict HRR_INT__d_s_n_s = INT__d_s_n_s + abcd * 396;
590 double const * restrict HRR_INT__d_s_o_s = INT__d_s_o_s + abcd * 468;
591 double const * restrict HRR_INT__f_s_k_s = INT__f_s_k_s + abcd * 360;
592 double const * restrict HRR_INT__f_s_l_s = INT__f_s_l_s + abcd * 450;
593 double const * restrict HRR_INT__f_s_m_s = INT__f_s_m_s + abcd * 550;
594 double const * restrict HRR_INT__f_s_n_s = INT__f_s_n_s + abcd * 660;
595 double const * restrict HRR_INT__f_s_o_s = INT__f_s_o_s + abcd * 780;
596 double * restrict HRR_INT__d_p_k_g = INT__d_p_k_g + real_abcd * 9720;
597
598 // form INT__d_p_k_s
599 for(iket = 0; iket < 36; ++iket)
600 {
601 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] );
602
603 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] );
604
605 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] );
606
607 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] );
608
609 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] );
610
611 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] );
612
613 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] );
614
615 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] );
616
617 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] );
618
619 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] );
620
621 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] );
622
623 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] );
624
625 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] );
626
627 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] );
628
629 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] );
630
631 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] );
632
633 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] );
634
635 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] );
636
637 }
638
639
640 // form INT__d_p_l_s
641 for(iket = 0; iket < 45; ++iket)
642 {
643 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] );
644
645 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] );
646
647 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] );
648
649 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] );
650
651 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] );
652
653 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] );
654
655 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] );
656
657 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] );
658
659 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] );
660
661 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] );
662
663 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] );
664
665 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] );
666
667 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] );
668
669 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] );
670
671 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] );
672
673 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] );
674
675 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] );
676
677 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] );
678
679 }
680
681
682 // form INT__d_p_m_s
683 for(iket = 0; iket < 55; ++iket)
684 {
685 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] );
686
687 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] );
688
689 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] );
690
691 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] );
692
693 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] );
694
695 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] );
696
697 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] );
698
699 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] );
700
701 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] );
702
703 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] );
704
705 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] );
706
707 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] );
708
709 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] );
710
711 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] );
712
713 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] );
714
715 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] );
716
717 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] );
718
719 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] );
720
721 }
722
723
724 // form INT__d_p_n_s
725 for(iket = 0; iket < 66; ++iket)
726 {
727 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] );
728
729 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] );
