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_g_d_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_g_d_p)8 int ostei_k_g_d_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_g_d_p)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__k_g_d_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_d_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__k_s_f_s = work + (SIMINT_NSHELL_SIMD * 216);
31 double * const INT__l_s_d_s = work + (SIMINT_NSHELL_SIMD * 576);
32 double * const INT__l_s_f_s = work + (SIMINT_NSHELL_SIMD * 846);
33 double * const INT__m_s_d_s = work + (SIMINT_NSHELL_SIMD * 1296);
34 double * const INT__m_s_f_s = work + (SIMINT_NSHELL_SIMD * 1626);
35 double * const INT__n_s_d_s = work + (SIMINT_NSHELL_SIMD * 2176);
36 double * const INT__n_s_f_s = work + (SIMINT_NSHELL_SIMD * 2572);
37 double * const INT__o_s_d_s = work + (SIMINT_NSHELL_SIMD * 3232);
38 double * const INT__o_s_f_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__p_s_s_s = primwork + 15;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_s_s = primwork + 57;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_s_s = primwork + 135;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_s_s = primwork + 255;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_s_s = primwork + 420;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_p_s = primwork + 630;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_s_s = primwork + 819;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_p_s = primwork + 1071;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_d_s = primwork + 1323;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_s_s = primwork + 1659;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_p_s = primwork + 1947;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_d_s = primwork + 2271;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__k_s_f_s = primwork + 2703;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_s_s = primwork + 3063;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_p_s = primwork + 3378;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_d_s = primwork + 3783;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__l_s_f_s = primwork + 4323;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_s_s = primwork + 4773;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_p_s = primwork + 5103;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_d_s = primwork + 5598;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__m_s_f_s = primwork + 6258;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_s_s = primwork + 6808;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_p_s = primwork + 7138;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_d_s = primwork + 7732;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__n_s_f_s = primwork + 8524;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_s_s = primwork + 9184;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_p_s = primwork + 9496;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_d_s = primwork + 10198;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__o_s_f_s = primwork + 11134;
70 double * const hrrwork = (double *)(primwork + 11914);
71 double * const HRR_INT__k_p_d_s = hrrwork + 0;
72 double * const HRR_INT__k_p_f_s = hrrwork + 648;
73 double * const HRR_INT__k_d_d_s = hrrwork + 1728;
74 double * const HRR_INT__k_d_f_s = hrrwork + 3024;
75 double * const HRR_INT__k_f_d_s = hrrwork + 5184;
76 double * const HRR_INT__k_f_f_s = hrrwork + 7344;
77 double * const HRR_INT__k_g_d_s = hrrwork + 10944;
78 double * const HRR_INT__k_g_f_s = hrrwork + 14184;
79 double * const HRR_INT__l_p_d_s = hrrwork + 19584;
80 double * const HRR_INT__l_p_f_s = hrrwork + 20394;
81 double * const HRR_INT__l_d_d_s = hrrwork + 21744;
82 double * const HRR_INT__l_d_f_s = hrrwork + 23364;
83 double * const HRR_INT__l_f_d_s = hrrwork + 26064;
84 double * const HRR_INT__l_f_f_s = hrrwork + 28764;
85 double * const HRR_INT__m_p_d_s = hrrwork + 33264;
86 double * const HRR_INT__m_p_f_s = hrrwork + 34254;
87 double * const HRR_INT__m_d_d_s = hrrwork + 35904;
88 double * const HRR_INT__m_d_f_s = hrrwork + 37884;
89 double * const HRR_INT__n_p_d_s = hrrwork + 41184;
90 double * const HRR_INT__n_p_f_s = hrrwork + 42372;
91
92
93 // Create constants
94 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
95 const SIMINT_DBLTYPE const_10 = SIMINT_DBLSET1(10);
96 const SIMINT_DBLTYPE const_11 = SIMINT_DBLSET1(11);
97 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
98 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
99 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
100 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
101 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
102 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
103 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
104 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
105 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
106
107
108 ////////////////////////////////////////
109 // Loop over shells and primitives
