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_k_i_p_p(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__k_i_p_p)8 int ostei_k_i_p_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__k_i_p_p)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__k_i_p_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__k_s_p_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__k_s_d_s = work + (SIMINT_NSHELL_SIMD * 108);
31 double * const INT__l_s_p_s = work + (SIMINT_NSHELL_SIMD * 324);
32 double * const INT__l_s_d_s = work + (SIMINT_NSHELL_SIMD * 459);
33 double * const INT__m_s_p_s = work + (SIMINT_NSHELL_SIMD * 729);
34 double * const INT__m_s_d_s = work + (SIMINT_NSHELL_SIMD * 894);
35 double * const INT__n_s_p_s = work + (SIMINT_NSHELL_SIMD * 1224);
36 double * const INT__n_s_d_s = work + (SIMINT_NSHELL_SIMD * 1422);
37 double * const INT__o_s_p_s = work + (SIMINT_NSHELL_SIMD * 1818);
38 double * const INT__o_s_d_s = work + (SIMINT_NSHELL_SIMD * 2052);
39 double * const INT__q_s_p_s = work + (SIMINT_NSHELL_SIMD * 2520);
40 double * const INT__q_s_d_s = work + (SIMINT_NSHELL_SIMD * 2793);
41 double * const INT__r_s_p_s = work + (SIMINT_NSHELL_SIMD * 3339);
42 double * const INT__r_s_d_s = work + (SIMINT_NSHELL_SIMD * 3654);
43 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*4284);
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_s_s = primwork + 16;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_s_s = primwork + 61;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_s_s = primwork + 145;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_s_s = primwork + 275;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_s_s = primwork + 455;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_s_s = primwork + 686;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_p_s = primwork + 966;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_s_s = primwork + 1134;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_p_s = primwork + 1458;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_d_s = primwork + 1674;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_s_s = primwork + 1890;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_p_s = primwork + 2250;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_d_s = primwork + 2520;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_s_s = primwork + 2790;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_p_s = primwork + 3175;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_d_s = primwork + 3505;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_s_s = primwork + 3835;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_p_s = primwork + 4231;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_d_s = primwork + 4627;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_s_s = primwork + 5023;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_p_s = primwork + 5413;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_d_s = primwork + 5881;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__q_s_s_s = primwork + 6349;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__q_s_p_s = primwork + 6713;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__q_s_d_s = primwork + 7259;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__r_s_s_s = primwork + 7805;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__r_s_p_s = primwork + 8120;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__r_s_d_s = primwork + 8750;
73 double * const hrrwork = (double *)(primwork + 9380);
74 double * const HRR_INT__k_p_p_s = hrrwork + 0;
75 double * const HRR_INT__k_p_d_s = hrrwork + 324;
76 double * const HRR_INT__k_d_p_s = hrrwork + 972;
77 double * const HRR_INT__k_d_d_s = hrrwork + 1620;
78 double * const HRR_INT__k_f_p_s = hrrwork + 2916;
79 double * const HRR_INT__k_f_d_s = hrrwork + 3996;
80 double * const HRR_INT__k_g_p_s = hrrwork + 6156;
81 double * const HRR_INT__k_g_d_s = hrrwork + 7776;
82 double * const HRR_INT__k_h_p_s = hrrwork + 11016;
83 double * const HRR_INT__k_h_d_s = hrrwork + 13284;
84 double * const HRR_INT__k_i_p_s = hrrwork + 17820;
85 double * const HRR_INT__k_i_d_s = hrrwork + 20844;
86 double * const HRR_INT__l_p_p_s = hrrwork + 26892;
87 double * const HRR_INT__l_p_d_s = hrrwork + 27297;
88 double * const HRR_INT__l_d_p_s = hrrwork + 28107;
89 double * const HRR_INT__l_d_d_s = hrrwork + 28917;
90 double * const HRR_INT__l_f_p_s = hrrwork + 30537;
91 double * const HRR_INT__l_f_d_s = hrrwork + 31887;
92 double * const HRR_INT__l_g_p_s = hrrwork + 34587;
93 double * const HRR_INT__l_g_d_s = hrrwork + 36612;
94 double * const HRR_INT__l_h_p_s = hrrwork + 40662;
95 double * const HRR_INT__l_h_d_s = hrrwork + 43497;
96 double * const HRR_INT__m_p_p_s = hrrwork + 49167;
97 double * const HRR_INT__m_p_d_s = hrrwork + 49662;
98 double * const HRR_INT__m_d_p_s = hrrwork + 50652;
99 double * const HRR_INT__m_d_d_s = hrrwork + 51642;
100 double * const HRR_INT__m_f_p_s = hrrwork + 53622;
101 double * const HRR_INT__m_f_d_s = hrrwork + 55272;
102 double * const HRR_INT__m_g_p_s = hrrwork + 58572;
103 double * const HRR_INT__m_g_d_s = hrrwork + 61047;
104 double * const HRR_INT__n_p_p_s = hrrwork + 65997;
105 double * const HRR_INT__n_p_d_s = hrrwork + 66591;
