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_h_p_k_d(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_p_k_d)8 int ostei_h_p_k_d(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__h_p_k_d)
13 {
14
15 SIMINT_ASSUME_ALIGN_DBL(work);
16 SIMINT_ASSUME_ALIGN_DBL(INT__h_p_k_d);
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__h_s_k_s = work + (SIMINT_NSHELL_SIMD * 0);
30 double * const INT__h_s_l_s = work + (SIMINT_NSHELL_SIMD * 756);
31 double * const INT__h_s_m_s = work + (SIMINT_NSHELL_SIMD * 1701);
32 double * const INT__i_s_k_s = work + (SIMINT_NSHELL_SIMD * 2856);
33 double * const INT__i_s_l_s = work + (SIMINT_NSHELL_SIMD * 3864);
34 double * const INT__i_s_m_s = work + (SIMINT_NSHELL_SIMD * 5124);
35 SIMINT_DBLTYPE * const primwork = (SIMINT_DBLTYPE *)(work + SIMINT_NSHELL_SIMD*6664);
36 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_s_s = primwork + 0;
37 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_p_s = primwork + 16;
38 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_d_s = primwork + 61;
39 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_f_s = primwork + 145;
40 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_g_s = primwork + 275;
41 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_h_s = primwork + 455;
42 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_i_s = primwork + 686;
43 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_k_s = primwork + 966;
44 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_l_s = primwork + 1290;
45 SIMINT_DBLTYPE * const restrict PRIM_INT__s_s_m_s = primwork + 1650;
46 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_d_s = primwork + 2035;
47 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_f_s = primwork + 2143;
48 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_g_s = primwork + 2323;
49 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_h_s = primwork + 2593;
50 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_i_s = primwork + 2971;
51 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_k_s = primwork + 3475;
52 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_l_s = primwork + 4123;
53 SIMINT_DBLTYPE * const restrict PRIM_INT__p_s_m_s = primwork + 4933;
54 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_f_s = primwork + 5923;
55 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_g_s = primwork + 6223;
56 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_h_s = primwork + 6673;
57 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_i_s = primwork + 7303;
58 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_k_s = primwork + 8143;
59 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_l_s = primwork + 9223;
60 SIMINT_DBLTYPE * const restrict PRIM_INT__d_s_m_s = primwork + 10573;
61 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_g_s = primwork + 12223;
62 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_h_s = primwork + 12823;
63 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_i_s = primwork + 13663;
64 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_k_s = primwork + 14783;
65 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_l_s = primwork + 16223;
66 SIMINT_DBLTYPE * const restrict PRIM_INT__f_s_m_s = primwork + 18023;
67 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_h_s = primwork + 20223;
68 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_i_s = primwork + 21168;
69 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_k_s = primwork + 22428;
70 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_l_s = primwork + 24048;
71 SIMINT_DBLTYPE * const restrict PRIM_INT__g_s_m_s = primwork + 26073;
72 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_i_s = primwork + 28548;
73 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_k_s = primwork + 29724;
74 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_l_s = primwork + 31236;
75 SIMINT_DBLTYPE * const restrict PRIM_INT__h_s_m_s = primwork + 33126;
76 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_k_s = primwork + 35436;
77 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_l_s = primwork + 36444;
