Mercurial > hg > sv-dependency-builds
comparison src/fftw-3.3.3/rdft/scalar/r2cb/r2cb_5.c @ 10:37bf6b4a2645
Add FFTW3
| author | Chris Cannam |
|---|---|
| date | Wed, 20 Mar 2013 15:35:50 +0000 |
| parents | |
| children |
comparison
equal
deleted
inserted
replaced
| 9:c0fb53affa76 | 10:37bf6b4a2645 |
|---|---|
| 1 /* | |
| 2 * Copyright (c) 2003, 2007-11 Matteo Frigo | |
| 3 * Copyright (c) 2003, 2007-11 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 Sun Nov 25 07:41:07 EST 2012 */ | |
| 23 | |
| 24 #include "codelet-rdft.h" | |
| 25 | |
| 26 #ifdef HAVE_FMA | |
| 27 | |
| 28 /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include r2cb.h */ | |
| 29 | |
| 30 /* | |
| 31 * This function contains 12 FP additions, 10 FP multiplications, | |
| 32 * (or, 2 additions, 0 multiplications, 10 fused multiply/add), | |
| 33 * 18 stack variables, 5 constants, and 10 memory accesses | |
| 34 */ | |
| 35 #include "r2cb.h" | |
| 36 | |
| 37 static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) | |
| 38 { | |
| 39 DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); | |
| 40 DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); | |
| 41 DK(KP500000000, +0.500000000000000000000000000000000000000000000); | |
| 42 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); | |
| 43 DK(KP618033988, +0.618033988749894848204586834365638117720309180); | |
| 44 { | |
| 45 INT i; | |
| 46 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { | |
| 47 E T1, T2, T3, Tc, Ta, T8, T9; | |
| 48 T8 = Ci[WS(csi, 1)]; | |
| 49 T9 = Ci[WS(csi, 2)]; | |
| 50 T1 = Cr[0]; | |
| 51 T2 = Cr[WS(csr, 1)]; | |
| 52 T3 = Cr[WS(csr, 2)]; | |
| 53 Tc = FMS(KP618033988, T8, T9); | |
| 54 Ta = FMA(KP618033988, T9, T8); | |
| 55 { | |
| 56 E T6, T4, T5, T7, Tb; | |
| 57 T6 = T2 - T3; | |
| 58 T4 = T2 + T3; | |
| 59 R0[0] = FMA(KP2_000000000, T4, T1); | |
| 60 T5 = FNMS(KP500000000, T4, T1); | |
| 61 T7 = FMA(KP1_118033988, T6, T5); | |
| 62 Tb = FNMS(KP1_118033988, T6, T5); | |
| 63 R0[WS(rs, 2)] = FMA(KP1_902113032, Ta, T7); | |
| 64 R1[0] = FNMS(KP1_902113032, Ta, T7); | |
| 65 R1[WS(rs, 1)] = FMA(KP1_902113032, Tc, Tb); | |
| 66 R0[WS(rs, 1)] = FNMS(KP1_902113032, Tc, Tb); | |
| 67 } | |
| 68 } | |
| 69 } | |
| 70 } | |
| 71 | |
| 72 static const kr2c_desc desc = { 5, "r2cb_5", {2, 0, 10, 0}, &GENUS }; | |
| 73 | |
| 74 void X(codelet_r2cb_5) (planner *p) { | |
| 75 X(kr2c_register) (p, r2cb_5, &desc); | |
| 76 } | |
| 77 | |
| 78 #else /* HAVE_FMA */ | |
| 79 | |
| 80 /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include r2cb.h */ | |
| 81 | |
| 82 /* | |
| 83 * This function contains 12 FP additions, 7 FP multiplications, | |
| 84 * (or, 8 additions, 3 multiplications, 4 fused multiply/add), | |
| 85 * 18 stack variables, 5 constants, and 10 memory accesses | |
| 86 */ | |
| 87 #include "r2cb.h" | |
| 88 | |
| 89 static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) | |
| 90 { | |
| 91 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); | |
| 92 DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); | |
| 93 DK(KP500000000, +0.500000000000000000000000000000000000000000000); | |
| 94 DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); | |
| 95 DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); | |
| 96 { | |
| 97 INT i; | |
| 98 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { | |
| 99 E Ta, Tc, T1, T4, T5, T6, Tb, T7; | |
| 100 { | |
| 101 E T8, T9, T2, T3; | |
| 102 T8 = Ci[WS(csi, 1)]; | |
| 103 T9 = Ci[WS(csi, 2)]; | |
| 104 Ta = FNMS(KP1_902113032, T9, KP1_175570504 * T8); | |
| 105 Tc = FMA(KP1_902113032, T8, KP1_175570504 * T9); | |
| 106 T1 = Cr[0]; | |
| 107 T2 = Cr[WS(csr, 1)]; | |
| 108 T3 = Cr[WS(csr, 2)]; | |
| 109 T4 = T2 + T3; | |
| 110 T5 = FNMS(KP500000000, T4, T1); | |
| 111 T6 = KP1_118033988 * (T2 - T3); | |
| 112 } | |
| 113 R0[0] = FMA(KP2_000000000, T4, T1); | |
| 114 Tb = T6 + T5; | |
| 115 R1[0] = Tb - Tc; | |
| 116 R0[WS(rs, 2)] = Tb + Tc; | |
| 117 T7 = T5 - T6; | |
| 118 R0[WS(rs, 1)] = T7 - Ta; | |
| 119 R1[WS(rs, 1)] = T7 + Ta; | |
| 120 } | |
| 121 } | |
| 122 } | |
| 123 | |
| 124 static const kr2c_desc desc = { 5, "r2cb_5", {8, 3, 4, 0}, &GENUS }; | |
| 125 | |
| 126 void X(codelet_r2cb_5) (planner *p) { | |
| 127 X(kr2c_register) (p, r2cb_5, &desc); | |
| 128 } | |
| 129 | |
| 130 #endif /* HAVE_FMA */ |