cannam@167: /* cannam@167: * Copyright (c) 2003, 2007-14 Matteo Frigo cannam@167: * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology cannam@167: * cannam@167: * This program is free software; you can redistribute it and/or modify cannam@167: * it under the terms of the GNU General Public License as published by cannam@167: * the Free Software Foundation; either version 2 of the License, or cannam@167: * (at your option) any later version. cannam@167: * cannam@167: * This program is distributed in the hope that it will be useful, cannam@167: * but WITHOUT ANY WARRANTY; without even the implied warranty of cannam@167: * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the cannam@167: * GNU General Public License for more details. cannam@167: * cannam@167: * You should have received a copy of the GNU General Public License cannam@167: * along with this program; if not, write to the Free Software cannam@167: * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA cannam@167: * cannam@167: */ cannam@167: cannam@167: /* This file was automatically generated --- DO NOT EDIT */ cannam@167: /* Generated on Thu May 24 08:07:28 EDT 2018 */ cannam@167: cannam@167: #include "rdft/codelet-rdft.h" cannam@167: cannam@167: #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA) cannam@167: cannam@167: /* Generated by: ../../../genfft/gen_r2cb.native -fma -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include rdft/scalar/r2cb.h */ cannam@167: cannam@167: /* cannam@167: * This function contains 12 FP additions, 10 FP multiplications, cannam@167: * (or, 2 additions, 0 multiplications, 10 fused multiply/add), cannam@167: * 18 stack variables, 5 constants, and 10 memory accesses cannam@167: */ cannam@167: #include "rdft/scalar/r2cb.h" cannam@167: cannam@167: static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) cannam@167: { cannam@167: DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); cannam@167: DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); cannam@167: DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); cannam@167: DK(KP500000000, +0.500000000000000000000000000000000000000000000); cannam@167: DK(KP618033988, +0.618033988749894848204586834365638117720309180); cannam@167: { cannam@167: INT i; cannam@167: 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)) { cannam@167: E Ta, Tc, T1, T4, T5, T6, Tb, T7; cannam@167: { cannam@167: E T8, T9, T2, T3; cannam@167: T8 = Ci[WS(csi, 1)]; cannam@167: T9 = Ci[WS(csi, 2)]; cannam@167: Ta = FMA(KP618033988, T9, T8); cannam@167: Tc = FMS(KP618033988, T8, T9); cannam@167: T1 = Cr[0]; cannam@167: T2 = Cr[WS(csr, 1)]; cannam@167: T3 = Cr[WS(csr, 2)]; cannam@167: T4 = T2 + T3; cannam@167: T5 = FNMS(KP500000000, T4, T1); cannam@167: T6 = T2 - T3; cannam@167: } cannam@167: R0[0] = FMA(KP2_000000000, T4, T1); cannam@167: Tb = FNMS(KP1_118033988, T6, T5); cannam@167: R0[WS(rs, 1)] = FNMS(KP1_902113032, Tc, Tb); cannam@167: R1[WS(rs, 1)] = FMA(KP1_902113032, Tc, Tb); cannam@167: T7 = FMA(KP1_118033988, T6, T5); cannam@167: R1[0] = FNMS(KP1_902113032, Ta, T7); cannam@167: R0[WS(rs, 2)] = FMA(KP1_902113032, Ta, T7); cannam@167: } cannam@167: } cannam@167: } cannam@167: cannam@167: static const kr2c_desc desc = { 5, "r2cb_5", {2, 0, 10, 0}, &GENUS }; cannam@167: cannam@167: void X(codelet_r2cb_5) (planner *p) { cannam@167: X(kr2c_register) (p, r2cb_5, &desc); cannam@167: } cannam@167: cannam@167: #else cannam@167: cannam@167: /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include rdft/scalar/r2cb.h */ cannam@167: cannam@167: /* cannam@167: * This function contains 12 FP additions, 7 FP multiplications, cannam@167: * (or, 8 additions, 3 multiplications, 4 fused multiply/add), cannam@167: * 18 stack variables, 5 constants, and 10 memory accesses cannam@167: */ cannam@167: #include "rdft/scalar/r2cb.h" cannam@167: cannam@167: static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) cannam@167: { cannam@167: DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); cannam@167: DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); cannam@167: DK(KP500000000, +0.500000000000000000000000000000000000000000000); cannam@167: DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); cannam@167: DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); cannam@167: { cannam@167: INT i; cannam@167: 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)) { cannam@167: E Ta, Tc, T1, T4, T5, T6, Tb, T7; cannam@167: { cannam@167: E T8, T9, T2, T3; cannam@167: T8 = Ci[WS(csi, 1)]; cannam@167: T9 = Ci[WS(csi, 2)]; cannam@167: Ta = FNMS(KP1_902113032, T9, KP1_175570504 * T8); cannam@167: Tc = FMA(KP1_902113032, T8, KP1_175570504 * T9); cannam@167: T1 = Cr[0]; cannam@167: T2 = Cr[WS(csr, 1)]; cannam@167: T3 = Cr[WS(csr, 2)]; cannam@167: T4 = T2 + T3; cannam@167: T5 = FNMS(KP500000000, T4, T1); cannam@167: T6 = KP1_118033988 * (T2 - T3); cannam@167: } cannam@167: R0[0] = FMA(KP2_000000000, T4, T1); cannam@167: Tb = T6 + T5; cannam@167: R1[0] = Tb - Tc; cannam@167: R0[WS(rs, 2)] = Tb + Tc; cannam@167: T7 = T5 - T6; cannam@167: R0[WS(rs, 1)] = T7 - Ta; cannam@167: R1[WS(rs, 1)] = T7 + Ta; cannam@167: } cannam@167: } cannam@167: } cannam@167: cannam@167: static const kr2c_desc desc = { 5, "r2cb_5", {8, 3, 4, 0}, &GENUS }; cannam@167: cannam@167: void X(codelet_r2cb_5) (planner *p) { cannam@167: X(kr2c_register) (p, r2cb_5, &desc); cannam@167: } cannam@167: cannam@167: #endif