78 SIMINT_DBLTYPE * const restrict PRIM_INT__i_s_m_s = primwork + 37704;
79 double * const hrrwork = (double *)(primwork + 39244);
80 double * const HRR_INT__h_p_k_s = hrrwork + 0;
81 double * const HRR_INT__h_p_k_p = hrrwork + 2268;
82 double * const HRR_INT__h_p_l_s = hrrwork + 9072;
83 double * const HRR_INT__h_p_l_p = hrrwork + 11907;
84 double * const HRR_INT__h_p_m_s = hrrwork + 20412;
85
86
87 // Create constants
88 const SIMINT_DBLTYPE const_1 = SIMINT_DBLSET1(1);
89 const SIMINT_DBLTYPE const_2 = SIMINT_DBLSET1(2);
90 const SIMINT_DBLTYPE const_3 = SIMINT_DBLSET1(3);
91 const SIMINT_DBLTYPE const_4 = SIMINT_DBLSET1(4);
92 const SIMINT_DBLTYPE const_5 = SIMINT_DBLSET1(5);
93 const SIMINT_DBLTYPE const_6 = SIMINT_DBLSET1(6);
94 const SIMINT_DBLTYPE const_7 = SIMINT_DBLSET1(7);
95 const SIMINT_DBLTYPE const_8 = SIMINT_DBLSET1(8);
96 const SIMINT_DBLTYPE const_9 = SIMINT_DBLSET1(9);
97 const SIMINT_DBLTYPE one_half = SIMINT_DBLSET1(0.5);
98
99
100 ////////////////////////////////////////
101 // Loop over shells and primitives
102 ////////////////////////////////////////
103
104 real_abcd = 0;
105 istart = 0;
106 for(ab = 0; ab < P.nshell12_clip; ++ab)
107 {
108 const int iend = istart + P.nprim12[ab];
109
110 cd = 0;
111 jstart = 0;
112
113 for(cd = 0; cd < Q.nshell12_clip; cd += SIMINT_NSHELL_SIMD)
114 {
115 const int nshellbatch = ((cd + SIMINT_NSHELL_SIMD) > Q.nshell12_clip) ? Q.nshell12_clip - cd : SIMINT_NSHELL_SIMD;
116 int jend = jstart;
117 for(i = 0; i < nshellbatch; i++)
118 jend += Q.nprim12[cd+i];
119
120 // Clear the beginning of the workspace (where we are accumulating integrals)
121 memset(work, 0, SIMINT_NSHELL_SIMD * 6664 * sizeof(double));
122 abcd = 0;
123
124
125 for(i = istart; i < iend; ++i)
126 {
127 SIMINT_DBLTYPE bra_screen_max; // only used if check_screen
128
129 if(check_screen)
130 {
131 // Skip this whole thing if always insignificant
132 if((P.screen[i] * Q.screen_max) < screen_tol)
133 continue;
134 bra_screen_max = SIMINT_DBLSET1(P.screen[i]);
135 }
136
137 icd = 0;
138 iprimcd = 0;
139 nprim_icd = Q.nprim12[cd];
140 double * restrict PRIM_PTR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
141 double * restrict PRIM_PTR_INT__h_s_l_s = INT__h_s_l_s + abcd * 945;
142 double * restrict PRIM_PTR_INT__h_s_m_s = INT__h_s_m_s + abcd * 1155;
143 double * restrict PRIM_PTR_INT__i_s_k_s = INT__i_s_k_s + abcd * 1008;
144 double * restrict PRIM_PTR_INT__i_s_l_s = INT__i_s_l_s + abcd * 1260;
145 double * restrict PRIM_PTR_INT__i_s_m_s = INT__i_s_m_s + abcd * 1540;
146
147
148
149 // Load these one per loop over i
150 const SIMINT_DBLTYPE P_alpha = SIMINT_DBLSET1(P.alpha[i]);
151 const SIMINT_DBLTYPE P_prefac = SIMINT_DBLSET1(P.prefac[i]);
152 const SIMINT_DBLTYPE Pxyz[3] = { SIMINT_DBLSET1(P.x[i]), SIMINT_DBLSET1(P.y[i]), SIMINT_DBLSET1(P.z[i]) };
153
154 const SIMINT_DBLTYPE P_PA[3] = { SIMINT_DBLSET1(P.PA_x[i]), SIMINT_DBLSET1(P.PA_y[i]), SIMINT_DBLSET1(P.PA_z[i]) };
155
156 for(j = jstart; j < jend; j += SIMINT_SIMD_LEN)
157 {
158 // calculate the shell offsets
159 // these are the offset from the shell pointed to by cd
160 // for each element
161 int shelloffsets[SIMINT_SIMD_LEN] = {0};
162 int lastoffset = 0;
163 const int nlane = ( ((j + SIMINT_SIMD_LEN) < jend) ? SIMINT_SIMD_LEN : (jend - j));
164
165 if((iprimcd + SIMINT_SIMD_LEN) >= nprim_icd)
166 {
167 // Handle if the first element of the vector is a new shell
168 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
169 {
170 nprim_icd += Q.nprim12[cd + (++icd)];
171 PRIM_PTR_INT__h_s_k_s += 756;
172 PRIM_PTR_INT__h_s_l_s += 945;
173 PRIM_PTR_INT__h_s_m_s += 1155;
174 PRIM_PTR_INT__i_s_k_s += 1008;
175 PRIM_PTR_INT__i_s_l_s += 1260;
176 PRIM_PTR_INT__i_s_m_s += 1540;
177 }
178 iprimcd++;
179 for(n = 1; n < SIMINT_SIMD_LEN; ++n)
180 {
181 if(iprimcd >= nprim_icd && ((icd+1) < nshellbatch))
182 {
183 shelloffsets[n] = shelloffsets[n-1] + 1;
184 lastoffset++;
185 nprim_icd += Q.nprim12[cd + (++icd)];
186 }
187 else
188 shelloffsets[n] = shelloffsets[n-1];
189 iprimcd++;
190 }
191 }
192 else
193 iprimcd += SIMINT_SIMD_LEN;
194
195 // Do we have to compute this vector (or has it been screened out)?
