Mercurial > hg > sv-dependency-builds
comparison src/fftw-3.3.8/rdft/scalar/r2cb/r2cb_6.c @ 167:bd3cc4d1df30
Add FFTW 3.3.8 source, and a Linux build
| author | Chris Cannam <cannam@all-day-breakfast.com> |
|---|---|
| date | Tue, 19 Nov 2019 14:52:55 +0000 |
| parents | |
| children |
comparison
equal
deleted
inserted
replaced
| 166:cbd6d7e562c7 | 167:bd3cc4d1df30 |
|---|---|
| 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 Thu May 24 08:07:28 EDT 2018 */ | |
| 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_r2cb.native -fma -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include rdft/scalar/r2cb.h */ | |
| 29 | |
| 30 /* | |
| 31 * This function contains 14 FP additions, 6 FP multiplications, | |
| 32 * (or, 8 additions, 0 multiplications, 6 fused multiply/add), | |
| 33 * 17 stack variables, 2 constants, and 12 memory accesses | |
| 34 */ | |
| 35 #include "rdft/scalar/r2cb.h" | |
| 36 | |
| 37 static void r2cb_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) | |
| 38 { | |
| 39 DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); | |
| 40 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); | |
| 41 { | |
| 42 INT i; | |
| 43 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { | |
| 44 E T3, T7, Tc, Te, T6, T8, T1, T2, T9, Td; | |
| 45 T1 = Cr[0]; | |
| 46 T2 = Cr[WS(csr, 3)]; | |
| 47 T3 = T1 + T2; | |
| 48 T7 = T1 - T2; | |
| 49 { | |
| 50 E Ta, Tb, T4, T5; | |
| 51 Ta = Ci[WS(csi, 2)]; | |
| 52 Tb = Ci[WS(csi, 1)]; | |
| 53 Tc = Ta - Tb; | |
| 54 Te = Ta + Tb; | |
| 55 T4 = Cr[WS(csr, 2)]; | |
| 56 T5 = Cr[WS(csr, 1)]; | |
| 57 T6 = T4 + T5; | |
| 58 T8 = T5 - T4; | |
| 59 } | |
| 60 R0[0] = FMA(KP2_000000000, T6, T3); | |
| 61 R1[WS(rs, 1)] = FNMS(KP2_000000000, T8, T7); | |
| 62 T9 = T3 - T6; | |
| 63 R0[WS(rs, 2)] = FNMS(KP1_732050807, Tc, T9); | |
| 64 R0[WS(rs, 1)] = FMA(KP1_732050807, Tc, T9); | |
| 65 Td = T7 + T8; | |
| 66 R1[0] = FNMS(KP1_732050807, Te, Td); | |
| 67 R1[WS(rs, 2)] = FMA(KP1_732050807, Te, Td); | |
| 68 } | |
| 69 } | |
| 70 } | |
| 71 | |
| 72 static const kr2c_desc desc = { 6, "r2cb_6", {8, 0, 6, 0}, &GENUS }; | |
| 73 | |
| 74 void X(codelet_r2cb_6) (planner *p) { | |
| 75 X(kr2c_register) (p, r2cb_6, &desc); | |
| 76 } | |
| 77 | |
| 78 #else | |
| 79 | |
| 80 /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include rdft/scalar/r2cb.h */ | |
| 81 | |
| 82 /* | |
| 83 * This function contains 14 FP additions, 4 FP multiplications, | |
| 84 * (or, 12 additions, 2 multiplications, 2 fused multiply/add), | |
| 85 * 17 stack variables, 2 constants, and 12 memory accesses | |
| 86 */ | |
| 87 #include "rdft/scalar/r2cb.h" | |
| 88 | |
| 89 static void r2cb_6(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_732050807, +1.732050807568877293527446341505872366942805254); | |
| 93 { | |
| 94 INT i; | |
| 95 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { | |
| 96 E T3, T7, Tc, Te, T6, T8, T1, T2, T9, Td; | |
| 97 T1 = Cr[0]; | |
| 98 T2 = Cr[WS(csr, 3)]; | |
| 99 T3 = T1 - T2; | |
| 100 T7 = T1 + T2; | |
| 101 { | |
| 102 E Ta, Tb, T4, T5; | |
| 103 Ta = Ci[WS(csi, 2)]; | |
| 104 Tb = Ci[WS(csi, 1)]; | |
| 105 Tc = KP1_732050807 * (Ta - Tb); | |
| 106 Te = KP1_732050807 * (Ta + Tb); | |
| 107 T4 = Cr[WS(csr, 2)]; | |
| 108 T5 = Cr[WS(csr, 1)]; | |
| 109 T6 = T4 - T5; | |
| 110 T8 = T4 + T5; | |
| 111 } | |
| 112 R1[WS(rs, 1)] = FMA(KP2_000000000, T6, T3); | |
| 113 R0[0] = FMA(KP2_000000000, T8, T7); | |
| 114 T9 = T7 - T8; | |
| 115 R0[WS(rs, 2)] = T9 - Tc; | |
| 116 R0[WS(rs, 1)] = T9 + Tc; | |
| 117 Td = T3 - T6; | |
| 118 R1[0] = Td - Te; | |
| 119 R1[WS(rs, 2)] = Td + Te; | |
| 120 } | |
| 121 } | |
| 122 } | |
| 123 | |
| 124 static const kr2c_desc desc = { 6, "r2cb_6", {12, 2, 2, 0}, &GENUS }; | |
| 125 | |
| 126 void X(codelet_r2cb_6) (planner *p) { | |
| 127 X(kr2c_register) (p, r2cb_6, &desc); | |
| 128 } | |
| 129 | |
| 130 #endif |