110 ////////////////////////////////////////
111
112 real_abcd = 0;
113 istart = 0;
114 for(ab = 0; ab < P.nshell12_clip; ++ab)
115 {
116 const int iend = istart + P.nprim12[ab];
117
118 cd = 0;
119 jstart = 0;
120
121 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
122 {
123 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
124 int jend = jstart;
125 for(i = 0; i < nshellbatch; i++)
126 jend += Q.nprim12[cd+i];
127
128 // Clear the beginning of the workspace (where we are accumulating integrals)
129 memset(work, 0, SIMINT_NSHELL_SIMD * 4480 * sizeof(double));
130 abcd = 0;
131
132
133 for(i = istart; i < iend; ++i)
134 {
135 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
136
137 if(check_screen)
138 {
139 // Skip this whole thing if always insignificant
140 if((P.screen[i] * Q.screen_max) < screen_tol)
141 continue;
142 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
143 }
144
145 icd = 0;
146 iprimcd = 0;
147 nprim_icd = Q.nprim12[cd];
148 double * restrict PRIM_PTR_INT__k_s_d_s = INT__k_s_d_s + abcd * 216;
149 double * restrict PRIM_PTR_INT__k_s_f_s = INT__k_s_f_s + abcd * 360;
150 double * restrict PRIM_PTR_INT__l_s_d_s = INT__l_s_d_s + abcd * 270;
151 double * restrict PRIM_PTR_INT__l_s_f_s = INT__l_s_f_s + abcd * 450;
152 double * restrict PRIM_PTR_INT__m_s_d_s = INT__m_s_d_s + abcd * 330;
153 double * restrict PRIM_PTR_INT__m_s_f_s = INT__m_s_f_s + abcd * 550;
154 double * restrict PRIM_PTR_INT__n_s_d_s = INT__n_s_d_s + abcd * 396;
155 double * restrict PRIM_PTR_INT__n_s_f_s = INT__n_s_f_s + abcd * 660;
156 double * restrict PRIM_PTR_INT__o_s_d_s = INT__o_s_d_s + abcd * 468;
157 double * restrict PRIM_PTR_INT__o_s_f_s = INT__o_s_f_s + abcd * 780;
158
159
160
161 // Load these one per loop over i
162 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
163 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
164 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
165
166 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
167
168 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
169 {
170 // calculate the shell offsets
171 // these are the offset from the shell pointed to by cd
172 // for each element
173 int shelloffsets[SIMINT_SIMD_LEN] = {0};
174 int lastoffset = 0;
175 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
176
177 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
178 {
179 // Handle if the first element of the vector is a new shell
180 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
181 {
182 nprim_icd += Q.nprim12[cd + (++icd)];
183 PRIM_PTR_INT__k_s_d_s += 216;
184 PRIM_PTR_INT__k_s_f_s += 360;
185 PRIM_PTR_INT__l_s_d_s += 270;
186 PRIM_PTR_INT__l_s_f_s += 450;
187 PRIM_PTR_INT__m_s_d_s += 330;
188 PRIM_PTR_INT__m_s_f_s += 550;
189 PRIM_PTR_INT__n_s_d_s += 396;
190 PRIM_PTR_INT__n_s_f_s += 660;
191 PRIM_PTR_INT__o_s_d_s += 468;
192 PRIM_PTR_INT__o_s_f_s += 780;
193 }
194 iprimcd++;
195 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
196 {
197 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
198 {
199 shelloffsets[n] = shelloffsets[n-1] + 1;
200 lastoffset++;
201 nprim_icd += Q.nprim12[cd + (++icd)];
202 }
203 else
204 shelloffsets[n] = shelloffsets[n-1];
205 iprimcd++;
206 }
207 }
208 else
209 iprimcd += SIMINT_SIMD_LEN;
210
211 // Do we have to compute this vector (or has it been screened out)?
212 // (not_screened != 0 means we have to do this vector)
213 if(check_screen)
214 {
215 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
216 if(vmax < screen_tol)
217 {
218 PRIM_PTR_INT__k_s_d_s += lastoffset*216;
219 PRIM_PTR_INT__k_s_f_s += lastoffset*360;
220 PRIM_PTR_INT__l_s_d_s += lastoffset*270;
221 PRIM_PTR_INT__l_s_f_s += lastoffset*450;
222 PRIM_PTR_INT__m_s_d_s += lastoffset*330;
223 PRIM_PTR_INT__m_s_f_s += lastoffset*550;
224 PRIM_PTR_INT__n_s_d_s += lastoffset*396;
225 PRIM_PTR_INT__n_s_f_s += lastoffset*660;
226 PRIM_PTR_INT__o_s_d_s += lastoffset*468;
227 PRIM_PTR_INT__o_s_f_s += lastoffset*780;
228 continue;
229 }
230 }
231
232 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
233 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
234 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
235 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
236
237
238 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
239 SIMINT_DBLTYPE PQ[3];
240 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
241 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
242 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
243 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
244 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
245 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
246
247 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
248 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
249 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
250 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
251 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
252 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
253 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
254
255 // NOTE: Minus sign!