196 // (not_screened != 0 means we have to do this vector)
197 if(check_screen)
198 {
199 const double vmax = vector_max(SIMINT_MUL(bra_screen_max, SIMINT_DBLLOAD(Q.screen, j)));
200 if(vmax < screen_tol)
201 {
202 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
203 PRIM_PTR_INT__h_s_l_s += lastoffset*945;
204 PRIM_PTR_INT__h_s_m_s += lastoffset*1155;
205 PRIM_PTR_INT__i_s_k_s += lastoffset*1008;
206 PRIM_PTR_INT__i_s_l_s += lastoffset*1260;
207 PRIM_PTR_INT__i_s_m_s += lastoffset*1540;
208 continue;
209 }
210 }
211
212 const SIMINT_DBLTYPE Q_alpha = SIMINT_DBLLOAD(Q.alpha, j);
213 const SIMINT_DBLTYPE PQalpha_mul = SIMINT_MUL(P_alpha, Q_alpha);
214 const SIMINT_DBLTYPE PQalpha_sum = SIMINT_ADD(P_alpha, Q_alpha);
215 const SIMINT_DBLTYPE one_over_PQalpha_sum = SIMINT_DIV(const_1, PQalpha_sum);
216
217
218 /* construct R2 = (Px - Qx)**2 + (Py - Qy)**2 + (Pz -Qz)**2 */
219 SIMINT_DBLTYPE PQ[3];
220 PQ[0] = SIMINT_SUB(Pxyz[0], SIMINT_DBLLOAD(Q.x, j));
221 PQ[1] = SIMINT_SUB(Pxyz[1], SIMINT_DBLLOAD(Q.y, j));
222 PQ[2] = SIMINT_SUB(Pxyz[2], SIMINT_DBLLOAD(Q.z, j));
223 SIMINT_DBLTYPE R2 = SIMINT_MUL(PQ[0], PQ[0]);
224 R2 = SIMINT_FMADD(PQ[1], PQ[1], R2);
225 R2 = SIMINT_FMADD(PQ[2], PQ[2], R2);
226
227 const SIMINT_DBLTYPE alpha = SIMINT_MUL(PQalpha_mul, one_over_PQalpha_sum); // alpha from MEST
228 const SIMINT_DBLTYPE one_over_p = SIMINT_DIV(const_1, P_alpha);
229 const SIMINT_DBLTYPE one_over_q = SIMINT_DIV(const_1, Q_alpha);
230 const SIMINT_DBLTYPE one_over_2p = SIMINT_MUL(one_half, one_over_p);
231 const SIMINT_DBLTYPE one_over_2q = SIMINT_MUL(one_half, one_over_q);
232 const SIMINT_DBLTYPE one_over_2pq = SIMINT_MUL(one_half, one_over_PQalpha_sum);
233 const SIMINT_DBLTYPE Q_PA[3] = { SIMINT_DBLLOAD(Q.PA_x, j), SIMINT_DBLLOAD(Q.PA_y, j), SIMINT_DBLLOAD(Q.PA_z, j) };
234
235 // NOTE: Minus sign!
236 const SIMINT_DBLTYPE a_over_p = SIMINT_MUL(SIMINT_NEG(alpha), one_over_p);
237 SIMINT_DBLTYPE aop_PQ[3];
238 aop_PQ[0] = SIMINT_MUL(a_over_p, PQ[0]);
239 aop_PQ[1] = SIMINT_MUL(a_over_p, PQ[1]);
240 aop_PQ[2] = SIMINT_MUL(a_over_p, PQ[2]);
241
242 SIMINT_DBLTYPE a_over_q = SIMINT_MUL(alpha, one_over_q);
243 SIMINT_DBLTYPE aoq_PQ[3];
244 aoq_PQ[0] = SIMINT_MUL(a_over_q, PQ[0]);
245 aoq_PQ[1] = SIMINT_MUL(a_over_q, PQ[1]);
246 aoq_PQ[2] = SIMINT_MUL(a_over_q, PQ[2]);
247 // Put a minus sign here so we don't have to in RR routines
248 a_over_q = SIMINT_NEG(a_over_q);
249
250
251 //////////////////////////////////////////////
252 // Fjt function section
253 // Maximum v value: 15
254 //////////////////////////////////////////////
255 // The parameter to the Fjt function
256 const SIMINT_DBLTYPE F_x = SIMINT_MUL(R2, alpha);
257
258
259 const SIMINT_DBLTYPE Q_prefac = mask_load(nlane, Q.prefac + j);
260
261
262 boys_F_split(PRIM_INT__s_s_s_s, F_x, 15);
263 SIMINT_DBLTYPE prefac = SIMINT_SQRT(one_over_PQalpha_sum);
264 prefac = SIMINT_MUL(SIMINT_MUL(P_prefac, Q_prefac), prefac);
265 for(n = 0; n <= 15; n++)
266 PRIM_INT__s_s_s_s[n] = SIMINT_MUL(PRIM_INT__s_s_s_s[n], prefac);
267
268 //////////////////////////////////////////////
269 // Primitive integrals: Vertical recurrance
270 //////////////////////////////////////////////
271
272 const SIMINT_DBLTYPE vrr_const_1_over_2p = one_over_2p;
273 const SIMINT_DBLTYPE vrr_const_2_over_2p = SIMINT_MUL(const_2, one_over_2p);
274 const SIMINT_DBLTYPE vrr_const_3_over_2p = SIMINT_MUL(const_3, one_over_2p);
275 const SIMINT_DBLTYPE vrr_const_4_over_2p = SIMINT_MUL(const_4, one_over_2p);
276 const SIMINT_DBLTYPE vrr_const_5_over_2p = SIMINT_MUL(const_5, one_over_2p);
277 const SIMINT_DBLTYPE vrr_const_1_over_2q = one_over_2q;
278 const SIMINT_DBLTYPE vrr_const_2_over_2q = SIMINT_MUL(const_2, one_over_2q);
279 const SIMINT_DBLTYPE vrr_const_3_over_2q = SIMINT_MUL(const_3, one_over_2q);
280 const SIMINT_DBLTYPE vrr_const_4_over_2q = SIMINT_MUL(const_4, one_over_2q);
281 const SIMINT_DBLTYPE vrr_const_5_over_2q = SIMINT_MUL(const_5, one_over_2q);
282 const SIMINT_DBLTYPE vrr_const_6_over_2q = SIMINT_MUL(const_6, one_over_2q);
283 const SIMINT_DBLTYPE vrr_const_7_over_2q = SIMINT_MUL(const_7, one_over_2q);
284 const SIMINT_DBLTYPE vrr_const_8_over_2q = SIMINT_MUL(const_8, one_over_2q);
285 const SIMINT_DBLTYPE vrr_const_1_over_2pq = one_over_2pq;
286 const SIMINT_DBLTYPE vrr_const_2_over_2pq = SIMINT_MUL(const_2, one_over_2pq);
287 const SIMINT_DBLTYPE vrr_const_3_over_2pq = SIMINT_MUL(const_3, one_over_2pq);