106 double * const HRR_INT__n_d_p_s = hrrwork + 67779;
107 double * const HRR_INT__n_d_d_s = hrrwork + 68967;
108 double * const HRR_INT__n_f_p_s = hrrwork + 71343;
109 double * const HRR_INT__n_f_d_s = hrrwork + 73323;
110 double * const HRR_INT__o_p_p_s = hrrwork + 77283;
111 double * const HRR_INT__o_p_d_s = hrrwork + 77985;
112 double * const HRR_INT__o_d_p_s = hrrwork + 79389;
113 double * const HRR_INT__o_d_d_s = hrrwork + 80793;
114 double * const HRR_INT__q_p_p_s = hrrwork + 83601;
115 double * const HRR_INT__q_p_d_s = hrrwork + 84420;
116
117
118 // Create constants
119 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
120 const SIMINT_DBLTYPE const_10 = SIMINT_DBLSET1(10);
121 const SIMINT_DBLTYPE const_11 = SIMINT_DBLSET1(11);
122 const SIMINT_DBLTYPE const_12 = SIMINT_DBLSET1(12);
123 const SIMINT_DBLTYPE const_13 = SIMINT_DBLSET1(13);
124 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
125 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
126 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
127 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
128 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
129 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
130 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
131 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
132 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
133
134
135 ////////////////////////////////////////
136 // Loop over shells and primitives
137 ////////////////////////////////////////
138
139 real_abcd = 0;
140 istart = 0;
141 for(ab = 0; ab < P.nshell12_clip; ++ab)
142 {
143 const int iend = istart + P.nprim12[ab];
144
145 cd = 0;
146 jstart = 0;
147
148 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
149 {
150 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
151 int jend = jstart;
152 for(i = 0; i < nshellbatch; i++)
153 jend += Q.nprim12[cd+i];
154
155 // Clear the beginning of the workspace (where we are accumulating integrals)
156 memset(work, 0, SIMINT_NSHELL_SIMD * 4284 * sizeof(double));
157 abcd = 0;
158
159
160 for(i = istart; i < iend; ++i)
161 {
162 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
163
164 if(check_screen)
165 {
166 // Skip this whole thing if always insignificant
167 if((P.screen[i] * Q.screen_max) < screen_tol)
168 continue;
169 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
170 }
171
172 icd = 0;
173 iprimcd = 0;
174 nprim_icd = Q.nprim12[cd];
175 double * restrict PRIM_PTR_INT__k_s_p_s = INT__k_s_p_s + abcd * 108;
176 double * restrict PRIM_PTR_INT__k_s_d_s = INT__k_s_d_s + abcd * 216;
177 double * restrict PRIM_PTR_INT__l_s_p_s = INT__l_s_p_s + abcd * 135;
178 double * restrict PRIM_PTR_INT__l_s_d_s = INT__l_s_d_s + abcd * 270;
179 double * restrict PRIM_PTR_INT__m_s_p_s = INT__m_s_p_s + abcd * 165;
180 double * restrict PRIM_PTR_INT__m_s_d_s = INT__m_s_d_s + abcd * 330;
181 double * restrict PRIM_PTR_INT__n_s_p_s = INT__n_s_p_s + abcd * 198;
182 double * restrict PRIM_PTR_INT__n_s_d_s = INT__n_s_d_s + abcd * 396;
183 double * restrict PRIM_PTR_INT__o_s_p_s = INT__o_s_p_s + abcd * 234;
184 double * restrict PRIM_PTR_INT__o_s_d_s = INT__o_s_d_s + abcd * 468;
185 double * restrict PRIM_PTR_INT__q_s_p_s = INT__q_s_p_s + abcd * 273;
186 double * restrict PRIM_PTR_INT__q_s_d_s = INT__q_s_d_s + abcd * 546;
187 double * restrict PRIM_PTR_INT__r_s_p_s = INT__r_s_p_s + abcd * 315;
188 double * restrict PRIM_PTR_INT__r_s_d_s = INT__r_s_d_s + abcd * 630;
189
190
191
192 // Load these one per loop over i
193 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
194 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
195 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
196
197 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
198
199 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
200 {
201 // calculate the shell offsets
202 // these are the offset from the shell pointed to by cd
203 // for each element
204 int shelloffsets[SIMINT_SIMD_LEN] = {0};
205 int lastoffset = 0;
206 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
207
208 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
209 {
210 // Handle if the first element of the vector is a new shell
211 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
212 {
213 nprim_icd += Q.nprim12[cd + (++icd)];
214 PRIM_PTR_INT__k_s_p_s += 108;
215 PRIM_PTR_INT__k_s_d_s += 216;
216 PRIM_PTR_INT__l_s_p_s += 135;
217 PRIM_PTR_INT__l_s_d_s += 270;
218 PRIM_PTR_INT__m_s_p_s += 165;
219 PRIM_PTR_INT__m_s_d_s += 330;
220 PRIM_PTR_INT__n_s_p_s += 198;
221 PRIM_PTR_INT__n_s_d_s += 396;
222 PRIM_PTR_INT__o_s_p_s += 234;
223 PRIM_PTR_INT__o_s_d_s += 468;
224 PRIM_PTR_INT__q_s_p_s += 273;
225 PRIM_PTR_INT__q_s_d_s += 546;
226 PRIM_PTR_INT__r_s_p_s += 315;
227 PRIM_PTR_INT__r_s_d_s += 630;
228 }
229 iprimcd++;
230 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
231 {
232 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
233 {
234 shelloffsets[n] = shelloffsets[n-1] + 1;
235 lastoffset++;
236 nprim_icd += Q.nprim12[cd + (++icd)];
237 }
238 else
239 shelloffsets[n] = shelloffsets[n-1];
240 iprimcd++;
241 }
242 }
243 else
244 iprimcd += SIMINT_SIMD_LEN;
245
246 // Do we have to compute this vector (or has it been screened out)?