256 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
257 SIMINT_DBLTYPE aop_PQ[3];
258 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
259 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
260 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
261
262 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
263 SIMINT_DBLTYPE aoq_PQ[3];
264 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
265 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
266 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
267 // Put a minus sign here so we don't have to in RR routines
268 a_over_q = SIMINT_NEG(a_over_q);
269
270
271 //////////////////////////////////////////////
272 // Fjt function section
273 // Maximum v value: 14
274 //////////////////////////////////////////////
275 // The parameter to the Fjt function
276 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
277
278
279 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
280
281
282 boys_F_split(PRIM_INT__s_s_s_s, F_x, 14);
283 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
284 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
285 for(n = 0; n <= 14; n++)
286 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
287
288 //////////////////////////////////////////////
289 // Primitive integrals: Vertical recurrance
290 //////////////////////////////////////////////
291
292 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
293 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
294 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
295 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
296 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
297 const SIMINT_DBLTYPE vrr_const_6_over_2p = SIMINT_MUL(const_6, one_over_2p);
298 const SIMINT_DBLTYPE vrr_const_7_over_2p = SIMINT_MUL(const_7, one_over_2p);
299 const SIMINT_DBLTYPE vrr_const_8_over_2p = SIMINT_MUL(const_8, one_over_2p);
300 const SIMINT_DBLTYPE vrr_const_9_over_2p = SIMINT_MUL(const_9, one_over_2p);
301 const SIMINT_DBLTYPE vrr_const_10_over_2p = SIMINT_MUL(const_10, 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_1_over_2pq = one_over_2pq;
305 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
306 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
307 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
308 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
309 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
310 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
311 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
312 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
313 const SIMINT_DBLTYPE vrr_const_10_over_2pq = SIMINT_MUL(const_10, one_over_2pq);
314 const SIMINT_DBLTYPE vrr_const_11_over_2pq = SIMINT_MUL(const_11, one_over_2pq);
315
316
317
318 // Forming PRIM_INT__p_s_s_s[14 * 3];
319 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
320 {
321
322 PRIM_INT__p_s_s_s[n * 3 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
323 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]);
324
325 PRIM_INT__p_s_s_s[n * 3 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
326 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]);
327
328 PRIM_INT__p_s_s_s[n * 3 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
329 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]);
330
331 }
332
333
334
335 // Forming PRIM_INT__d_s_s_s[13 * 6];
336 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
337 {
338
339 PRIM_INT__d_s_s_s[n * 6 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__p_s_s_s[n * 3 + 0]);
340 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]);
341 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]);
342
343 PRIM_INT__d_s_s_s[n * 6 + 3] = SIMINT_MUL(P_PA[1], PRIM_INT__p_s_s_s[n * 3 + 1]);