288 const SIMINT_DBLTYPE vrr_const_4_over_2pq = SIMINT_MUL(const_4, one_over_2pq);
289 const SIMINT_DBLTYPE vrr_const_5_over_2pq = SIMINT_MUL(const_5, one_over_2pq);
290 const SIMINT_DBLTYPE vrr_const_6_over_2pq = SIMINT_MUL(const_6, one_over_2pq);
291 const SIMINT_DBLTYPE vrr_const_7_over_2pq = SIMINT_MUL(const_7, one_over_2pq);
292 const SIMINT_DBLTYPE vrr_const_8_over_2pq = SIMINT_MUL(const_8, one_over_2pq);
293 const SIMINT_DBLTYPE vrr_const_9_over_2pq = SIMINT_MUL(const_9, one_over_2pq);
294
295
296
297 // Forming PRIM_INT__s_s_p_s[15 * 3];
298 for(n = 0; n < 15; ++n) // loop over orders of auxiliary function
299 {
300
301 PRIM_INT__s_s_p_s[n * 3 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_s_s[n * 1 + 0]);
302 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]);
303
304 PRIM_INT__s_s_p_s[n * 3 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_s_s[n * 1 + 0]);
305 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]);
306
307 PRIM_INT__s_s_p_s[n * 3 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_s_s[n * 1 + 0]);
308 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]);
309
310 }
311
312
313
314 // Forming PRIM_INT__s_s_d_s[14 * 6];
315 for(n = 0; n < 14; ++n) // loop over orders of auxiliary function
316 {
317
318 PRIM_INT__s_s_d_s[n * 6 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_p_s[n * 3 + 0]);
319 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]);
320 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]);
321
322 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 0]);
323 PRIM_INT__s_s_d_s[n * 6 + 1] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_d_s[n * 6 + 1]);
324
325 PRIM_INT__s_s_d_s[n * 6 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 0]);
326 PRIM_INT__s_s_d_s[n * 6 + 2] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__s_s_d_s[n * 6 + 2]);
327
328 PRIM_INT__s_s_d_s[n * 6 + 3] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_p_s[n * 3 + 1]);
329 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]);
330 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]);
331
332 PRIM_INT__s_s_d_s[n * 6 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 1]);
333 PRIM_INT__s_s_d_s[n * 6 + 4] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__s_s_d_s[n * 6 + 4]);
334
335 PRIM_INT__s_s_d_s[n * 6 + 5] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_p_s[n * 3 + 2]);
336 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]);
337 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]);
338
339 }
340
341
342
343 // Forming PRIM_INT__s_s_f_s[13 * 10];
344 for(n = 0; n < 13; ++n) // loop over orders of auxiliary function
345 {
346
347 PRIM_INT__s_s_f_s[n * 10 + 0] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
348 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]);
349 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]);
350
351 PRIM_INT__s_s_f_s[n * 10 + 1] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
352 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]);
353
354 PRIM_INT__s_s_f_s[n * 10 + 2] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
355 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]);
356
357 PRIM_INT__s_s_f_s[n * 10 + 3] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
358 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]);
359
360 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
361 PRIM_INT__s_s_f_s[n * 10 + 4] = SIMINT_FMADD( aoq_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__s_s_f_s[n * 10 + 4]);
362
363 PRIM_INT__s_s_f_s[n * 10 + 5] = SIMINT_MUL(Q_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
364 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]);
365
366 PRIM_INT__s_s_f_s[n * 10 + 6] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
367 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]);
368 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]);
369
370 PRIM_INT__s_s_f_s[n * 10 + 7] = SIMINT_MUL(Q_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
371 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]);
372
373 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_MUL(Q_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
374 PRIM_INT__s_s_f_s[n * 10 + 8] = SIMINT_FMADD( aoq_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__s_s_f_s[n * 10 + 8]);
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 12);
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 11);
403
404
405 ostei_general_vrr1_K(6, 10,
