/* * Copyright (c) 2003, 2007-14 Matteo Frigo * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Thu May 24 08:07:43 EDT 2018 */ #include "rdft/codelet-rdft.h" #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA) /* Generated by: ../../../genfft/gen_r2cb.native -fma -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cbIII_9 -dft-III -include rdft/scalar/r2cbIII.h */ /* * This function contains 32 FP additions, 24 FP multiplications, * (or, 8 additions, 0 multiplications, 24 fused multiply/add), * 35 stack variables, 12 constants, and 18 memory accesses */ #include "rdft/scalar/r2cbIII.h" static void r2cbIII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP1_969615506, +1.969615506024416118733486049179046027341286503); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_532088886, +1.532088886237956070404785301110833347871664914); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T3, Tr, Th, Td, Tc, T8, Tn, Ts, Tk, Tt, T9, Te; { E Tg, T1, T2, Tf; Tg = Ci[WS(csi, 1)]; T1 = Cr[WS(csr, 4)]; T2 = Cr[WS(csr, 1)]; Tf = T2 - T1; T3 = FMA(KP2_000000000, T2, T1); Tr = FMA(KP1_732050807, Tg, Tf); Th = FNMS(KP1_732050807, Tg, Tf); } { E T4, T7, Tm, Tj, Tl, Ti; T4 = Cr[WS(csr, 3)]; Td = Ci[WS(csi, 3)]; { E T5, T6, Ta, Tb; T5 = Cr[0]; T6 = Cr[WS(csr, 2)]; T7 = T5 + T6; Tm = T5 - T6; Ta = Ci[WS(csi, 2)]; Tb = Ci[0]; Tc = Ta - Tb; Tj = Tb + Ta; } T8 = T4 + T7; Tl = FMA(KP500000000, Tc, Td); Tn = FNMS(KP866025403, Tm, Tl); Ts = FMA(KP866025403, Tm, Tl); Ti = FNMS(KP500000000, T7, T4); Tk = FMA(KP866025403, Tj, Ti); Tt = FNMS(KP866025403, Tj, Ti); } R0[0] = FMA(KP2_000000000, T8, T3); T9 = T8 - T3; Te = Tc - Td; R1[WS(rs, 1)] = FMA(KP1_732050807, Te, T9); R0[WS(rs, 3)] = FMS(KP1_732050807, Te, T9); { E Tq, To, Tp, Tw, Tu, Tv; Tq = FNMS(KP839099631, Tk, Tn); To = FMA(KP839099631, Tn, Tk); Tp = FMA(KP766044443, To, Th); R1[0] = FNMS(KP1_532088886, To, Th); R1[WS(rs, 3)] = FMA(KP1_326827896, Tq, Tp); R0[WS(rs, 2)] = FMS(KP1_326827896, Tq, Tp); Tw = FNMS(KP176326980, Ts, Tt); Tu = FMA(KP176326980, Tt, Ts); Tv = FMA(KP984807753, Tu, Tr); R0[WS(rs, 1)] = FMS(KP1_969615506, Tu, Tr); R1[WS(rs, 2)] = FMA(KP1_705737063, Tw, Tv); R0[WS(rs, 4)] = FMS(KP1_705737063, Tw, Tv); } } } } static const kr2c_desc desc = { 9, "r2cbIII_9", {8, 0, 24, 0}, &GENUS }; void X(codelet_r2cbIII_9) (planner *p) { X(kr2c_register) (p, r2cbIII_9, &desc); } #else /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cbIII_9 -dft-III -include rdft/scalar/r2cbIII.h */ /* * This function contains 32 FP additions, 18 FP multiplications, * (or, 22 additions, 8 multiplications, 10 fused multiply/add), * 35 stack variables, 12 constants, and 18 memory accesses */ #include "rdft/scalar/r2cbIII.h" static void r2cbIII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_113340798, +1.113340798452838732905825904094046265936583811); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP300767466, +0.300767466360870593278543795225003852144476517); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T3, Ts, Ti, Td, Tc, T8, To, Tu, Tl, Tt, T9, Te; { E Th, T1, T2, Tf, Tg; Tg = Ci[WS(csi, 1)]; Th = KP1_732050807 * Tg; T1 = Cr[WS(csr, 4)]; T2 = Cr[WS(csr, 1)]; Tf = T2 - T1; T3 = FMA(KP2_000000000, T2, T1); Ts = Tf - Th; Ti = Tf + Th; } { E T4, T7, Tm, Tk, Tn, Tj; T4 = Cr[WS(csr, 3)]; Td = Ci[WS(csi, 3)]; { E T5, T6, Ta, Tb; T5 = Cr[0]; T6 = Cr[WS(csr, 2)]; T7 = T5 + T6; Tm = KP866025403 * (T6 - T5); Ta = Ci[WS(csi, 2)]; Tb = Ci[0]; Tc = Ta - Tb; Tk = KP866025403 * (Tb + Ta); } T8 = T4 + T7; Tn = FMA(KP500000000, Tc, Td); To = Tm - Tn; Tu = Tm + Tn; Tj = FMS(KP500000000, T7, T4); Tl = Tj + Tk; Tt = Tj - Tk; } R0[0] = FMA(KP2_000000000, T8, T3); T9 = T8 - T3; Te = KP1_732050807 * (Tc - Td); R1[WS(rs, 1)] = T9 + Te; R0[WS(rs, 3)] = Te - T9; { E Tr, Tp, Tq, Tx, Tv, Tw; Tr = FNMS(KP1_705737063, Tl, KP300767466 * To); Tp = FMA(KP173648177, Tl, KP984807753 * To); Tq = Ti - Tp; R0[WS(rs, 1)] = -(FMA(KP2_000000000, Tp, Ti)); R0[WS(rs, 4)] = Tr - Tq; R1[WS(rs, 2)] = Tq + Tr; Tx = FMA(KP1_113340798, Tt, KP1_326827896 * Tu); Tv = FNMS(KP642787609, Tu, KP766044443 * Tt); Tw = Tv - Ts; R1[0] = FMA(KP2_000000000, Tv, Ts); R1[WS(rs, 3)] = Tx - Tw; R0[WS(rs, 2)] = Tw + Tx; } } } } static const kr2c_desc desc = { 9, "r2cbIII_9", {22, 8, 10, 0}, &GENUS }; void X(codelet_r2cbIII_9) (planner *p) { X(kr2c_register) (p, r2cbIII_9, &desc); } #endif