344 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]);
345 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]);
346
347 PRIM_INT__d_s_s_s[n * 6 + 5] = SIMINT_MUL(P_PA[2], PRIM_INT__p_s_s_s[n * 3 + 2]);
348 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]);
349 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]);
350
351 }
352
353
354
355 // Forming PRIM_INT__f_s_s_s[12 * 10];
356 for(n = 0; n < 12; ++n) // loop over orders of auxiliary function
357 {
358
359 PRIM_INT__f_s_s_s[n * 10 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 0]);
360 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]);
361 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]);
362
363 PRIM_INT__f_s_s_s[n * 10 + 1] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 0]);
364 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]);
365
366 PRIM_INT__f_s_s_s[n * 10 + 2] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 0]);
367 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]);
368
369 PRIM_INT__f_s_s_s[n * 10 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 3]);
370 PRIM_INT__f_s_s_s[n * 10 + 3] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 3], PRIM_INT__f_s_s_s[n * 10 + 3]);
371
372 PRIM_INT__f_s_s_s[n * 10 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__d_s_s_s[n * 6 + 5]);
373 PRIM_INT__f_s_s_s[n * 10 + 5] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__d_s_s_s[(n+1) * 6 + 5], PRIM_INT__f_s_s_s[n * 10 + 5]);
374
375 PRIM_INT__f_s_s_s[n * 10 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__d_s_s_s[n * 6 + 3]);
376 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]);
377 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]);
378
379 PRIM_INT__f_s_s_s[n * 10 + 7] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 3]);
380 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]);
381
382 PRIM_INT__f_s_s_s[n * 10 + 9] = SIMINT_MUL(P_PA[2], PRIM_INT__d_s_s_s[n * 6 + 5]);
383 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]);
384 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]);
385
386 }
387
388
389 VRR_I_g_s_s_s(
390 PRIM_INT__g_s_s_s,
391 PRIM_INT__f_s_s_s,
392 PRIM_INT__d_s_s_s,
393 P_PA,
394 a_over_p,
395 aop_PQ,
396 one_over_2p,
397 11);
398
399
400 VRR_I_h_s_s_s(
401 PRIM_INT__h_s_s_s,
402 PRIM_INT__g_s_s_s,
403 PRIM_INT__f_s_s_s,
404 P_PA,
405 a_over_p,
406 aop_PQ,
407 one_over_2p,
408 10);
409
410
411 ostei_general_vrr1_I(6, 9,
412 one_over_2p, a_over_p, aop_PQ, P_PA,
413 PRIM_INT__h_s_s_s, PRIM_INT__g_s_s_s, PRIM_INT__i_s_s_s);
414
415
416 ostei_general_vrr1_I(7, 8,
417 one_over_2p, a_over_p, aop_PQ, P_PA,
418 PRIM_INT__i_s_s_s, PRIM_INT__h_s_s_s, PRIM_INT__k_s_s_s);
419
420
421 ostei_general_vrr_K(7, 0, 1, 0, 3,
422 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
423 PRIM_INT__k_s_s_s, NULL, NULL, PRIM_INT__i_s_s_s, NULL, PRIM_INT__k_s_p_s);
424
425
426 ostei_general_vrr_K(6, 0, 1, 0, 3,
427 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
428 PRIM_INT__i_s_s_s, NULL, NULL, PRIM_INT__h_s_s_s, NULL, PRIM_INT__i_s_p_s);
429
430
431 ostei_general_vrr_K(7, 0, 2, 0, 2,
432 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
433 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);
434
435
436 ostei_general_vrr_K(5, 0, 1, 0, 3,
437 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
438 PRIM_INT__h_s_s_s, NULL, NULL, PRIM_INT__g_s_s_s, NULL, PRIM_INT__h_s_p_s);
439
440
441 ostei_general_vrr_K(6, 0, 2, 0, 2,
442 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
443 PRIM_INT__i_s_p_s, PRIM_INT__i_s_s_s, NULL, PRIM_INT__h_s_p_s, NULL, PRIM_INT__i_s_d_s);
444
445
446 ostei_general_vrr_K(7, 0, 3, 0, 1,
447 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
