/src/fftw3/rdft/scalar/r2cf/r2cfII_9.c
Line | Count | Source |
1 | | /* |
2 | | * Copyright (c) 2003, 2007-14 Matteo Frigo |
3 | | * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology |
4 | | * |
5 | | * This program is free software; you can redistribute it and/or modify |
6 | | * it under the terms of the GNU General Public License as published by |
7 | | * the Free Software Foundation; either version 2 of the License, or |
8 | | * (at your option) any later version. |
9 | | * |
10 | | * This program is distributed in the hope that it will be useful, |
11 | | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
12 | | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
13 | | * GNU General Public License for more details. |
14 | | * |
15 | | * You should have received a copy of the GNU General Public License |
16 | | * along with this program; if not, write to the Free Software |
17 | | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
18 | | * |
19 | | */ |
20 | | |
21 | | /* This file was automatically generated --- DO NOT EDIT */ |
22 | | /* Generated on Mon Jun 8 06:21:33 UTC 2026 */ |
23 | | |
24 | | #include "rdft/codelet-rdft.h" |
25 | | |
26 | | #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA) |
27 | | |
28 | | /* Generated by: ../../../genfft/gen_r2cf.native -fma -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cfII_9 -dft-II -include rdft/scalar/r2cfII.h */ |
29 | | |
30 | | /* |
31 | | * This function contains 42 FP additions, 34 FP multiplications, |
32 | | * (or, 12 additions, 4 multiplications, 30 fused multiply/add), |
33 | | * 48 stack variables, 17 constants, and 18 memory accesses |
34 | | */ |
35 | | #include "rdft/scalar/r2cfII.h" |
36 | | |
37 | | static void r2cfII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) |
38 | | { |
39 | | DK(KP852868531, +0.852868531952443209628250963940074071936020296); |
40 | | DK(KP879385241, +0.879385241571816768108218554649462939872416269); |
41 | | DK(KP984807753, +0.984807753012208059366743024589523013670643252); |
42 | | DK(KP898197570, +0.898197570222573798468955502359086394667167570); |
43 | | DK(KP673648177, +0.673648177666930348851716626769314796000375677); |
44 | | DK(KP939692620, +0.939692620785908384054109277324731469936208134); |
45 | | DK(KP907603734, +0.907603734547952313649323976213898122064543220); |
46 | | DK(KP666666666, +0.666666666666666666666666666666666666666666667); |
47 | | DK(KP826351822, +0.826351822333069651148283373230685203999624323); |
48 | | DK(KP866025403, +0.866025403784438646763723170752936183471402627); |
49 | | DK(KP315207469, +0.315207469095904627298647952427796244129086440); |
50 | | DK(KP420276625, +0.420276625461206169731530603237061658838781920); |
51 | | DK(KP203604859, +0.203604859554852403062088995281827210665664861); |
52 | | DK(KP152703644, +0.152703644666139302296566746461370407999248646); |
53 | | DK(KP726681596, +0.726681596905677465811651808188092531873167623); |
54 | | DK(KP968908795, +0.968908795874236621082202410917456709164223497); |
55 | | DK(KP500000000, +0.500000000000000000000000000000000000000000000); |
56 | | { |
57 | | INT i; |
58 | | for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { |
59 | | E T1, T4, To, Ta, Tm, TB, Tq, Tt, Tf, Tj, TA, Tr, Ts, T2, T3; |