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, 9,
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, 6,
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, 6,
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, 5,
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_vrr_I(1, 0, 5, 0, 6,
431 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
432 PRIM_INT__s_s_h_s, NULL, NULL, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_h_s);
433
434
435 ostei_general_vrr_I(2, 0, 6, 0, 5,
436 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
437 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);
438
439
440 ostei_general_vrr_I(3, 0, 7, 0, 4,
441 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
442 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);
443
444
445 VRR_I_p_s_g_s(
446 PRIM_INT__p_s_g_s,
447 PRIM_INT__s_s_g_s,
448 PRIM_INT__s_s_f_s,
449 P_PA,
450 aop_PQ,
451 one_over_2pq,
452 6);
453
454
455 ostei_general_vrr_I(2, 0, 5, 0, 5,
456 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
457 PRIM_INT__p_s_h_s, PRIM_INT__s_s_h_s, NULL, PRIM_INT__p_s_g_s, NULL, PRIM_INT__d_s_h_s);
458
459
460 ostei_general_vrr_I(3, 0, 6, 0, 4,
461 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
462 PRIM_INT__d_s_i_s, PRIM_INT__p_s_i_s, NULL, PRIM_INT__d_s_h_s, NULL, PRIM_INT__f_s_i_s);
463
464
465 ostei_general_vrr_I(4, 0, 7, 0, 3,
466 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
467 PRIM_INT__f_s_k_s, PRIM_INT__d_s_k_s, NULL, PRIM_INT__f_s_i_s, NULL, PRIM_INT__g_s_k_s);
468
469
470 VRR_I_p_s_f_s(
471 PRIM_INT__p_s_f_s,
472 PRIM_INT__s_s_f_s,
473 PRIM_INT__s_s_d_s,
474 P_PA,
475 aop_PQ,
476 one_over_2pq,
477 6);
478
479
480 ostei_general_vrr_I(2, 0, 4, 0, 5,
481 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
482 PRIM_INT__p_s_g_s, PRIM_INT__s_s_g_s, NULL, PRIM_INT__p_s_f_s, NULL, PRIM_INT__d_s_g_s);
483
484
485 ostei_general_vrr_I(3, 0, 5, 0, 4,
486 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
487 PRIM_INT__d_s_h_s, PRIM_INT__p_s_h_s, NULL, PRIM_INT__d_s_g_s, NULL, PRIM_INT__f_s_h_s);
488
489
490 ostei_general_vrr_I(4, 0, 6, 0, 3,
491 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
492 PRIM_INT__f_s_i_s, PRIM_INT__d_s_i_s, NULL, PRIM_INT__f_s_h_s, NULL, PRIM_INT__g_s_i_s);
493
494
495 ostei_general_vrr_I(5, 0, 7, 0, 2,
496 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
497 PRIM_INT__g_s_k_s, PRIM_INT__f_s_k_s, NULL, PRIM_INT__g_s_i_s, NULL, PRIM_INT__h_s_k_s);
498
499
500 ostei_general_vrr1_K(8, 8,
501 one_over_2q, a_over_q, aoq_PQ, Q_PA,
502 PRIM_INT__s_s_k_s, PRIM_INT__s_s_i_s, PRIM_INT__s_s_l_s);
503
504
505 ostei_general_vrr_I(1, 0, 8, 0, 6,
506 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
507 PRIM_INT__s_s_l_s, NULL, NULL, PRIM_INT__s_s_k_s, NULL, PRIM_INT__p_s_l_s);
508
509
510 ostei_general_vrr_I(2, 0, 8, 0, 5,
511 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
512 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);
513
514
515 ostei_general_vrr_I(3, 0, 8, 0, 4,
516 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
517 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);
518
519
520 ostei_general_vrr_I(4, 0, 8, 0, 3,
521 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
522 PRIM_INT__f_s_l_s, PRIM_INT__d_s_l_s, NULL, PRIM_INT__f_s_k_s, NULL, PRIM_INT__g_s_l_s);
523
524
525 ostei_general_vrr_I(5, 0, 8, 0, 2,
526 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
527 PRIM_INT__g_s_l_s, PRIM_INT__f_s_l_s, NULL, PRIM_INT__g_s_k_s, NULL, PRIM_INT__h_s_l_s);
528
529
530
531 // Forming PRIM_INT__p_s_d_s[6 * 18];
532 for(n = 0; n < 6; ++n) // loop over orders of auxiliary function
533 {
534
535 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 0]);
536 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 0]);
537 PRIM_INT__p_s_d_s[n * 18 + 0] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 0]);
538
539 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 1]);
540 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 1]);
541 PRIM_INT__p_s_d_s[n * 18 + 1] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 1]);
542
543 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 2]);
544 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 2]);
545 PRIM_INT__p_s_d_s[n * 18 + 2] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 2]);
546
547 PRIM_INT__p_s_d_s[n * 18 + 3] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 3]);