448 PRIM_INT__k_s_d_s, PRIM_INT__k_s_p_s, NULL, PRIM_INT__i_s_d_s, NULL, PRIM_INT__k_s_f_s);
449
450
451 ostei_general_vrr1_I(8, 7,
452 one_over_2p, a_over_p, aop_PQ, P_PA,
453 PRIM_INT__k_s_s_s, PRIM_INT__i_s_s_s, PRIM_INT__l_s_s_s);
454
455
456 ostei_general_vrr_K(8, 0, 1, 0, 3,
457 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
458 PRIM_INT__l_s_s_s, NULL, NULL, PRIM_INT__k_s_s_s, NULL, PRIM_INT__l_s_p_s);
459
460
461 ostei_general_vrr_K(8, 0, 2, 0, 2,
462 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
463 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);
464
465
466 ostei_general_vrr_K(8, 0, 3, 0, 1,
467 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
468 PRIM_INT__l_s_d_s, PRIM_INT__l_s_p_s, NULL, PRIM_INT__k_s_d_s, NULL, PRIM_INT__l_s_f_s);
469
470
471 ostei_general_vrr1_I(9, 6,
472 one_over_2p, a_over_p, aop_PQ, P_PA,
473 PRIM_INT__l_s_s_s, PRIM_INT__k_s_s_s, PRIM_INT__m_s_s_s);
474
475
476 ostei_general_vrr_K(9, 0, 1, 0, 3,
477 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
478 PRIM_INT__m_s_s_s, NULL, NULL, PRIM_INT__l_s_s_s, NULL, PRIM_INT__m_s_p_s);
479
480
481 ostei_general_vrr_K(9, 0, 2, 0, 2,
482 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
483 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);
484
485
486 ostei_general_vrr_K(9, 0, 3, 0, 1,
487 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
488 PRIM_INT__m_s_d_s, PRIM_INT__m_s_p_s, NULL, PRIM_INT__l_s_d_s, NULL, PRIM_INT__m_s_f_s);
489
490
491 ostei_general_vrr1_I(10, 5,
492 one_over_2p, a_over_p, aop_PQ, P_PA,
493 PRIM_INT__m_s_s_s, PRIM_INT__l_s_s_s, PRIM_INT__n_s_s_s);
494
495
496 ostei_general_vrr_K(10, 0, 1, 0, 3,
497 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
498 PRIM_INT__n_s_s_s, NULL, NULL, PRIM_INT__m_s_s_s, NULL, PRIM_INT__n_s_p_s);
499
500
501 ostei_general_vrr_K(10, 0, 2, 0, 2,
502 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
503 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);
504
505
506 ostei_general_vrr_K(10, 0, 3, 0, 1,
507 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
508 PRIM_INT__n_s_d_s, PRIM_INT__n_s_p_s, NULL, PRIM_INT__m_s_d_s, NULL, PRIM_INT__n_s_f_s);
509
510
511 ostei_general_vrr1_I(11, 4,
512 one_over_2p, a_over_p, aop_PQ, P_PA,
513 PRIM_INT__n_s_s_s, PRIM_INT__m_s_s_s, PRIM_INT__o_s_s_s);
514
515
516 ostei_general_vrr_K(11, 0, 1, 0, 3,
517 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
518 PRIM_INT__o_s_s_s, NULL, NULL, PRIM_INT__n_s_s_s, NULL, PRIM_INT__o_s_p_s);
519
520
521 ostei_general_vrr_K(11, 0, 2, 0, 2,
522 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
523 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);
524
525
526 ostei_general_vrr_K(11, 0, 3, 0, 1,
527 one_over_2q, a_over_q, one_over_2pq, aoq_PQ, Q_PA,
528 PRIM_INT__o_s_d_s, PRIM_INT__o_s_p_s, NULL, PRIM_INT__n_s_d_s, NULL, PRIM_INT__o_s_f_s);
529
530
531
532
533 ////////////////////////////////////
534 // Accumulate contracted integrals
535 ////////////////////////////////////
536 if(lastoffset == 0)
537 {
538 contract_all(216, PRIM_INT__k_s_d_s, PRIM_PTR_INT__k_s_d_s);
539 contract_all(360, PRIM_INT__k_s_f_s, PRIM_PTR_INT__k_s_f_s);
540 contract_all(270, PRIM_INT__l_s_d_s, PRIM_PTR_INT__l_s_d_s);
541 contract_all(450, PRIM_INT__l_s_f_s, PRIM_PTR_INT__l_s_f_s);
542 contract_all(330, PRIM_INT__m_s_d_s, PRIM_PTR_INT__m_s_d_s);
543 contract_all(550, PRIM_INT__m_s_f_s, PRIM_PTR_INT__m_s_f_s);
544 contract_all(396, PRIM_INT__n_s_d_s, PRIM_PTR_INT__n_s_d_s);
545 contract_all(660, PRIM_INT__n_s_f_s, PRIM_PTR_INT__n_s_f_s);
546 contract_all(468, PRIM_INT__o_s_d_s, PRIM_PTR_INT__o_s_d_s);
547 contract_all(780, PRIM_INT__o_s_f_s, PRIM_PTR_INT__o_s_f_s);
548 }
549 else
550 {
551 contract(216, shelloffsets, PRIM_INT__k_s_d_s, PRIM_PTR_INT__k_s_d_s);