730
731 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] );
732
733 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] );
734
735 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] );
736
737 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] );
738
739 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] );
740
741 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] );
742
743 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] );
744
745 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] );
746
747 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] );
748
749 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] );
750
751 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] );
752
753 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] );
754
755 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] );
756
757 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] );
758
759 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] );
760
761 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] );
762
763 }
764
765
766 // form INT__d_p_o_s
767 for(iket = 0; iket < 78; ++iket)
768 {
769 HRR_INT__d_p_o_s[0 * 78 + iket] = HRR_INT__f_s_o_s[0 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[0 * 78 + iket] );
770
771 HRR_INT__d_p_o_s[1 * 78 + iket] = HRR_INT__f_s_o_s[1 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[0 * 78 + iket] );
772
773 HRR_INT__d_p_o_s[2 * 78 + iket] = HRR_INT__f_s_o_s[2 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[0 * 78 + iket] );
774
775 HRR_INT__d_p_o_s[3 * 78 + iket] = HRR_INT__f_s_o_s[1 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[1 * 78 + iket] );
776
777 HRR_INT__d_p_o_s[4 * 78 + iket] = HRR_INT__f_s_o_s[3 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[1 * 78 + iket] );
778
779 HRR_INT__d_p_o_s[5 * 78 + iket] = HRR_INT__f_s_o_s[4 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[1 * 78 + iket] );
780
781 HRR_INT__d_p_o_s[6 * 78 + iket] = HRR_INT__f_s_o_s[2 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[2 * 78 + iket] );
782
783 HRR_INT__d_p_o_s[7 * 78 + iket] = HRR_INT__f_s_o_s[4 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[2 * 78 + iket] );
784
785 HRR_INT__d_p_o_s[8 * 78 + iket] = HRR_INT__f_s_o_s[5 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[2 * 78 + iket] );
786
787 HRR_INT__d_p_o_s[9 * 78 + iket] = HRR_INT__f_s_o_s[3 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[3 * 78 + iket] );
788
789 HRR_INT__d_p_o_s[10 * 78 + iket] = HRR_INT__f_s_o_s[6 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[3 * 78 + iket] );
790
791 HRR_INT__d_p_o_s[11 * 78 + iket] = HRR_INT__f_s_o_s[7 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[3 * 78 + iket] );
792
793 HRR_INT__d_p_o_s[12 * 78 + iket] = HRR_INT__f_s_o_s[4 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[4 * 78 + iket] );
794
795 HRR_INT__d_p_o_s[13 * 78 + iket] = HRR_INT__f_s_o_s[7 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[4 * 78 + iket] );
796
797 HRR_INT__d_p_o_s[14 * 78 + iket] = HRR_INT__f_s_o_s[8 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[4 * 78 + iket] );
798
799 HRR_INT__d_p_o_s[15 * 78 + iket] = HRR_INT__f_s_o_s[5 * 78 + iket] + ( hAB[0] * HRR_INT__d_s_o_s[5 * 78 + iket] );
800
801 HRR_INT__d_p_o_s[16 * 78 + iket] = HRR_INT__f_s_o_s[8 * 78 + iket] + ( hAB[1] * HRR_INT__d_s_o_s[5 * 78 + iket] );
802
803 HRR_INT__d_p_o_s[17 * 78 + iket] = HRR_INT__f_s_o_s[9 * 78 + iket] + ( hAB[2] * HRR_INT__d_s_o_s[5 * 78 + iket] );
804
805 }
806
807
808 // form INT__d_p_k_p
809 ostei_general_hrr_L(2, 1, 7, 1, hCD, HRR_INT__d_p_l_s, HRR_INT__d_p_k_s, HRR_INT__d_p_k_p);
810
811 // form INT__d_p_l_p
812 ostei_general_hrr_L(2, 1, 8, 1, hCD, HRR_INT__d_p_m_s, HRR_INT__d_p_l_s, HRR_INT__d_p_l_p);
813
814 // form INT__d_p_m_p
815 ostei_general_hrr_L(2, 1, 9, 1, hCD, HRR_INT__d_p_n_s, HRR_INT__d_p_m_s, HRR_INT__d_p_m_p);
816
817 // form INT__d_p_n_p
818 ostei_general_hrr_L(2, 1, 10, 1, hCD, HRR_INT__d_p_o_s, HRR_INT__d_p_n_s, HRR_INT__d_p_n_p);
819
820 // form INT__d_p_k_d
821 ostei_general_hrr_L(2, 1, 7, 2, hCD, HRR_INT__d_p_l_p, HRR_INT__d_p_k_p, HRR_INT__d_p_k_d);
822
823 // form INT__d_p_l_d
824 ostei_general_hrr_L(2, 1, 8, 2, hCD, HRR_INT__d_p_m_p, HRR_INT__d_p_l_p, HRR_INT__d_p_l_d);
825
826 // form INT__d_p_m_d
827 ostei_general_hrr_L(2, 1, 9, 2, hCD, HRR_INT__d_p_n_p, HRR_INT__d_p_m_p, HRR_INT__d_p_m_d);
828
829 // form INT__d_p_k_f
830 ostei_general_hrr_L(2, 1, 7, 3, hCD, HRR_INT__d_p_l_d, HRR_INT__d_p_k_d, HRR_INT__d_p_k_f);