548 PRIM_INT__p_s_d_s[n * 18 + 3] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 3]);
549
550 PRIM_INT__p_s_d_s[n * 18 + 4] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 4]);
551 PRIM_INT__p_s_d_s[n * 18 + 4] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 4]);
552
553 PRIM_INT__p_s_d_s[n * 18 + 5] = SIMINT_MUL(P_PA[0], PRIM_INT__s_s_d_s[n * 6 + 5]);
554 PRIM_INT__p_s_d_s[n * 18 + 5] = SIMINT_FMADD( aop_PQ[0], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 5]);
555
556 PRIM_INT__p_s_d_s[n * 18 + 6] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 0]);
557 PRIM_INT__p_s_d_s[n * 18 + 6] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 6]);
558
559 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 1]);
560 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 7]);
561 PRIM_INT__p_s_d_s[n * 18 + 7] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 7]);
562
563 PRIM_INT__p_s_d_s[n * 18 + 8] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 2]);
564 PRIM_INT__p_s_d_s[n * 18 + 8] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 8]);
565
566 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 3]);
567 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 9]);
568 PRIM_INT__p_s_d_s[n * 18 + 9] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 9]);
569
570 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 4]);
571 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 10]);
572 PRIM_INT__p_s_d_s[n * 18 + 10] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 10]);
573
574 PRIM_INT__p_s_d_s[n * 18 + 11] = SIMINT_MUL(P_PA[1], PRIM_INT__s_s_d_s[n * 6 + 5]);
575 PRIM_INT__p_s_d_s[n * 18 + 11] = SIMINT_FMADD( aop_PQ[1], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 11]);
576
577 PRIM_INT__p_s_d_s[n * 18 + 12] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 0]);
578 PRIM_INT__p_s_d_s[n * 18 + 12] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 0], PRIM_INT__p_s_d_s[n * 18 + 12]);
579
580 PRIM_INT__p_s_d_s[n * 18 + 13] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 1]);
581 PRIM_INT__p_s_d_s[n * 18 + 13] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 1], PRIM_INT__p_s_d_s[n * 18 + 13]);
582
583 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 2]);
584 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 2], PRIM_INT__p_s_d_s[n * 18 + 14]);
585 PRIM_INT__p_s_d_s[n * 18 + 14] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 0], PRIM_INT__p_s_d_s[n * 18 + 14]);
586
587 PRIM_INT__p_s_d_s[n * 18 + 15] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 3]);
588 PRIM_INT__p_s_d_s[n * 18 + 15] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 3], PRIM_INT__p_s_d_s[n * 18 + 15]);
589
590 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 4]);
591 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 4], PRIM_INT__p_s_d_s[n * 18 + 16]);
592 PRIM_INT__p_s_d_s[n * 18 + 16] = SIMINT_FMADD( vrr_const_1_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 1], PRIM_INT__p_s_d_s[n * 18 + 16]);
593
594 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_MUL(P_PA[2], PRIM_INT__s_s_d_s[n * 6 + 5]);
595 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_FMADD( aop_PQ[2], PRIM_INT__s_s_d_s[(n+1) * 6 + 5], PRIM_INT__p_s_d_s[n * 18 + 17]);
596 PRIM_INT__p_s_d_s[n * 18 + 17] = SIMINT_FMADD( vrr_const_2_over_2pq, PRIM_INT__s_s_p_s[(n+1) * 3 + 2], PRIM_INT__p_s_d_s[n * 18 + 17]);
597
598 }
599
600
601 VRR_I_d_s_f_s(
602 PRIM_INT__d_s_f_s,
603 PRIM_INT__p_s_f_s,
604 PRIM_INT__s_s_f_s,
605 PRIM_INT__p_s_d_s,
606 P_PA,
607 a_over_p,
608 aop_PQ,
609 one_over_2p,
610 one_over_2pq,
611 5);
612
613
614 ostei_general_vrr_I(3, 0, 4, 0, 4,
615 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
616 PRIM_INT__d_s_g_s, PRIM_INT__p_s_g_s, NULL, PRIM_INT__d_s_f_s, NULL, PRIM_INT__f_s_g_s);
617
618
619 ostei_general_vrr_I(4, 0, 5, 0, 3,
620 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
621 PRIM_INT__f_s_h_s, PRIM_INT__d_s_h_s, NULL, PRIM_INT__f_s_g_s, NULL, PRIM_INT__g_s_h_s);
622
623
624 ostei_general_vrr_I(5, 0, 6, 0, 2,
625 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
626 PRIM_INT__g_s_i_s, PRIM_INT__f_s_i_s, NULL, PRIM_INT__g_s_h_s, NULL, PRIM_INT__h_s_i_s);
627
628
629 ostei_general_vrr_I(6, 0, 7, 0, 1,
630 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
631 PRIM_INT__h_s_k_s, PRIM_INT__g_s_k_s, NULL, PRIM_INT__h_s_i_s, NULL, PRIM_INT__i_s_k_s);
632
633
634 ostei_general_vrr1_K(9, 7,
635 one_over_2q, a_over_q, aoq_PQ, Q_PA,