552 contract(360, shelloffsets, PRIM_INT__k_s_f_s, PRIM_PTR_INT__k_s_f_s);
553 contract(270, shelloffsets, PRIM_INT__l_s_d_s, PRIM_PTR_INT__l_s_d_s);
554 contract(450, shelloffsets, PRIM_INT__l_s_f_s, PRIM_PTR_INT__l_s_f_s);
555 contract(330, shelloffsets, PRIM_INT__m_s_d_s, PRIM_PTR_INT__m_s_d_s);
556 contract(550, shelloffsets, PRIM_INT__m_s_f_s, PRIM_PTR_INT__m_s_f_s);
557 contract(396, shelloffsets, PRIM_INT__n_s_d_s, PRIM_PTR_INT__n_s_d_s);
558 contract(660, shelloffsets, PRIM_INT__n_s_f_s, PRIM_PTR_INT__n_s_f_s);
559 contract(468, shelloffsets, PRIM_INT__o_s_d_s, PRIM_PTR_INT__o_s_d_s);
560 contract(780, shelloffsets, PRIM_INT__o_s_f_s, PRIM_PTR_INT__o_s_f_s);
561 PRIM_PTR_INT__k_s_d_s += lastoffset*216;
562 PRIM_PTR_INT__k_s_f_s += lastoffset*360;
563 PRIM_PTR_INT__l_s_d_s += lastoffset*270;
564 PRIM_PTR_INT__l_s_f_s += lastoffset*450;
565 PRIM_PTR_INT__m_s_d_s += lastoffset*330;
566 PRIM_PTR_INT__m_s_f_s += lastoffset*550;
567 PRIM_PTR_INT__n_s_d_s += lastoffset*396;
568 PRIM_PTR_INT__n_s_f_s += lastoffset*660;
569 PRIM_PTR_INT__o_s_d_s += lastoffset*468;
570 PRIM_PTR_INT__o_s_f_s += lastoffset*780;
571 }
572
573 } // close loop over j
574 } // close loop over i
575
576 //Advance to the next batch
577 jstart = SIMINT_SIMD_ROUND(jend);
578
579 //////////////////////////////////////////////
580 // Contracted integrals: Horizontal recurrance
581 //////////////////////////////////////////////
582
583
584 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
585
586
587 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
588 {
589 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
590
591 // set up HRR pointers
592 double const * restrict HRR_INT__k_s_d_s = INT__k_s_d_s + abcd * 216;
593 double const * restrict HRR_INT__k_s_f_s = INT__k_s_f_s + abcd * 360;
594 double const * restrict HRR_INT__l_s_d_s = INT__l_s_d_s + abcd * 270;
595 double const * restrict HRR_INT__l_s_f_s = INT__l_s_f_s + abcd * 450;
596 double const * restrict HRR_INT__m_s_d_s = INT__m_s_d_s + abcd * 330;
597 double const * restrict HRR_INT__m_s_f_s = INT__m_s_f_s + abcd * 550;
598 double const * restrict HRR_INT__n_s_d_s = INT__n_s_d_s + abcd * 396;
599 double const * restrict HRR_INT__n_s_f_s = INT__n_s_f_s + abcd * 660;
600 double const * restrict HRR_INT__o_s_d_s = INT__o_s_d_s + abcd * 468;
601 double const * restrict HRR_INT__o_s_f_s = INT__o_s_f_s + abcd * 780;
602 double * restrict HRR_INT__k_g_d_p = INT__k_g_d_p + real_abcd * 9720;
603
604 // form INT__k_p_d_s
605 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);
606
607 // form INT__k_p_f_s
608 ostei_general_hrr_J(7, 1, 3, 0, hAB, HRR_INT__l_s_f_s, HRR_INT__k_s_f_s, HRR_INT__k_p_f_s);
609
610 // form INT__l_p_d_s
611 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);
612
613 // form INT__l_p_f_s
614 ostei_general_hrr_J(8, 1, 3, 0, hAB, HRR_INT__m_s_f_s, HRR_INT__l_s_f_s, HRR_INT__l_p_f_s);
615
616 // form INT__m_p_d_s
617 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);
618
619 // form INT__m_p_f_s
620 ostei_general_hrr_J(9, 1, 3, 0, hAB, HRR_INT__n_s_f_s, HRR_INT__m_s_f_s, HRR_INT__m_p_f_s);
621
622 // form INT__n_p_d_s
623 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);
624
625 // form INT__n_p_f_s
626 ostei_general_hrr_J(10, 1, 3, 0, hAB, HRR_INT__o_s_f_s, HRR_INT__n_s_f_s, HRR_INT__n_p_f_s);
627
628 // form INT__k_d_d_s
629 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);
630
631 // form INT__k_d_f_s
632 ostei_general_hrr_J(7, 2, 3, 0, hAB, HRR_INT__l_p_f_s, HRR_INT__k_p_f_s, HRR_INT__k_d_f_s);
633
634 // form INT__l_d_d_s
635 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);
636
637 // form INT__l_d_f_s
638 ostei_general_hrr_J(8, 2, 3, 0, hAB, HRR_INT__m_p_f_s, HRR_INT__l_p_f_s, HRR_INT__l_d_f_s);
639
640 // form INT__m_d_d_s