247 // (not_screened != 0 means we have to do this vector)
248 if(check_screen)
249 {
250 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
251 if(vmax < screen_tol)
252 {
253 PRIM_PTR_INT__k_s_p_s += lastoffset*108;
254 PRIM_PTR_INT__k_s_d_s += lastoffset*216;
255 PRIM_PTR_INT__l_s_p_s += lastoffset*135;
256 PRIM_PTR_INT__l_s_d_s += lastoffset*270;
257 PRIM_PTR_INT__m_s_p_s += lastoffset*165;
258 PRIM_PTR_INT__m_s_d_s += lastoffset*330;
259 PRIM_PTR_INT__n_s_p_s += lastoffset*198;
260 PRIM_PTR_INT__n_s_d_s += lastoffset*396;
261 PRIM_PTR_INT__o_s_p_s += lastoffset*234;
262 PRIM_PTR_INT__o_s_d_s += lastoffset*468;
263 PRIM_PTR_INT__q_s_p_s += lastoffset*273;
264 PRIM_PTR_INT__q_s_d_s += lastoffset*546;
265 PRIM_PTR_INT__r_s_p_s += lastoffset*315;
266 PRIM_PTR_INT__r_s_d_s += lastoffset*630;
267 continue;
268 }
269 }
270
271 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
272 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
273 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
274 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
275
276
277 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
278 SIMINT_DBLTYPE PQ[3];
279 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
280 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
281 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
282 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
283 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
284 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
285
286 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
287 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
288 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
289 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
290 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
291 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
292 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
293
294 // NOTE: Minus sign!
295 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
296 SIMINT_DBLTYPE aop_PQ[3];
297 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
298 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
299 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
300
301 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
302 SIMINT_DBLTYPE aoq_PQ[3];
303 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
304 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
305 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
306 // Put a minus sign here so we don't have to in RR routines
307 a_over_q = SIMINT_NEG(a_over_q);
308
309
310 //////////////////////////////////////////////
311 // Fjt function section
312 // Maximum v value: 15
313 //////////////////////////////////////////////
314 // The parameter to the Fjt function
315 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
316
317
318 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
319
320
321 boys_F_split(PRIM_INT__s_s_s_s, F_x, 15);
322 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
323 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
324 for(n = 0; n <= 15; n++)
325 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
326
327 //////////////////////////////////////////////
328 // Primitive integrals: Vertical recurrance
329 //////////////////////////////////////////////
330
331 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
332 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
333 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
334 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
335 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
336 const SIMINT_DBLTYPE vrr_const_6_over_2p = SIMINT_MUL(const_6, one_over_2p);
337 const SIMINT_DBLTYPE vrr_const_7_over_2p = SIMINT_MUL(const_7, one_over_2p);
338 const SIMINT_DBLTYPE vrr_const_8_over_2p = SIMINT_MUL(const_8, one_over_2p);
339 const SIMINT_DBLTYPE vrr_const_9_over_2p = SIMINT_MUL(const_9, one_over_2p);