60 | | E T5, Tg; |
61 | | T1 = R0[0]; |
62 | | T2 = R0[WS(rs, 3)]; |
63 | | T3 = R1[WS(rs, 1)]; |
64 | | T4 = T2 - T3; |
65 | | To = T2 + T3; |
66 | | { |
67 | | E T6, T9, Tk, T7, T8, Tl; |
68 | | T6 = R0[WS(rs, 1)]; |
69 | | T7 = R0[WS(rs, 4)]; |
70 | | T8 = R1[WS(rs, 2)]; |
71 | | T9 = T7 - T8; |
72 | | Tk = T7 + T8; |
73 | | Ta = T6 + T9; |
74 | | Tl = FNMS(KP500000000, T9, T6); |
75 | | Tm = FMA(KP968908795, Tl, Tk); |
76 | | TB = FNMS(KP726681596, Tk, Tl); |
77 | | Tq = FNMS(KP152703644, Tk, Tl); |
78 | | Tt = FMA(KP203604859, Tl, Tk); |
79 | | } |
80 | | { |
81 | | E Tb, Te, Ti, Tc, Td, Th; |
82 | | Tb = R0[WS(rs, 2)]; |
83 | | Tc = R1[0]; |
84 | | Td = R1[WS(rs, 3)]; |
85 | | Te = Tc + Td; |
86 | | Ti = Tc - Td; |
87 | | Tf = Tb - Te; |
88 | | Th = FMA(KP500000000, Te, Tb); |
89 | | Tj = FNMS(KP152703644, Ti, Th); |
90 | | TA = FMA(KP203604859, Th, Ti); |
91 | | Tr = FNMS(KP420276625, Th, Ti); |
92 | | Ts = FMA(KP315207469, Ti, Th); |
93 | | } |
94 | | Ci[WS(csi, 1)] = KP866025403 * (Tf - Ta); |
95 | | T5 = T1 + T4; |
96 | | Tg = Ta + Tf; |
97 | | Cr[WS(csr, 1)] = FNMS(KP500000000, Tg, T5); |
98 | | Cr[WS(csr, 4)] = T5 + Tg; |
99 | | { |
100 | | E Ty, Tx, Tz, Tn, TD, TC; |
101 | | Tx = FNMS(KP826351822, Tr, Tq); |
102 | | Ty = FNMS(KP666666666, Tx, Tt); |
103 | | Tz = FMA(KP907603734, Ty, Ts); |
104 | | Ci[WS(csi, 2)] = KP866025403 * (FNMS(KP939692620, Tz, To)); |
105 | | Tn = FMA(KP673648177, Tm, Tj); |
106 | | TC = FNMS(KP898197570, TB, TA); |
107 | | TD = FNMS(KP666666666, Tn, TC); |
108 | | Ci[0] = -(KP984807753 * (FMA(KP879385241, To, Tn))); |
109 | | Ci[WS(csi, 3)] = -(KP866025403 * (FMA(KP852868531, TD, To))); |
110 | | { |
111 | | E Tp, Tv, TF, TG, Tu, TE, Tw; |
112 | | Tp = FNMS(KP500000000, T4, T1); |
113 | | Tu = FNMS(KP907603734, Tt, Ts); |
114 | | Tv = FNMS(KP666666666, Tu, Tr); |
115 | | TE = FNMS(KP673648177, Tm, Tj); |
116 | | TF = FMA(KP898197570, TB, TA); |
117 | | TG = FMA(KP500000000, TF, TE); |
118 | | Cr[WS(csr, 3)] = FNMS(KP852868531, TG, Tp); |
119 | | Cr[0] = FMA(KP852868531, TF, Tp); |
120 | | Tw = FMA(KP826351822, Tv, Tq); |
121 | | Cr[WS(csr, 2)] = FNMS(KP852868531, Tw, Tp); |
122 | | } |
123 | | } |
124 | | } |
125 | | } |
126 | | } |
127 | | |
128 | | static const kr2c_desc desc = { 9, "r2cfII_9", { 12, 4, 30, 0 }, &GENUS }; |
129 | | |
130 | | void X(codelet_r2cfII_9) (planner *p) { X(kr2c_register) (p, r2cfII_9, &desc); |
131 | | } |
132 | | |
133 | | #else |
134 | | |
135 | | /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cfII_9 -dft-II -include rdft/scalar/r2cfII.h */ |
136 | | |
137 | | /* |
138 | | * This function contains 42 FP additions, 30 FP multiplications, |
139 | | * (or, 25 additions, 13 multiplications, 17 fused multiply/add), |
140 | | * 39 stack variables, 14 constants, and 18 memory accesses |
141 | | */ |
142 | | #include "rdft/scalar/r2cfII.h" |
143 | | |
144 | | static void r2cfII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) |
145 | 0 | { |
146 | 0 | DK(KP663413948, +0.663413948168938396205421319635891297216863310); |
147 | 0 | DK(KP642787609, +0.642787609686539326322643409907263432907559884); |
148 | 0 | DK(KP556670399, +0.556670399226419366452912952047023132968291906); |