641 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);
642
643 // form INT__m_d_f_s
644 ostei_general_hrr_J(9, 2, 3, 0, hAB, HRR_INT__n_p_f_s, HRR_INT__m_p_f_s, HRR_INT__m_d_f_s);
645
646 // form INT__k_f_d_s
647 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);
648
649 // form INT__k_f_f_s
650 ostei_general_hrr_J(7, 3, 3, 0, hAB, HRR_INT__l_d_f_s, HRR_INT__k_d_f_s, HRR_INT__k_f_f_s);
651
652 // form INT__l_f_d_s
653 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);
654
655 // form INT__l_f_f_s
656 ostei_general_hrr_J(8, 3, 3, 0, hAB, HRR_INT__m_d_f_s, HRR_INT__l_d_f_s, HRR_INT__l_f_f_s);
657
658 // form INT__k_g_d_s
659 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);
660
661 // form INT__k_g_f_s
662 ostei_general_hrr_J(7, 4, 3, 0, hAB, HRR_INT__l_f_f_s, HRR_INT__k_f_f_s, HRR_INT__k_g_f_s);
663
664 // form INT__k_g_d_p
665 for(ibra = 0; ibra < 540; ++ibra)
666 {
667 HRR_INT__k_g_d_p[ibra * 18 + 0] = HRR_INT__k_g_f_s[ibra * 10 + 0] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 0] );
668
669 HRR_INT__k_g_d_p[ibra * 18 + 1] = HRR_INT__k_g_f_s[ibra * 10 + 1] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 0] );
670
671 HRR_INT__k_g_d_p[ibra * 18 + 2] = HRR_INT__k_g_f_s[ibra * 10 + 2] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 0] );
672
673 HRR_INT__k_g_d_p[ibra * 18 + 3] = HRR_INT__k_g_f_s[ibra * 10 + 1] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 1] );
674
675 HRR_INT__k_g_d_p[ibra * 18 + 4] = HRR_INT__k_g_f_s[ibra * 10 + 3] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 1] );
676
677 HRR_INT__k_g_d_p[ibra * 18 + 5] = HRR_INT__k_g_f_s[ibra * 10 + 4] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 1] );
678
679 HRR_INT__k_g_d_p[ibra * 18 + 6] = HRR_INT__k_g_f_s[ibra * 10 + 2] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 2] );
680
681 HRR_INT__k_g_d_p[ibra * 18 + 7] = HRR_INT__k_g_f_s[ibra * 10 + 4] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 2] );
682
683 HRR_INT__k_g_d_p[ibra * 18 + 8] = HRR_INT__k_g_f_s[ibra * 10 + 5] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 2] );
684
685 HRR_INT__k_g_d_p[ibra * 18 + 9] = HRR_INT__k_g_f_s[ibra * 10 + 3] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 3] );
686
687 HRR_INT__k_g_d_p[ibra * 18 + 10] = HRR_INT__k_g_f_s[ibra * 10 + 6] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 3] );
688
689 HRR_INT__k_g_d_p[ibra * 18 + 11] = HRR_INT__k_g_f_s[ibra * 10 + 7] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 3] );
690
691 HRR_INT__k_g_d_p[ibra * 18 + 12] = HRR_INT__k_g_f_s[ibra * 10 + 4] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 4] );
692
693 HRR_INT__k_g_d_p[ibra * 18 + 13] = HRR_INT__k_g_f_s[ibra * 10 + 7] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 4] );
694
695 HRR_INT__k_g_d_p[ibra * 18 + 14] = HRR_INT__k_g_f_s[ibra * 10 + 8] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 4] );
696
697 HRR_INT__k_g_d_p[ibra * 18 + 15] = HRR_INT__k_g_f_s[ibra * 10 + 5] + ( hCD[0] * HRR_INT__k_g_d_s[ibra * 6 + 5] );
698
699 HRR_INT__k_g_d_p[ibra * 18 + 16] = HRR_INT__k_g_f_s[ibra * 10 + 8] + ( hCD[1] * HRR_INT__k_g_d_s[ibra * 6 + 5] );
700
701 HRR_INT__k_g_d_p[ibra * 18 + 17] = HRR_INT__k_g_f_s[ibra * 10 + 9] + ( hCD[2] * HRR_INT__k_g_d_s[ibra * 6 + 5] );
702
703 }
704
705
706 } // close HRR loop
707
708
709 } // close loop cdbatch
710
711 istart = iend;
712 } // close loop over ab
713
714 return P.nshell12_clip * Q.nshell12_clip;
715 }
716
ostei_g_k_d_p(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__g_k_d_p)717 int ostei_g_k_d_p(struct simint_multi_shellpair const P,
718 struct simint_multi_shellpair const Q,
719 double screen_tol,
720 double * const restrict work,
721 double * const restrict INT__g_k_d_p)
722 {
723 double P_AB[3*P.nshell12];