831
832 // form INT__d_p_l_f
833 ostei_general_hrr_L(2, 1, 8, 3, hCD, HRR_INT__d_p_m_d, HRR_INT__d_p_l_d, HRR_INT__d_p_l_f);
834
835 // form INT__d_p_k_g
836 ostei_general_hrr_L(2, 1, 7, 4, hCD, HRR_INT__d_p_l_f, HRR_INT__d_p_k_f, HRR_INT__d_p_k_g);
837
838
839 } // close HRR loop
840
841
842 } // close loop cdbatch
843
844 istart = iend;
845 } // close loop over ab
846
847 return P.nshell12_clip * Q.nshell12_clip;
848 }
849
ostei_p_d_k_g(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__p_d_k_g)850 int ostei_p_d_k_g(struct simint_multi_shellpair const P,
851 struct simint_multi_shellpair const Q,
852 double screen_tol,
853 double * const restrict work,
854 double * const restrict INT__p_d_k_g)
855 {
856 double P_AB[3*P.nshell12];
857 struct simint_multi_shellpair P_tmp = P;
858 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
859 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
860 P_tmp.AB_x = P_AB;
861 P_tmp.AB_y = P_AB + P.nshell12;
862 P_tmp.AB_z = P_AB + 2*P.nshell12;
863
864 for(int i = 0; i < P.nshell12; i++)
865 {
866 P_tmp.AB_x[i] = -P.AB_x[i];
867 P_tmp.AB_y[i] = -P.AB_y[i];
868 P_tmp.AB_z[i] = -P.AB_z[i];
869 }
870
871 int ret = ostei_d_p_k_g(P_tmp, Q, screen_tol, work, INT__p_d_k_g);
872 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
873
874 for(int q = 0; q < ret; q++)
875 {
876 int idx = 0;
877 for(int a = 0; a < 3; ++a)
878 for(int b = 0; b < 6; ++b)
879 for(int c = 0; c < 36; ++c)
880 for(int d = 0; d < 15; ++d)
881 buffer[idx++] = INT__p_d_k_g[q*9720+b*1620+a*540+c*15+d];
882
883 memcpy(INT__p_d_k_g+q*9720, buffer, 9720*sizeof(double));
884 }
885
886 return ret;
887 }
888
ostei_d_p_g_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_p_g_k)889 int ostei_d_p_g_k(struct simint_multi_shellpair const P,
890 struct simint_multi_shellpair const Q,
891 double screen_tol,
892 double * const restrict work,
893 double * const restrict INT__d_p_g_k)
894 {
895 double Q_AB[3*Q.nshell12];
896 struct simint_multi_shellpair Q_tmp = Q;
897 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
898 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
899 Q_tmp.AB_x = Q_AB;
900 Q_tmp.AB_y = Q_AB + Q.nshell12;
901 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
902
903 for(int i = 0; i < Q.nshell12; i++)
904 {
905 Q_tmp.AB_x[i] = -Q.AB_x[i];
906 Q_tmp.AB_y[i] = -Q.AB_y[i];
907 Q_tmp.AB_z[i] = -Q.AB_z[i];
908 }
909
910 int ret = ostei_d_p_k_g(P, Q_tmp, screen_tol, work, INT__d_p_g_k);
911 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
912
913 for(int q = 0; q < ret; q++)
914 {
915 int idx = 0;
916 for(int a = 0; a < 6; ++a)
917 for(int b = 0; b < 3; ++b)
918 for(int c = 0; c < 15; ++c)
919 for(int d = 0; d < 36; ++d)
920 buffer[idx++] = INT__d_p_g_k[q*9720+a*1620+b*540+d*15+c];
921
922 memcpy(INT__d_p_g_k+q*9720, buffer, 9720*sizeof(double));
923 }
924
925 return ret;
926 }
927
ostei_p_d_g_k(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__p_d_g_k)928 int ostei_p_d_g_k(struct simint_multi_shellpair const P,
929 struct simint_multi_shellpair const Q,
930 double screen_tol,
931 double * const restrict work,
932 double * const restrict INT__p_d_g_k)
933 {
934 double P_AB[3*P.nshell12];
935 struct simint_multi_shellpair P_tmp = P;
936 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
937 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
938 P_tmp.AB_x = P_AB;
939 P_tmp.AB_y = P_AB + P.nshell12;
940 P_tmp.AB_z = P_AB + 2*P.nshell12;
941
942 for(int i = 0; i < P.nshell12; i++)
943 {
944 P_tmp.AB_x[i] = -P.AB_x[i];
945 P_tmp.AB_y[i] = -P.AB_y[i];
946 P_tmp.AB_z[i] = -P.AB_z[i];
947 }
948
949 double Q_AB[3*Q.nshell12];
950 struct simint_multi_shellpair Q_tmp = Q;
951 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
952 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
953 Q_tmp.AB_x = Q_AB;
954 Q_tmp.AB_y = Q_AB + Q.nshell12;
955 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
956
957 for(int i = 0; i < Q.nshell12; i++)
958 {
959 Q_tmp.AB_x[i] = -Q.AB_x[i];
960 Q_tmp.AB_y[i] = -Q.AB_y[i];
961 Q_tmp.AB_z[i] = -Q.AB_z[i];
962 }
963
964 int ret = ostei_d_p_k_g(P_tmp, Q_tmp, screen_tol, work, INT__p_d_g_k);
965 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
966
967 for(int q = 0; q < ret; q++)
968 {
969 int idx = 0;
970 for(int a = 0; a < 3; ++a)
971 for(int b = 0; b < 6; ++b)
972 for(int c = 0; c < 15; ++c)
973 for(int d = 0; d < 36; ++d)
974 buffer[idx++] = INT__p_d_g_k[q*9720+b*1620+a*540+d*15+c];
975
976 memcpy(INT__p_d_g_k+q*9720, buffer, 9720*sizeof(double));
977 }
978
979 return ret;
980 }
981
982