149 | 0 | DK(KP766044443, +0.766044443118978035202392650555416673935832457); |
150 | 0 | DK(KP852868531, +0.852868531952443209628250963940074071936020296); |
151 | 0 | DK(KP173648177, +0.173648177666930348851716626769314796000375677); |
152 | 0 | DK(KP984807753, +0.984807753012208059366743024589523013670643252); |
153 | 0 | DK(KP150383733, +0.150383733180435296639271897612501926072238258); |
154 | 0 | DK(KP813797681, +0.813797681349373692844693217248393223289101568); |
155 | 0 | DK(KP342020143, +0.342020143325668733044099614682259580763083368); |
156 | 0 | DK(KP939692620, +0.939692620785908384054109277324731469936208134); |
157 | 0 | DK(KP296198132, +0.296198132726023843175338011893050938967728390); |
158 | 0 | DK(KP866025403, +0.866025403784438646763723170752936183471402627); |
159 | 0 | DK(KP500000000, +0.500000000000000000000000000000000000000000000); |
160 | 0 | { |
161 | 0 | INT i; |
162 | 0 | for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { |
163 | 0 | E T1, T4, To, Ta, Tl, Tk, Tf, Ti, Th, T2, T3, T5, Tg; |
164 | 0 | T1 = R0[0]; |
165 | 0 | T2 = R1[WS(rs, 1)]; |
166 | 0 | T3 = R0[WS(rs, 3)]; |
167 | 0 | T4 = T2 - T3; |
168 | 0 | To = T2 + T3; |
169 | 0 | { |
170 | 0 | E T6, T7, T8, T9; |
171 | 0 | T6 = R0[WS(rs, 1)]; |
172 | 0 | T7 = R1[WS(rs, 2)]; |
173 | 0 | T8 = R0[WS(rs, 4)]; |
174 | 0 | T9 = T7 - T8; |
175 | 0 | Ta = T6 - T9; |
176 | 0 | Tl = T7 + T8; |
177 | 0 | Tk = FMA(KP500000000, T9, T6); |
178 | 0 | } |
179 | 0 | { |
180 | 0 | E Tb, Tc, Td, Te; |
181 | 0 | Tb = R0[WS(rs, 2)]; |
182 | 0 | Tc = R1[0]; |
183 | 0 | Td = R1[WS(rs, 3)]; |
184 | 0 | Te = Tc + Td; |
185 | 0 | Tf = Tb - Te; |
186 | 0 | Ti = FMA(KP500000000, Te, Tb); |
187 | 0 | Th = Tc - Td; |
188 | 0 | } |
189 | 0 | Ci[WS(csi, 1)] = KP866025403 * (Tf - Ta); |
190 | 0 | T5 = T1 - T4; |
191 | 0 | Tg = Ta + Tf; |
192 | 0 | Cr[WS(csr, 1)] = FNMS(KP500000000, Tg, T5); |
193 | 0 | Cr[WS(csr, 4)] = T5 + Tg; |
194 | 0 | { |
195 | 0 | E Tr, Tt, Tw, Tv, Tu, Tp, Tq, Ts, Tj, Tm, Tn; |
196 | 0 | Tr = FMA(KP500000000, T4, T1); |
197 | 0 | Tt = FMA(KP296198132, Th, KP939692620 * Ti); |
198 | 0 | Tw = FNMS(KP813797681, Th, KP342020143 * Ti); |
199 | 0 | Tv = FNMS(KP984807753, Tk, KP150383733 * Tl); |
200 | 0 | Tu = FMA(KP173648177, Tk, KP852868531 * Tl); |
201 | 0 | Tp = FNMS(KP556670399, Tl, KP766044443 * Tk); |
202 | 0 | Tq = FMA(KP852868531, Th, KP173648177 * Ti); |
203 | 0 | Ts = Tp + Tq; |
204 | 0 | Tj = FNMS(KP984807753, Ti, KP150383733 * Th); |
205 | 0 | Tm = FMA(KP642787609, Tk, KP663413948 * Tl); |
206 | 0 | Tn = Tj - Tm; |
207 | 0 | Ci[0] = FNMS(KP866025403, To, Tn); |
208 | 0 | Cr[0] = Tr + Ts; |
209 | 0 | Ci[WS(csi, 3)] = FNMS(KP500000000, Tn, KP866025403 * ((Tp - Tq) - To)); |
210 | 0 | Cr[WS(csr, 3)] = FMA(KP866025403, Tm + Tj, Tr) - (KP500000000 * Ts); |
211 | 0 | Ci[WS(csi, 2)] = FMA(KP866025403, To - (Tu + Tt), KP500000000 * (Tw - Tv)); |
212 | 0 | Cr[WS(csr, 2)] = FMA(KP500000000, Tt - Tu, Tr) + (KP866025403 * (Tv + Tw)); |
213 | 0 | } |
214 | 0 | } |
215 | 0 | } |
216 | 0 | } |
217 | | |
218 | | static const kr2c_desc desc = { 9, "r2cfII_9", { 25, 13, 17, 0 }, &GENUS }; |
219 | | |
220 | 1 | void X(codelet_r2cfII_9) (planner *p) { X(kr2c_register) (p, r2cfII_9, &desc); |
221 | 1 | } |
222 | | |
223 | | #endif |