636 PRIM_INT__s_s_l_s, PRIM_INT__s_s_k_s, PRIM_INT__s_s_m_s);
637
638
639 ostei_general_vrr_I(1, 0, 9, 0, 6,
640 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
641 PRIM_INT__s_s_m_s, NULL, NULL, PRIM_INT__s_s_l_s, NULL, PRIM_INT__p_s_m_s);
642
643
644 ostei_general_vrr_I(2, 0, 9, 0, 5,
645 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
646 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);
647
648
649 ostei_general_vrr_I(3, 0, 9, 0, 4,
650 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
651 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);
652
653
654 ostei_general_vrr_I(4, 0, 9, 0, 3,
655 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
656 PRIM_INT__f_s_m_s, PRIM_INT__d_s_m_s, NULL, PRIM_INT__f_s_l_s, NULL, PRIM_INT__g_s_m_s);
657
658
659 ostei_general_vrr_I(5, 0, 9, 0, 2,
660 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
661 PRIM_INT__g_s_m_s, PRIM_INT__f_s_m_s, NULL, PRIM_INT__g_s_l_s, NULL, PRIM_INT__h_s_m_s);
662
663
664 ostei_general_vrr_I(6, 0, 8, 0, 1,
665 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
666 PRIM_INT__h_s_l_s, PRIM_INT__g_s_l_s, NULL, PRIM_INT__h_s_k_s, NULL, PRIM_INT__i_s_l_s);
667
668
669 ostei_general_vrr_I(6, 0, 9, 0, 1,
670 one_over_2p, a_over_p, one_over_2pq, aop_PQ, P_PA,
671 PRIM_INT__h_s_m_s, PRIM_INT__g_s_m_s, NULL, PRIM_INT__h_s_l_s, NULL, PRIM_INT__i_s_m_s);
672
673
674
675
676 ////////////////////////////////////
677 // Accumulate contracted integrals
678 ////////////////////////////////////
679 if(lastoffset == 0)
680 {
681 contract_all(756, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
682 contract_all(945, PRIM_INT__h_s_l_s, PRIM_PTR_INT__h_s_l_s);
683 contract_all(1155, PRIM_INT__h_s_m_s, PRIM_PTR_INT__h_s_m_s);
684 contract_all(1008, PRIM_INT__i_s_k_s, PRIM_PTR_INT__i_s_k_s);
685 contract_all(1260, PRIM_INT__i_s_l_s, PRIM_PTR_INT__i_s_l_s);
686 contract_all(1540, PRIM_INT__i_s_m_s, PRIM_PTR_INT__i_s_m_s);
687 }
688 else
689 {
690 contract(756, shelloffsets, PRIM_INT__h_s_k_s, PRIM_PTR_INT__h_s_k_s);
691 contract(945, shelloffsets, PRIM_INT__h_s_l_s, PRIM_PTR_INT__h_s_l_s);
692 contract(1155, shelloffsets, PRIM_INT__h_s_m_s, PRIM_PTR_INT__h_s_m_s);
693 contract(1008, shelloffsets, PRIM_INT__i_s_k_s, PRIM_PTR_INT__i_s_k_s);
694 contract(1260, shelloffsets, PRIM_INT__i_s_l_s, PRIM_PTR_INT__i_s_l_s);
695 contract(1540, shelloffsets, PRIM_INT__i_s_m_s, PRIM_PTR_INT__i_s_m_s);
696 PRIM_PTR_INT__h_s_k_s += lastoffset*756;
697 PRIM_PTR_INT__h_s_l_s += lastoffset*945;
698 PRIM_PTR_INT__h_s_m_s += lastoffset*1155;
699 PRIM_PTR_INT__i_s_k_s += lastoffset*1008;
700 PRIM_PTR_INT__i_s_l_s += lastoffset*1260;
701 PRIM_PTR_INT__i_s_m_s += lastoffset*1540;
702 }
703
704 } // close loop over j
705 } // close loop over i
706
707 //Advance to the next batch
708 jstart = SIMINT_SIMD_ROUND(jend);
709
710 //////////////////////////////////////////////
711 // Contracted integrals: Horizontal recurrance
712 //////////////////////////////////////////////
713
714
715 const double hAB[3] = { P.AB_x[ab], P.AB_y[ab], P.AB_z[ab] };
716
717
718 for(abcd = 0; abcd < nshellbatch; ++abcd, ++real_abcd)
719 {
720 const double hCD[3] = { Q.AB_x[cd+abcd], Q.AB_y[cd+abcd], Q.AB_z[cd+abcd] };
721
722 // set up HRR pointers
723 double const * restrict HRR_INT__h_s_k_s = INT__h_s_k_s + abcd * 756;
724 double const * restrict HRR_INT__h_s_l_s = INT__h_s_l_s + abcd * 945;
725 double const * restrict HRR_INT__h_s_m_s = INT__h_s_m_s + abcd * 1155;
726 double const * restrict HRR_INT__i_s_k_s = INT__i_s_k_s + abcd * 1008;
727 double const * restrict HRR_INT__i_s_l_s = INT__i_s_l_s + abcd * 1260;
728 double const * restrict HRR_INT__i_s_m_s = INT__i_s_m_s + abcd * 1540;
729 double * restrict HRR_INT__h_p_k_d = INT__h_p_k_d + real_abcd * 13608;
730
731 // form INT__h_p_k_s
732 ostei_general_hrr_J(5, 1, 7, 0, hAB, HRR_INT__i_s_k_s, HRR_INT__h_s_k_s, HRR_INT__h_p_k_s);
733
734 // form INT__h_p_l_s
735 ostei_general_hrr_J(5, 1, 8, 0, hAB, HRR_INT__i_s_l_s, HRR_INT__h_s_l_s, HRR_INT__h_p_l_s);
736
737 // form INT__h_p_m_s
738 ostei_general_hrr_J(5, 1, 9, 0, hAB, HRR_INT__i_s_m_s, HRR_INT__h_s_m_s, HRR_INT__h_p_m_s);
739
740 // form INT__h_p_k_p
741 ostei_general_hrr_L(5, 1, 7, 1, hCD, HRR_INT__h_p_l_s, HRR_INT__h_p_k_s, HRR_INT__h_p_k_p);
742
743 // form INT__h_p_l_p
744 ostei_general_hrr_L(5, 1, 8, 1, hCD, HRR_INT__h_p_m_s, HRR_INT__h_p_l_s, HRR_INT__h_p_l_p);
745
746 // form INT__h_p_k_d