340 const SIMINT_DBLTYPE vrr_const_10_over_2p = SIMINT_MUL(const_10, one_over_2p);
341 const SIMINT_DBLTYPE vrr_const_11_over_2p = SIMINT_MUL(const_11, one_over_2p);
342 const SIMINT_DBLTYPE vrr_const_12_over_2p = SIMINT_MUL(const_12, one_over_2p);
343 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
344 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
345 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
346 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
347 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
348 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
349 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
350 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
351 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
352 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
353 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
354 const SIMINT_DBLTYPE vrr_const_11_over_2pq = SIMINT_MUL(const_11, one_over_2pq);
355 const SIMINT_DBLTYPE vrr_const_12_over_2pq = SIMINT_MUL(const_12, one_over_2pq);
356 const SIMINT_DBLTYPE vrr_const_13_over_2pq = SIMINT_MUL(const_13, one_over_2pq);
357
358
359
360 // Forming PRIM_INT__p_s_s_s[15 * 3];
361 for(n = 0; n < 15; ++n) // loop over orders of auxiliary function
362 {
363
364 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
365 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]);
366
367 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
368 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]);
369
370 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
371 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]);
372
373 }
374
375
376
377 // Forming PRIM_INT__d_s_s_s[14 * 6];
378 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
379 {
380
381 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_s_s[n * 3 + 0]);
382 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]);
383 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]);
384
385 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 1]);
386 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]);
387 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]);
388
389 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 2]);
390 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]);
391 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]);
392
393 }
394
395
396
397 // Forming PRIM_INT__f_s_s_s[13 * 10];
398 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
399 {
400
401 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 0]);
402 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]);
403 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]);
404
405 PRIM_INT__f_s_s_s[n * 10 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 0]);
406 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]);
407
408 PRIM_INT__f_s_s_s[n * 10 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 0]);
409 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]);
410
411 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 3]);
412 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]);
413 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]);
414
415 PRIM_INT__f_s_s_s[n * 10 + 7] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 3]);
416 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]);
417
418 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 5]);
419 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]);
420 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]);
421
422 }
423
424
425 VRR_I_g_s_s_s(
426 PRIM_INT__g_s_s_s,
427 PRIM_INT__f_s_s_s,
428 PRIM_INT__d_s_s_s,
429 P_PA,
430 a_over_p,
431 aop_PQ,
432 one_over_2p,
433 12);
434
435
436 VRR_I_h_s_s_s(
437 PRIM_INT__h_s_s_s,
438 PRIM_INT__g_s_s_s,
439 PRIM_INT__f_s_s_s,
440 P_PA,
441 a_over_p,
442 aop_PQ,
443 one_over_2p,
444 11);
445
446
447 ostei_general_vrr1_I(6, 10,
448 one_over_2p, a_over_p, aop_PQ, P_PA,
449 PRIM_INT__h_s_s_s, PRIM_INT__g_s_s_s, PRIM_INT__i_s_s_s);
450
451
452 ostei_general_vrr1_I(7, 9,