724 struct simint_multi_shellpair P_tmp = P;
725 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
726 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
727 P_tmp.AB_x = P_AB;
728 P_tmp.AB_y = P_AB + P.nshell12;
729 P_tmp.AB_z = P_AB + 2*P.nshell12;
730
731 for(int i = 0; i < P.nshell12; i++)
732 {
733 P_tmp.AB_x[i] = -P.AB_x[i];
734 P_tmp.AB_y[i] = -P.AB_y[i];
735 P_tmp.AB_z[i] = -P.AB_z[i];
736 }
737
738 int ret = ostei_k_g_d_p(P_tmp, Q, screen_tol, work, INT__g_k_d_p);
739 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
740
741 for(int q = 0; q < ret; q++)
742 {
743 int idx = 0;
744 for(int a = 0; a < 15; ++a)
745 for(int b = 0; b < 36; ++b)
746 for(int c = 0; c < 6; ++c)
747 for(int d = 0; d < 3; ++d)
748 buffer[idx++] = INT__g_k_d_p[q*9720+b*270+a*18+c*3+d];
749
750 memcpy(INT__g_k_d_p+q*9720, buffer, 9720*sizeof(double));
751 }
752
753 return ret;
754 }
755
ostei_k_g_p_d(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__k_g_p_d)756 int ostei_k_g_p_d(struct simint_multi_shellpair const P,
757 struct simint_multi_shellpair const Q,
758 double screen_tol,
759 double * const restrict work,
760 double * const restrict INT__k_g_p_d)
761 {
762 double Q_AB[3*Q.nshell12];
763 struct simint_multi_shellpair Q_tmp = Q;
764 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
765 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
766 Q_tmp.AB_x = Q_AB;
767 Q_tmp.AB_y = Q_AB + Q.nshell12;
768 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
769
770 for(int i = 0; i < Q.nshell12; i++)
771 {
772 Q_tmp.AB_x[i] = -Q.AB_x[i];
773 Q_tmp.AB_y[i] = -Q.AB_y[i];
774 Q_tmp.AB_z[i] = -Q.AB_z[i];
775 }
776
777 int ret = ostei_k_g_d_p(P, Q_tmp, screen_tol, work, INT__k_g_p_d);
778 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
779
780 for(int q = 0; q < ret; q++)
781 {
782 int idx = 0;
783 for(int a = 0; a < 36; ++a)
784 for(int b = 0; b < 15; ++b)
785 for(int c = 0; c < 3; ++c)
786 for(int d = 0; d < 6; ++d)
787 buffer[idx++] = INT__k_g_p_d[q*9720+a*270+b*18+d*3+c];
788
789 memcpy(INT__k_g_p_d+q*9720, buffer, 9720*sizeof(double));
790 }
791
792 return ret;
793 }
794
ostei_g_k_p_d(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__g_k_p_d)795 int ostei_g_k_p_d(struct simint_multi_shellpair const P,
796 struct simint_multi_shellpair const Q,
797 double screen_tol,
798 double * const restrict work,
799 double * const restrict INT__g_k_p_d)
800 {
801 double P_AB[3*P.nshell12];
802 struct simint_multi_shellpair P_tmp = P;
803 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
804 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
805 P_tmp.AB_x = P_AB;
806 P_tmp.AB_y = P_AB + P.nshell12;
807 P_tmp.AB_z = P_AB + 2*P.nshell12;
808
809 for(int i = 0; i < P.nshell12; i++)
810 {
811 P_tmp.AB_x[i] = -P.AB_x[i];
812 P_tmp.AB_y[i] = -P.AB_y[i];
813 P_tmp.AB_z[i] = -P.AB_z[i];
814 }
815
816 double Q_AB[3*Q.nshell12];
817 struct simint_multi_shellpair Q_tmp = Q;
818 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
819 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
820 Q_tmp.AB_x = Q_AB;
821 Q_tmp.AB_y = Q_AB + Q.nshell12;
822 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
823
824 for(int i = 0; i < Q.nshell12; i++)
825 {
826 Q_tmp.AB_x[i] = -Q.AB_x[i];
827 Q_tmp.AB_y[i] = -Q.AB_y[i];
828 Q_tmp.AB_z[i] = -Q.AB_z[i];
829 }
830
831 int ret = ostei_k_g_d_p(P_tmp, Q_tmp, screen_tol, work, INT__g_k_p_d);
832 double buffer[9720] SIMINT_ALIGN_ARRAY_DBL;
833
834 for(int q = 0; q < ret; q++)
835 {
836 int idx = 0;
837 for(int a = 0; a < 15; ++a)
838 for(int b = 0; b < 36; ++b)
839 for(int c = 0; c < 3; ++c)
840 for(int d = 0; d < 6; ++d)
841 buffer[idx++] = INT__g_k_p_d[q*9720+b*270+a*18+d*3+c];
842
843 memcpy(INT__g_k_p_d+q*9720, buffer, 9720*sizeof(double));
844 }
845
846 return ret;
847 }
848
849