747 ostei_general_hrr_L(5, 1, 7, 2, hCD, HRR_INT__h_p_l_p, HRR_INT__h_p_k_p, HRR_INT__h_p_k_d);
748
749
750 } // close HRR loop
751
752
753 } // close loop cdbatch
754
755 istart = iend;
756 } // close loop over ab
757
758 return P.nshell12_clip * Q.nshell12_clip;
759 }
760
ostei_p_h_k_d(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__p_h_k_d)761 int ostei_p_h_k_d(struct simint_multi_shellpair const P,
762 struct simint_multi_shellpair const Q,
763 double screen_tol,
764 double * const restrict work,
765 double * const restrict INT__p_h_k_d)
766 {
767 double P_AB[3*P.nshell12];
768 struct simint_multi_shellpair P_tmp = P;
769 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
770 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
771 P_tmp.AB_x = P_AB;
772 P_tmp.AB_y = P_AB + P.nshell12;
773 P_tmp.AB_z = P_AB + 2*P.nshell12;
774
775 for(int i = 0; i < P.nshell12; i++)
776 {
777 P_tmp.AB_x[i] = -P.AB_x[i];
778 P_tmp.AB_y[i] = -P.AB_y[i];
779 P_tmp.AB_z[i] = -P.AB_z[i];
780 }
781
782 int ret = ostei_h_p_k_d(P_tmp, Q, screen_tol, work, INT__p_h_k_d);
783 double buffer[13608] SIMINT_ALIGN_ARRAY_DBL;
784
785 for(int q = 0; q < ret; q++)
786 {
787 int idx = 0;
788 for(int a = 0; a < 3; ++a)
789 for(int b = 0; b < 21; ++b)
790 for(int c = 0; c < 36; ++c)
791 for(int d = 0; d < 6; ++d)
792 buffer[idx++] = INT__p_h_k_d[q*13608+b*648+a*216+c*6+d];
793
794 memcpy(INT__p_h_k_d+q*13608, buffer, 13608*sizeof(double));
795 }
796
797 return ret;
798 }
799
ostei_h_p_d_k(struct simint_multi_shellpair const P,struct simint_multi_shellpair const Q,double screen_tol,double * const restrict work,double * const restrict INT__h_p_d_k)800 int ostei_h_p_d_k(struct simint_multi_shellpair const P,
801 struct simint_multi_shellpair const Q,
802 double screen_tol,
803 double * const restrict work,
804 double * const restrict INT__h_p_d_k)
805 {
806 double Q_AB[3*Q.nshell12];
807 struct simint_multi_shellpair Q_tmp = Q;
808 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
809 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
810 Q_tmp.AB_x = Q_AB;
811 Q_tmp.AB_y = Q_AB + Q.nshell12;
812 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
813
814 for(int i = 0; i < Q.nshell12; i++)
815 {
816 Q_tmp.AB_x[i] = -Q.AB_x[i];
817 Q_tmp.AB_y[i] = -Q.AB_y[i];
818 Q_tmp.AB_z[i] = -Q.AB_z[i];
819 }
820
821 int ret = ostei_h_p_k_d(P, Q_tmp, screen_tol, work, INT__h_p_d_k);
822 double buffer[13608] SIMINT_ALIGN_ARRAY_DBL;
823
824 for(int q = 0; q < ret; q++)
825 {
826 int idx = 0;
827 for(int a = 0; a < 21; ++a)
828 for(int b = 0; b < 3; ++b)
829 for(int c = 0; c < 6; ++c)
830 for(int d = 0; d < 36; ++d)
831 buffer[idx++] = INT__h_p_d_k[q*13608+a*648+b*216+d*6+c];
832
833 memcpy(INT__h_p_d_k+q*13608, buffer, 13608*sizeof(double));
834 }
835
836 return ret;
837 }
838
ostei_p_h_d_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_h_d_k)839 int ostei_p_h_d_k(struct simint_multi_shellpair const P,
840 struct simint_multi_shellpair const Q,
841 double screen_tol,
842 double * const restrict work,
843 double * const restrict INT__p_h_d_k)
844 {
845 double P_AB[3*P.nshell12];
846 struct simint_multi_shellpair P_tmp = P;
847 P_tmp.PA_x = P.PB_x; P_tmp.PA_y = P.PB_y; P_tmp.PA_z = P.PB_z;
848 P_tmp.PB_x = P.PA_x; P_tmp.PB_y = P.PA_y; P_tmp.PB_z = P.PA_z;
849 P_tmp.AB_x = P_AB;
850 P_tmp.AB_y = P_AB + P.nshell12;
851 P_tmp.AB_z = P_AB + 2*P.nshell12;
852
853 for(int i = 0; i < P.nshell12; i++)
854 {
855 P_tmp.AB_x[i] = -P.AB_x[i];
856 P_tmp.AB_y[i] = -P.AB_y[i];
857 P_tmp.AB_z[i] = -P.AB_z[i];
858 }
859
860 double Q_AB[3*Q.nshell12];
861 struct simint_multi_shellpair Q_tmp = Q;
862 Q_tmp.PA_x = Q.PB_x; Q_tmp.PA_y = Q.PB_y; Q_tmp.PA_z = Q.PB_z;
863 Q_tmp.PB_x = Q.PA_x; Q_tmp.PB_y = Q.PA_y; Q_tmp.PB_z = Q.PA_z;
864 Q_tmp.AB_x = Q_AB;
865 Q_tmp.AB_y = Q_AB + Q.nshell12;
866 Q_tmp.AB_z = Q_AB + 2*Q.nshell12;
867
868 for(int i = 0; i < Q.nshell12; i++)
869 {
870 Q_tmp.AB_x[i] = -Q.AB_x[i];
871 Q_tmp.AB_y[i] = -Q.AB_y[i];
872 Q_tmp.AB_z[i] = -Q.AB_z[i];
873 }
874
875 int ret = ostei_h_p_k_d(P_tmp, Q_tmp, screen_tol, work, INT__p_h_d_k);
876 double buffer[13608] SIMINT_ALIGN_ARRAY_DBL;
877
878 for(int q = 0; q < ret; q++)
879 {
880 int idx = 0;
881 for(int a = 0; a < 3; ++a)
882 for(int b = 0; b < 21; ++b)
883 for(int c = 0; c < 6; ++c)
884 for(int d = 0; d < 36; ++d)
885 buffer[idx++] = INT__p_h_d_k[q*13608+b*648+a*216+d*6+c];
886
887 memcpy(INT__p_h_d_k+q*13608, buffer, 13608*sizeof(double));
888 }
889
890 return ret;
891 }
892
893