453 one_over_2p, a_over_p, aop_PQ, P_PA,
454 PRIM_INT__i_s_s_s, PRIM_INT__h_s_s_s, PRIM_INT__k_s_s_s);
455
456
457 ostei_general_vrr_K(7, 0, 1, 0, 2,
458 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
459 PRIM_INT__k_s_s_s, NULL, NULL, PRIM_INT__i_s_s_s, NULL, PRIM_INT__k_s_p_s);
460
461
462 ostei_general_vrr_K(6, 0, 1, 0, 2,
463 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
464 PRIM_INT__i_s_s_s, NULL, NULL, PRIM_INT__h_s_s_s, NULL, PRIM_INT__i_s_p_s);
465
466
467 ostei_general_vrr_K(7, 0, 2, 0, 1,
468 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
469 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);
470
471
472 ostei_general_vrr1_I(8, 8,
473 one_over_2p, a_over_p, aop_PQ, P_PA,
474 PRIM_INT__k_s_s_s, PRIM_INT__i_s_s_s, PRIM_INT__l_s_s_s);
475
476
477 ostei_general_vrr_K(8, 0, 1, 0, 2,
478 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
479 PRIM_INT__l_s_s_s, NULL, NULL, PRIM_INT__k_s_s_s, NULL, PRIM_INT__l_s_p_s);
480
481
482 ostei_general_vrr_K(8, 0, 2, 0, 1,
483 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
484 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);
485
486
487 ostei_general_vrr1_I(9, 7,
488 one_over_2p, a_over_p, aop_PQ, P_PA,
489 PRIM_INT__l_s_s_s, PRIM_INT__k_s_s_s, PRIM_INT__m_s_s_s);
490
491
492 ostei_general_vrr_K(9, 0, 1, 0, 2,
493 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
494 PRIM_INT__m_s_s_s, NULL, NULL, PRIM_INT__l_s_s_s, NULL, PRIM_INT__m_s_p_s);
495
496
497 ostei_general_vrr_K(9, 0, 2, 0, 1,
498 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
499 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);
500
501
502 ostei_general_vrr1_I(10, 6,
503 one_over_2p, a_over_p, aop_PQ, P_PA,
504 PRIM_INT__m_s_s_s, PRIM_INT__l_s_s_s, PRIM_INT__n_s_s_s);
505
506
507 ostei_general_vrr_K(10, 0, 1, 0, 2,
508 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
509 PRIM_INT__n_s_s_s, NULL, NULL, PRIM_INT__m_s_s_s, NULL, PRIM_INT__n_s_p_s);
510
511
512 ostei_general_vrr_K(10, 0, 2, 0, 1,
513 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
514 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);
515
516
517 ostei_general_vrr1_I(11, 5,
518 one_over_2p, a_over_p, aop_PQ, P_PA,
519 PRIM_INT__n_s_s_s, PRIM_INT__m_s_s_s, PRIM_INT__o_s_s_s);
520
521
522 ostei_general_vrr_K(11, 0, 1, 0, 2,
523 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
524 PRIM_INT__o_s_s_s, NULL, NULL, PRIM_INT__n_s_s_s, NULL, PRIM_INT__o_s_p_s);
525
526
527 ostei_general_vrr_K(11, 0, 2, 0, 1,
528 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
529 PRIM_INT__o_s_p_s, PRIM_INT__o_s_s_s, NULL, PRIM_INT__n_s_p_s, NULL, PRIM_INT__o_s_d_s);
530
531
532 ostei_general_vrr1_I(12, 4,
533 one_over_2p, a_over_p, aop_PQ, P_PA,
534 PRIM_INT__o_s_s_s, PRIM_INT__n_s_s_s, PRIM_INT__q_s_s_s);
535
536
537 ostei_general_vrr_K(12, 0, 1, 0, 2,
538 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
539 PRIM_INT__q_s_s_s, NULL, NULL, PRIM_INT__o_s_s_s, NULL, PRIM_INT__q_s_p_s);
540
541
542 ostei_general_vrr_K(12, 0, 2, 0, 1,
543 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
544 PRIM_INT__q_s_p_s, PRIM_INT__q_s_s_s, NULL, PRIM_INT__o_s_p_s, NULL, PRIM_INT__q_s_d_s);
545
546
547 ostei_general_vrr1_I(13, 3,
548 one_over_2p, a_over_p, aop_PQ, P_PA,
549 PRIM_INT__q_s_s_s, PRIM_INT__o_s_s_s, PRIM_INT__r_s_s_s);
550
551
552 ostei_general_vrr_K(13, 0, 1, 0, 2,
553 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
554 PRIM_INT__r_s_s_s, NULL, NULL, PRIM_INT__q_s_s_s, NULL, PRIM_INT__r_s_p_s);
555
556
557 ostei_general_vrr_K(13, 0, 2, 0, 1,
558 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
559 PRIM_INT__r_s_p_s, PRIM_INT__r_s_s_s, NULL, PRIM_INT__q_s_p_s, NULL, PRIM_INT__r_s_d_s);
560
561
562
563
564 ////////////////////////////////////
565 // Accumulate contracted integrals
566 ////////////////////////////////////
567 if(lastoffset == 0)
568 {
569 contract_all(108, PRIM_INT__k_s_p_s, PRIM_PTR_INT__k_s_p_s);
570 contract_all(216, PRIM_INT__k_s_d_s, PRIM_PTR_INT__k_s_d_s);
571 contract_all(135, PRIM_INT__l_s_p_s, PRIM_PTR_INT__l_s_p_s);
572 contract_all(270, PRIM_INT__l_s_d_s, PRIM_PTR_INT__l_s_d_s);
573 contract_all(165, PRIM_INT__m_s_p_s, PRIM_PTR_INT__m_s_p_s);
574 contract_all(330, PRIM_INT__m_s_d_s, PRIM_PTR_INT__m_s_d_s);
575 contract_all(198, PRIM_INT__n_s_p_s, PRIM_PTR_INT__n_s_p_s);
576 contract_all(396, PRIM_INT__n_s_d_s, PRIM_PTR_INT__n_s_d_s);
577 contract_all(234, PRIM_INT__o_s_p_s, PRIM_PTR_INT__o_s_p_s);
578 contract_all(468, PRIM_INT__o_s_d_s, PRIM_PTR_INT__o_s_d_s);
579 contract_all(273, PRIM_INT__q_s_p_s, PRIM_PTR_INT__q_s_p_s);
580 contract_all(546, PRIM_INT__q_s_d_s, PRIM_PTR_INT__q_s_d_s);
581 contract_all(315, PRIM_INT__r_s_p_s, PRIM_PTR_INT__r_s_p_s);
582 contract_all(630, PRIM_INT__r_s_d_s, PRIM_PTR_INT__r_s_d_s);
583 }
584 else
585 {
586 contract(108, shelloffsets, PRIM_INT__k_s_p_s, PRIM_PTR_INT__k_s_p_s);
587 contract(216, shelloffsets, PRIM_INT__k_s_d_s, PRIM_PTR_INT__k_s_d_s);
588 contract(135, shelloffsets, PRIM_INT__l_s_p_s, PRIM_PTR_INT__l_s_p_s);
589 contract(270, shelloffsets, PRIM_INT__l_s_d_s, PRIM_PTR_INT__l_s_d_s);
590 contract(165, shelloffsets, PRIM_INT__m_s_p_s, PRIM_PTR_INT__m_s_p_s);
591 contract(330, shelloffsets, PRIM_INT__m_s_d_s, PRIM_PTR_INT__m_s_d_s);
592 contract(198, shelloffsets, PRIM_INT__n_s_p_s, PRIM_PTR_INT__n_s_p_s);
593 contract(396, shelloffsets, PRIM_INT__n_s_d_s, PRIM_PTR_INT__n_s_d_s);
594 contract(234, shelloffsets, PRIM_INT__o_s_p_s, PRIM_PTR_INT__o_s_p_s);
595 contract(468, shelloffsets, PRIM_INT__o_s_d_s, PRIM_PTR_INT__o_s_d_s);
596 contract(273, shelloffsets, PRIM_INT__q_s_p_s, PRIM_PTR_INT__q_s_p_s);
597 contract(546, shelloffsets, PRIM_INT__q_s_d_s, PRIM_PTR_INT__q_s_d_s);
598 contract(315, shelloffsets, PRIM_INT__r_s_p_s, PRIM_PTR_INT__r_s_p_s);
599 contract(630, shelloffsets, PRIM_INT__r_s_d_s, PRIM_PTR_INT__r_s_d_s);
600 PRIM_PTR_INT__k_s_p_s += lastoffset*108;
601 PRIM_PTR_INT__k_s_d_s += lastoffset*216;
602 PRIM_PTR_INT__l_s_p_s += lastoffset*135;
603 PRIM_PTR_INT__l_s_d_s += lastoffset*270;
604 PRIM_PTR_INT__m_s_p_s += lastoffset*165;
605 PRIM_PTR_INT__m_s_d_s += lastoffset*330;
606 PRIM_PTR_INT__n_s_p_s += lastoffset*198;
607 PRIM_PTR_INT__n_s_d_s += lastoffset*396;
608 PRIM_PTR_INT__o_s_p_s += lastoffset*234;
609 PRIM_PTR_INT__o_s_d_s += lastoffset*468;
610 PRIM_PTR_INT__q_s_p_s += lastoffset*273;
611 PRIM_PTR_INT__q_s_d_s += lastoffset*546;
612 PRIM_PTR_INT__r_s_p_s += lastoffset*315;
613 PRIM_PTR_INT__r_s_d_s += lastoffset*630;
614 }
615
616 } // close loop over j
617 } // close loop over i
618
619 //Advance to the next batch
620 jstart = SIMINT_SIMD_ROUND(jend);
621
622 //////////////////////////////////////////////
623 // Contracted integrals: Horizontal recurrance
624 //////////////////////////////////////////////
625
626
627 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
628
629
630 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
631 {
632 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
633
634 // set up HRR pointers
635 double const * restrict HRR_INT__k_s_p_s = INT__k_s_p_s + abcd * 108;
636 double const * restrict HRR_INT__k_s_d_s = INT__k_s_d_s + abcd * 216;
637 double const * restrict HRR_INT__l_s_p_s = INT__l_s_p_s + abcd * 135;
638 double const * restrict HRR_INT__l_s_d_s = INT__l_s_d_s + abcd * 270;
639 double const * restrict HRR_INT__m_s_p_s = INT__m_s_p_s + abcd * 165;
640 double const * restrict HRR_INT__m_s_d_s = INT__m_s_d_s + abcd * 330;
641 double const * restrict HRR_INT__n_s_p_s = INT__n_s_p_s + abcd * 198;
642 double const * restrict HRR_INT__n_s_d_s = INT__n_s_d_s + abcd * 396;
643 double const * restrict HRR_INT__o_s_p_s = INT__o_s_p_s + abcd * 234;
644 double const * restrict HRR_INT__o_s_d_s = INT__o_s_d_s + abcd * 468;
645 double const * restrict HRR_INT__q_s_p_s = INT__q_s_p_s + abcd * 273;
646 double const * restrict HRR_INT__q_s_d_s = INT__q_s_d_s + abcd * 546;
647 double const * restrict HRR_INT__r_s_p_s = INT__r_s_p_s + abcd * 315;
648 double const * restrict HRR_INT__r_s_d_s = INT__r_s_d_s + abcd * 630;
649 double * restrict HRR_INT__k_i_p_p = INT__k_i_p_p + real_abcd * 9072;
650
651 // form INT__k_p_p_s
652 ostei_general_hrr_J(7, 1, 1, 0, hAB, HRR_INT__l_s_p_s, HRR_INT__k_s_p_s, HRR_INT__k_p_p_s);
653
654 // form INT__k_p_d_s
655 ostei_general_hrr_J(7, 1, 2, 0, hAB, HRR_INT__l_s_d_s, HRR_INT__k_s_d_s, HRR_INT__k_p_d_s);
656
657 // form INT__l_p_p_s
658 ostei_general_hrr_J(8, 1, 1, 0, hAB, HRR_INT__m_s_p_s, HRR_INT__l_s_p_s, HRR_INT__l_p_p_s);
659
660 // form INT__l_p_d_s
661 ostei_general_hrr_J(8, 1, 2, 0, hAB, HRR_INT__m_s_d_s, HRR_INT__l_s_d_s, HRR_INT__l_p_d_s);
662
663 // form INT__m_p_p_s
664 ostei_general_hrr_J(9, 1, 1, 0, hAB, HRR_INT__n_s_p_s, HRR_INT__m_s_p_s, HRR_INT__m_p_p_s);
665
666 // form INT__m_p_d_s
667 ostei_general_hrr_J(9, 1, 2, 0, hAB, HRR_INT__n_s_d_s, HRR_INT__m_s_d_s, HRR_INT__m_p_d_s);
668
669 // form INT__n_p_p_s
670 ostei_general_hrr_J(10, 1, 1, 0, hAB, HRR_INT__o_s_p_s, HRR_INT__n_s_p_s, HRR_INT__n_p_p_s);
671
672 // form INT__n_p_d_s
673 ostei_general_hrr_J(10, 1, 2, 0, hAB, HRR_INT__o_s_d_s, HRR_INT__n_s_d_s, HRR_INT__n_p_d_s);
674
675 // form INT__o_p_p_s
676 ostei_general_hrr_J(11, 1, 1, 0, hAB, HRR_INT__q_s_p_s, HRR_INT__o_s_p_s, HRR_INT__o_p_p_s);
677
678 // form INT__o_p_d_s
679 ostei_general_hrr_J(11, 1, 2, 0, hAB, HRR_INT__q_s_d_s, HRR_INT__o_s_d_s, HRR_INT__o_p_d_s);
680
681 // form INT__q_p_p_s
682 ostei_general_hrr_J(12, 1, 1, 0, hAB, HRR_INT__r_s_p_s, HRR_INT__q_s_p_s, HRR_INT__q_p_p_s);
683
684 // form INT__q_p_d_s
685 ostei_general_hrr_J(12, 1, 2, 0, hAB, HRR_INT__r_s_d_s, HRR_INT__q_s_d_s, HRR_INT__q_p_d_s);
686
687 // form INT__k_d_p_s
688 ostei_general_hrr_J(7, 2, 1, 0, hAB, HRR_INT__l_p_p_s, HRR_INT__k_p_p_s, HRR_INT__k_d_p_s);
689
690 // form INT__k_d_d_s
691 ostei_general_hrr_J(7, 2, 2, 0, hAB, HRR_INT__l_p_d_s, HRR_INT__k_p_d_s, HRR_INT__k_d_d_s);
692
693 // form INT__l_d_p_s
694 ostei_general_hrr_J(8, 2, 1, 0, hAB, HRR_INT__m_p_p_s, HRR_INT__l_p_p_s, HRR_INT__l_d_p_s);
695
696 // form INT__l_d_d_s
697 ostei_general_hrr_J(8, 2, 2, 0, hAB, HRR_INT__m_p_d_s, HRR_INT__l_p_d_s, HRR_INT__l_d_d_s);
698
699 // form INT__m_d_p_s
700 ostei_general_hrr_J(9, 2, 1, 0, hAB, HRR_INT__n_p_p_s, HRR_INT__m_p_p_s, HRR_INT__m_d_p_s);
701
702 // form INT__m_d_d_s
703 ostei_general_hrr_J(9, 2, 2, 0, hAB, HRR_INT__n_p_d_s, HRR_INT__m_p_d_s, HRR_INT__m_d_d_s);
704
705 // form INT__n_d_p_s
706 ostei_general_hrr_J(10, 2, 1, 0, hAB, HRR_INT__o_p_p_s, HRR_INT__n_p_p_s, HRR_INT__n_d_p_s);
707
708 // form INT__n_d_d_s
709 ostei_general_hrr_J(10, 2, 2, 0, hAB, HRR_INT__o_p_d_s, HRR_INT__n_p_d_s, HRR_INT__n_d_d_s);
710
711 // form INT__o_d_p_s
712 ostei_general_hrr_J(11, 2, 1, 0, hAB, HRR_INT__q_p_p_s, HRR_INT__o_p_p_s, HRR_INT__o_d_p_s);
713
714 // form INT__o_d_d_s
715 ostei_general_hrr_J(11, 2, 2, 0, hAB, HRR_INT__q_p_d_s, HRR_INT__o_p_d_s, HRR_INT__o_d_d_s);
716
717 // form INT__k_f_p_s
718 ostei_general_hrr_J(7, 3, 1, 0, hAB, HRR_INT__l_d_p_s, HRR_INT__k_d_p_s, HRR_INT__k_f_p_s);
719
720 // form INT__k_f_d_s
721 ostei_general_hrr_J(7, 3, 2, 0, hAB, HRR_INT__l_d_d_s, HRR_INT__k_d_d_s, HRR_INT__k_f_d_s);
722
723 // form INT__l_f_p_s
724 ostei_general_hrr_J(8, 3, 1, 0, hAB, HRR_INT__m_d_p_s, HRR_INT__l_d_p_s, HRR_INT__l_f_p_s);
725
726 // form INT__l_f_d_s
727 ostei_general_hrr_J(8, 3, 2, 0, hAB, HRR_INT__m_d_d_s, HRR_INT__l_d_d_s, HRR_INT__l_f_d_s);
728
729 // form INT__m_f_p_s
730 ostei_general_hrr_J(9, 3, 1, 0, hAB, HRR_INT__n_d_p_s, HRR_INT__m_d_p_s, HRR_INT__m_f_p_s);
731
732 // form INT__m_f_d_s
733 ostei_general_hrr_J(9, 3, 2, 0, hAB, HRR_INT__n_d_d_s, HRR_INT__m_d_d_s, HRR_INT__m_f_d_s);
734
735 // form INT__n_f_p_s
736 ostei_general_hrr_J(10, 3, 1, 0, hAB, HRR_INT__o_d_p_s, HRR_INT__n_d_p_s, HRR_INT__n_f_p_s);
737
738 // form INT__n_f_d_s
739 ostei_general_hrr_J(10, 3, 2, 0, hAB, HRR_INT__o_d_d_s, HRR_INT__n_d_d_s, HRR_INT__n_f_d_s);
740
741 // form INT__k_g_p_s
742 ostei_general_hrr_J(7, 4, 1, 0, hAB, HRR_INT__l_f_p_s, HRR_INT__k_f_p_s, HRR_INT__k_g_p_s);
743
744 // form INT__k_g_d_s
745 ostei_general_hrr_J(7, 4, 2, 0, hAB, HRR_INT__l_f_d_s, HRR_INT__k_f_d_s, HRR_INT__k_g_d_s);
746
747 // form INT__l_g_p_s
748 ostei_general_hrr_J(8, 4, 1, 0, hAB, HRR_INT__m_f_p_s, HRR_INT__l_f_p_s, HRR_INT__l_g_p_s);
749
750 // form INT__l_g_d_s
751 ostei_general_hrr_J(8, 4, 2, 0, hAB, HRR_INT__m_f_d_s, HRR_INT__l_f_d_s, HRR_INT__l_g_d_s);
752
753 // form INT__m_g_p_s
754 ostei_general_hrr_J(9, 4, 1, 0, hAB, HRR_INT__n_f_p_s, HRR_INT__m_f_p_s, HRR_INT__m_g_p_s);
755
756 // form INT__m_g_d_s
757 ostei_general_hrr_J(9, 4, 2, 0, hAB, HRR_INT__n_f_d_s, HRR_INT__m_f_d_s, HRR_INT__m_g_d_s);
758
759 // form INT__k_h_p_s
760 ostei_general_hrr_J(7, 5, 1, 0, hAB, HRR_INT__l_g_p_s, HRR_INT__k_g_p_s, HRR_INT__k_h_p_s);
761
762 // form INT__k_h_d_s
763 ostei_general_hrr_J(7, 5, 2, 0, hAB, HRR_INT__l_g_d_s, HRR_INT__k_g_d_s, HRR_INT__k_h_d_s);
764
765 // form INT__l_h_p_s
766 ostei_general_hrr_J(8, 5, 1, 0, hAB, HRR_INT__m_g_p_s, HRR_INT__l_g_p_s, HRR_INT__l_h_p_s);
767
768 // form INT__l_h_d_s
769 ostei_general_hrr_J(8, 5, 2, 0, hAB, HRR_INT__m_g_d_s, HRR_INT__l_g_d_s, HRR_INT__l_h_d_s);
770
771 // form INT__k_i_p_s
772 ostei_general_hrr_J(7, 6, 1, 0, hAB, HRR_INT__l_h_p_s, HRR_INT__k_h_p_s, HRR_INT__k_i_p_s);
773
774 // form INT__k_i_d_s
775 ostei_general_hrr_J(7, 6, 2, 0, hAB, HRR_INT__l_h_d_s, HRR_INT__k_h_d_s, HRR_INT__k_i_d_s);
776
777 // form INT__k_i_p_p
778 for(ibra = 0; ibra < 1008; ++ibra)
779 {
780 HRR_INT__k_i_p_p[ibra * 9 + 0] = HRR_INT__k_i_d_s[ibra * 6 + 0] + ( hCD[0] * HRR_INT__k_i_p_s[ibra * 3 + 0] );
781
782 HRR_INT__k_i_p_p[ibra * 9 + 1] = HRR_INT__k_i_d_s[ibra * 6 + 1] + ( hCD[1] * HRR_INT__k_i_p_s[ibra * 3 + 0] );
783
784 HRR_INT__k_i_p_p[ibra * 9 + 2] = HRR_INT__k_i_d_s[ibra * 6 + 2] + ( hCD[2] * HRR_INT__k_i_p_s[ibra * 3 + 0] );
785
786 HRR_INT__k_i_p_p[ibra * 9 + 3] = HRR_INT__k_i_d_s[ibra * 6 + 1] + ( hCD[0] * HRR_INT__k_i_p_s[ibra * 3 + 1] );
787
788 HRR_INT__k_i_p_p[ibra * 9 + 4] = HRR_INT__k_i_d_s[ibra * 6 + 3] + ( hCD[1] * HRR_INT__k_i_p_s[ibra * 3 + 1] );
789
790 HRR_INT__k_i_p_p[ibra * 9 + 5] = HRR_INT__k_i_d_s[ibra * 6 + 4] + ( hCD[2] * HRR_INT__k_i_p_s[ibra * 3 + 1] );
791
792 HRR_INT__k_i_p_p[ibra * 9 + 6] = HRR_INT__k_i_d_s[ibra * 6 + 2] + ( hCD[0] * HRR_INT__k_i_p_s[ibra * 3 + 2] );
793
794 HRR_INT__k_i_p_p[ibra * 9 + 7] = HRR_INT__k_i_d_s[ibra * 6 + 4] + ( hCD[1] * HRR_INT__k_i_p_s[ibra * 3 + 2] );
795
796 HRR_INT__k_i_p_p[ibra * 9 + 8] = HRR_INT__k_i_d_s[ibra * 6 + 5] + ( hCD[2] * HRR_INT__k_i_p_s[ibra * 3 + 2] );
797
798 }
799
800
801 } // close HRR loop
802
803
804 } // close loop cdbatch
805
806 istart = iend;
807 } // close loop over ab
808
809 return P.nshell12_clip * Q.nshell12_clip;
810 }
811
ostei_i_k_p_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_k_p_p)812 int ostei_i_k_p_p(struct simint_multi_shellpair const P,
813 struct simint_multi_shellpair const Q,
814 double screen_tol,
815 double * const restrict work,
816 double * const restrict INT__i_k_p_p)
817 {
818 double P_AB[3*P.nshell12];
819 struct simint_multi_shellpair P_tmp = P;
820 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
821 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
822 P_tmp.AB_x = P_AB;
823 P_tmp.AB_y = P_AB + P.nshell12;
824 P_tmp.AB_z = P_AB + 2*P.nshell12;
825
826 for(int i = 0; i < P.nshell12; i++)
827 {
828 P_tmp.AB_x[i] = -P.AB_x[i];
829 P_tmp.AB_y[i] = -P.AB_y[i];
830 P_tmp.AB_z[i] = -P.AB_z[i];
831 }
832
833 int ret = ostei_k_i_p_p(P_tmp, Q, screen_tol, work, INT__i_k_p_p);
834 double buffer[9072] SIMINT_ALIGN_ARRAY_DBL;
835
836 for(int q = 0; q < ret; q++)
837 {
838 int idx = 0;
839 for(int a = 0; a < 28; ++a)
840 for(int b = 0; b < 36; ++b)
841 for(int c = 0; c < 3; ++c)
842 for(int d = 0; d < 3; ++d)
843 buffer[idx++] = INT__i_k_p_p[q*9072+b*252+a*9+c*3+d];
844
845 memcpy(INT__i_k_p_p+q*9072, buffer, 9072*sizeof(double));
846 }
847
848 return ret;
849 }
850
851