Chris@42: /* Chris@42: * Copyright (c) 2003, 2007-14 Matteo Frigo Chris@42: * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology Chris@42: * Chris@42: * This program is free software; you can redistribute it and/or modify Chris@42: * it under the terms of the GNU General Public License as published by Chris@42: * the Free Software Foundation; either version 2 of the License, or Chris@42: * (at your option) any later version. Chris@42: * Chris@42: * This program is distributed in the hope that it will be useful, Chris@42: * but WITHOUT ANY WARRANTY; without even the implied warranty of Chris@42: * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the Chris@42: * GNU General Public License for more details. Chris@42: * Chris@42: * You should have received a copy of the GNU General Public License Chris@42: * along with this program; if not, write to the Free Software Chris@42: * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA Chris@42: * Chris@42: */ Chris@42: Chris@42: /* This file was automatically generated --- DO NOT EDIT */ Chris@42: /* Generated on Sat Jul 30 16:36:47 EDT 2016 */ Chris@42: Chris@42: #include "codelet-dft.h" Chris@42: Chris@42: #ifdef HAVE_FMA Chris@42: Chris@42: /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 64 -name t2_64 -include t.h */ Chris@42: Chris@42: /* Chris@42: * This function contains 1154 FP additions, 840 FP multiplications, Chris@42: * (or, 520 additions, 206 multiplications, 634 fused multiply/add), Chris@42: * 349 stack variables, 15 constants, and 256 memory accesses Chris@42: */ Chris@42: #include "t.h" Chris@42: Chris@42: static void t2_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) Chris@42: { Chris@42: DK(KP995184726, +0.995184726672196886244836953109479921575474869); Chris@42: DK(KP773010453, +0.773010453362736960810906609758469800971041293); Chris@42: DK(KP956940335, +0.956940335732208864935797886980269969482849206); Chris@42: DK(KP881921264, +0.881921264348355029712756863660388349508442621); Chris@42: DK(KP820678790, +0.820678790828660330972281985331011598767386482); Chris@42: DK(KP098491403, +0.098491403357164253077197521291327432293052451); Chris@42: DK(KP534511135, +0.534511135950791641089685961295362908582039528); Chris@42: DK(KP303346683, +0.303346683607342391675883946941299872384187453); Chris@42: DK(KP831469612, +0.831469612302545237078788377617905756738560812); Chris@42: DK(KP980785280, +0.980785280403230449126182236134239036973933731); Chris@42: DK(KP668178637, +0.668178637919298919997757686523080761552472251); Chris@42: DK(KP198912367, +0.198912367379658006911597622644676228597850501); Chris@42: DK(KP923879532, +0.923879532511286756128183189396788286822416626); Chris@42: DK(KP707106781, +0.707106781186547524400844362104849039284835938); Chris@42: DK(KP414213562, +0.414213562373095048801688724209698078569671875); Chris@42: { Chris@42: INT m; Chris@42: for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(128, rs)) { Chris@42: E Tg0, TlC, TlB, Tg3; Chris@42: { Chris@42: E T2, T3, Tc, T8, Te, T5, T6, T14, T3d, T3i, TJ, T7, Tr, T3g, TG; Chris@42: E T10, T3a, TL, TP, Tb, Tt, T17, Td, Ti, T3N, T3R, T1i, Tu, T1I, T2U; Chris@42: E T1t, T3U, T5O, T48, T2u, T7B, TK, T79, T3D, T2h, T2l, T3G, T1x, T3X, T2d; Chris@42: E T1M, T2X, T4B, T4x, T3j, T4T, T29, T5s, T81, T5w, T7X, T7N, T7h, T64, T6a; Chris@42: E T6e, T7l, T60, T7R, T6h, T5A, T7o, T6J, T6k, T5E, T6N, T7r, T6x, T6t, T7c; Chris@42: E TO, T2x, T7E, TU, TQ, T2C, T2y, T5R, T4b, T4c, T4g, T4W, T3m, T3r, T3n; Chris@42: E T1k, Tx, Ty, T4p, T4s, TC, T23, T1Z, T19, Th, T31, T35, T1e, T44, T41; Chris@42: E T1a, T6W, T70, T55, T59, T3v, T3z, Tf, T1R, T2N, T2Q, T1V, T1p, T1l, Tm; Chris@42: { Chris@42: E T1H, T1s, T2g, Tg, Tw, TH, T2t, T47, T3h, T3M, T4w, T28, T3Q, T4A, T2c; Chris@42: E Ts; Chris@42: { Chris@42: E T4, T13, TI, TF, TZ, Ta, T9; Chris@42: T2 = W[0]; Chris@42: T3 = W[2]; Chris@42: Tc = W[5]; Chris@42: T8 = W[4]; Chris@42: Te = W[6]; Chris@42: T4 = T2 * T3; Chris@42: T13 = T2 * Tc; Chris@42: TI = T3 * Tc; Chris@42: TF = T3 * T8; Chris@42: T1H = T8 * Te; Chris@42: TZ = T2 * T8; Chris@42: T5 = W[1]; Chris@42: T6 = W[3]; Chris@42: T1s = T3 * Te; Chris@42: T2g = T2 * Te; Chris@42: T14 = FNMS(T5, T8, T13); Chris@42: T3d = FMA(T5, T8, T13); Chris@42: T3i = FNMS(T6, T8, TI); Chris@42: TJ = FMA(T6, T8, TI); Chris@42: T7 = FNMS(T5, T6, T4); Chris@42: Tr = FMA(T5, T6, T4); Chris@42: Ta = T2 * T6; Chris@42: Tg = T7 * Tc; Chris@42: Tw = Tr * Tc; Chris@42: T3g = FMA(T6, Tc, TF); Chris@42: TG = FNMS(T6, Tc, TF); Chris@42: T10 = FMA(T5, Tc, TZ); Chris@42: T3a = FNMS(T5, Tc, TZ); Chris@42: TH = TG * Te; Chris@42: T2t = T10 * Te; Chris@42: T47 = T3a * Te; Chris@42: T3h = T3g * Te; Chris@42: TL = W[8]; Chris@42: TP = W[9]; Chris@42: T9 = T7 * T8; Chris@42: Tb = FMA(T5, T3, Ta); Chris@42: Tt = FNMS(T5, T3, Ta); Chris@42: T3M = T2 * TL; Chris@42: T4w = T8 * TL; Chris@42: T28 = T3 * TL; Chris@42: T3Q = T2 * TP; Chris@42: T4A = T8 * TP; Chris@42: T2c = T3 * TP; Chris@42: T17 = FNMS(Tb, Tc, T9); Chris@42: Td = FMA(Tb, Tc, T9); Chris@42: Ts = Tr * T8; Chris@42: Ti = W[7]; Chris@42: } Chris@42: { Chris@42: E T5r, T80, T1L, T2k, T1w, T5z, T2B, T2v; Chris@42: T3N = FMA(T5, TP, T3M); Chris@42: T3R = FNMS(T5, TL, T3Q); Chris@42: T1i = FMA(Tt, Tc, Ts); Chris@42: Tu = FNMS(Tt, Tc, Ts); Chris@42: T1I = FNMS(Tc, Ti, T1H); Chris@42: T2U = FMA(Tc, Ti, T1H); Chris@42: T1t = FMA(T6, Ti, T1s); Chris@42: T3U = FNMS(T6, Ti, T1s); Chris@42: T5O = FNMS(T3d, Ti, T47); Chris@42: T48 = FMA(T3d, Ti, T47); Chris@42: T2u = FMA(T14, Ti, T2t); Chris@42: T7B = FNMS(T14, Ti, T2t); Chris@42: T1L = T8 * Ti; Chris@42: T2k = T2 * Ti; Chris@42: T1w = T3 * Ti; Chris@42: TK = FMA(TJ, Ti, TH); Chris@42: T79 = FNMS(TJ, Ti, TH); Chris@42: T3D = FMA(T5, Ti, T2g); Chris@42: T2h = FNMS(T5, Ti, T2g); Chris@42: T2l = FMA(T5, Te, T2k); Chris@42: T3G = FNMS(T5, Te, T2k); Chris@42: T1x = FNMS(T6, Te, T1w); Chris@42: T3X = FMA(T6, Te, T1w); Chris@42: T2d = FNMS(T6, TL, T2c); Chris@42: T1M = FMA(Tc, Te, T1L); Chris@42: T2X = FNMS(Tc, Te, T1L); Chris@42: T4B = FNMS(Tc, TL, T4A); Chris@42: T4x = FMA(Tc, TP, T4w); Chris@42: T3j = FMA(T3i, Ti, T3h); Chris@42: T4T = FNMS(T3i, Ti, T3h); Chris@42: T29 = FMA(T6, TP, T28); Chris@42: T5r = T3g * TL; Chris@42: T80 = T7 * TP; Chris@42: { Chris@42: E T7M, T7g, T63, T5v, T7W; Chris@42: T5v = T3g * TP; Chris@42: T7W = T7 * TL; Chris@42: T5s = FMA(T3i, TP, T5r); Chris@42: T81 = FNMS(Tb, TL, T80); Chris@42: T5w = FNMS(T3i, TL, T5v); Chris@42: T7X = FMA(Tb, TP, T7W); Chris@42: T7M = TG * TL; Chris@42: T7g = T10 * TL; Chris@42: T63 = T3a * TP; Chris@42: { Chris@42: E T6d, T7k, T69, T5Z, T7Q; Chris@42: T69 = Tr * TL; Chris@42: T7N = FMA(TJ, TP, T7M); Chris@42: T7h = FMA(T14, TP, T7g); Chris@42: T64 = FNMS(T3d, TL, T63); Chris@42: T6a = FMA(Tt, TP, T69); Chris@42: T6d = Tr * TP; Chris@42: T7k = T10 * TP; Chris@42: T5Z = T3a * TL; Chris@42: T7Q = TG * TP; Chris@42: T6e = FNMS(Tt, TL, T6d); Chris@42: T7l = FNMS(T14, TL, T7k); Chris@42: T60 = FMA(T3d, TP, T5Z); Chris@42: T7R = FNMS(TJ, TL, T7Q); Chris@42: T5z = Tr * Te; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T6I, T5D, T6M, T6s, T6w; Chris@42: T6I = T7 * Te; Chris@42: T5D = Tr * Ti; Chris@42: T6M = T7 * Ti; Chris@42: T6h = FNMS(Tt, Ti, T5z); Chris@42: T5A = FMA(Tt, Ti, T5z); Chris@42: T7o = FMA(Tb, Ti, T6I); Chris@42: T6J = FNMS(Tb, Ti, T6I); Chris@42: T6k = FMA(Tt, Te, T5D); Chris@42: T5E = FNMS(Tt, Te, T5D); Chris@42: T6N = FMA(Tb, Te, T6M); Chris@42: T7r = FNMS(Tb, Te, T6M); Chris@42: T6s = T2U * TL; Chris@42: T6w = T2U * TP; Chris@42: { Chris@42: E TN, TT, TM, T2w; Chris@42: TN = TG * Ti; Chris@42: T2w = T10 * Ti; Chris@42: T6x = FNMS(T2X, TL, T6w); Chris@42: T6t = FMA(T2X, TP, T6s); Chris@42: T7c = FMA(TJ, Te, TN); Chris@42: TO = FNMS(TJ, Te, TN); Chris@42: TT = TK * TP; Chris@42: TM = TK * TL; Chris@42: T2x = FNMS(T14, Te, T2w); Chris@42: T7E = FMA(T14, Te, T2w); Chris@42: TU = FNMS(TO, TL, TT); Chris@42: TQ = FMA(TO, TP, TM); Chris@42: T2B = T2u * TP; Chris@42: T2v = T2u * TL; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T1Y, T22, Tv, TB; Chris@42: { Chris@42: E T49, T4f, T4a, T3l, T3q, T3k; Chris@42: T4a = T3a * Ti; Chris@42: T2C = FNMS(T2x, TL, T2B); Chris@42: T2y = FMA(T2x, TP, T2v); Chris@42: T5R = FMA(T3d, Te, T4a); Chris@42: T4b = FNMS(T3d, Te, T4a); Chris@42: T49 = T48 * TL; Chris@42: T4f = T48 * TP; Chris@42: T3l = T3g * Ti; Chris@42: T4c = FMA(T4b, TP, T49); Chris@42: T4g = FNMS(T4b, TL, T4f); Chris@42: T4W = FMA(T3i, Te, T3l); Chris@42: T3m = FNMS(T3i, Te, T3l); Chris@42: T1Y = Tu * TL; Chris@42: T3q = T3j * TP; Chris@42: T3k = T3j * TL; Chris@42: T22 = Tu * TP; Chris@42: Tv = Tu * Te; Chris@42: T3r = FNMS(T3m, TL, T3q); Chris@42: T3n = FMA(T3m, TP, T3k); Chris@42: TB = Tu * Ti; Chris@42: T1k = FNMS(Tt, T8, Tw); Chris@42: Tx = FMA(Tt, T8, Tw); Chris@42: } Chris@42: { Chris@42: E T30, T34, T18, T1d; Chris@42: T30 = T17 * TL; Chris@42: T34 = T17 * TP; Chris@42: T18 = T17 * Te; Chris@42: Ty = FMA(Tx, Ti, Tv); Chris@42: T4p = FNMS(Tx, Ti, Tv); Chris@42: T4s = FMA(Tx, Te, TB); Chris@42: TC = FNMS(Tx, Te, TB); Chris@42: T23 = FNMS(Tx, TL, T22); Chris@42: T1Z = FMA(Tx, TP, T1Y); Chris@42: T1d = T17 * Ti; Chris@42: T19 = FMA(Tb, T8, Tg); Chris@42: Th = FNMS(Tb, T8, Tg); Chris@42: { Chris@42: E T1j, T1o, T1Q, T1U; Chris@42: T1j = T1i * TL; Chris@42: { Chris@42: E T6V, T6Z, T54, T58; Chris@42: T6V = Ty * TL; Chris@42: T6Z = Ty * TP; Chris@42: T31 = FMA(T19, TP, T30); Chris@42: T35 = FNMS(T19, TL, T34); Chris@42: T1e = FMA(T19, Te, T1d); Chris@42: T44 = FNMS(T19, Te, T1d); Chris@42: T41 = FMA(T19, Ti, T18); Chris@42: T1a = FNMS(T19, Ti, T18); Chris@42: T6W = FMA(TC, TP, T6V); Chris@42: T70 = FNMS(TC, TL, T6Z); Chris@42: T1o = T1i * TP; Chris@42: T54 = T41 * TL; Chris@42: T58 = T41 * TP; Chris@42: T1Q = T1i * Te; Chris@42: T1U = T1i * Ti; Chris@42: T55 = FMA(T44, TP, T54); Chris@42: T59 = FNMS(T44, TL, T58); Chris@42: } Chris@42: T3v = Td * TL; Chris@42: T3z = Td * TP; Chris@42: Tf = Td * Te; Chris@42: T1R = FMA(T1k, Ti, T1Q); Chris@42: T2N = FNMS(T1k, Ti, T1Q); Chris@42: T2Q = FMA(T1k, Te, T1U); Chris@42: T1V = FNMS(T1k, Te, T1U); Chris@42: T1p = FNMS(T1k, TL, T1o); Chris@42: T1l = FMA(T1k, TP, T1j); Chris@42: Tm = Td * Ti; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tl9, TlD, TY, Tg4, T8w, TdS, TkE, Tkd, T2G, Tge, Tgh, TiK, Te1, T98, Te0; Chris@42: E T9f, Te5, T9p, Tgq, T39, Te8, T9M, TiN, Tgn, TeE, TbI, Thr, T74, TeP, TcB; Chris@42: E Tja, Thc, T8D, TdT, T1B, TkD, T8K, TdU, Tg7, Tk7, T8T, TdY, T27, Tg9, T90; Chris@42: E TdX, Tgc, TiJ, T9Y, Tec, T4k, TgB, Tal, Tef, Tgy, TiT, Taz, Tel, T5d, Th0; Chris@42: E Tbs, Tew, TgL, TiZ, T3K, Tgo, Tgt, TiO, T9P, Te6, T9E, Te9, T4L, Tgz, TgE; Chris@42: E TiU, Tao, Ted, Tad, Teg, T5I, TgM, Th3, Tj0, Tbv, Tem, TaO, Tex, T7v, Thd; Chris@42: E Thu, Tjb, TcE, TeF, TbX, TeQ, T68, Tj5, Tez, Teq, Tbj, Tbx, TgS, Th5, T6B; Chris@42: E Tj6, TeA, Tet, Tb4, Tby, TgX, Th6, T7V, Tjg, TeS, TeJ, Tcs, TcG, Thj, Thw; Chris@42: E T84, T83, T85, Tc7, T8k, Tc3, T86, T89, T8b; Chris@42: { Chris@42: E T3w, T3A, T4H, T4E, T8e, T8i, T5j, T5n, T4U, T4S, T4V, Tau, T5b, Tbq, T4X; Chris@42: E T50, T52; Chris@42: { Chris@42: E T72, Tcz, Tcv, T6Q, Tha, TbG, T6U, Tcx, T99, T9e; Chris@42: { Chris@42: E T1, Tkb, Tp, Tka, TR, TV, TE, T8s, TS, T8t; Chris@42: { Chris@42: E Tn, Tj, T8d, T8h, T5i, T5m; Chris@42: T1 = ri[0]; Chris@42: T8d = T1R * TL; Chris@42: T8h = T1R * TP; Chris@42: T3w = FMA(Th, TP, T3v); Chris@42: T3A = FNMS(Th, TL, T3z); Chris@42: Tn = FMA(Th, Te, Tm); Chris@42: T4H = FNMS(Th, Te, Tm); Chris@42: T4E = FMA(Th, Ti, Tf); Chris@42: Tj = FNMS(Th, Ti, Tf); Chris@42: T8e = FMA(T1V, TP, T8d); Chris@42: T8i = FNMS(T1V, TL, T8h); Chris@42: Tkb = ii[0]; Chris@42: T5i = T4E * TL; Chris@42: T5m = T4E * TP; Chris@42: { Chris@42: E Tk, To, Tl, Tk9; Chris@42: Tk = ri[WS(rs, 32)]; Chris@42: To = ii[WS(rs, 32)]; Chris@42: T5j = FMA(T4H, TP, T5i); Chris@42: T5n = FNMS(T4H, TL, T5m); Chris@42: Tl = Tj * Tk; Chris@42: Tk9 = Tj * To; Chris@42: { Chris@42: E Tz, TD, TA, T8r; Chris@42: Tz = ri[WS(rs, 16)]; Chris@42: TD = ii[WS(rs, 16)]; Chris@42: Tp = FMA(Tn, To, Tl); Chris@42: Tka = FNMS(Tn, Tk, Tk9); Chris@42: TA = Ty * Tz; Chris@42: T8r = Ty * TD; Chris@42: TR = ri[WS(rs, 48)]; Chris@42: TV = ii[WS(rs, 48)]; Chris@42: TE = FMA(TC, TD, TA); Chris@42: T8s = FNMS(TC, Tz, T8r); Chris@42: TS = TQ * TR; Chris@42: T8t = TQ * TV; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E T8q, Tq, Tl7, Tkc, TW, T8u; Chris@42: T8q = T1 - Tp; Chris@42: Tq = T1 + Tp; Chris@42: Tl7 = Tkb - Tka; Chris@42: Tkc = Tka + Tkb; Chris@42: TW = FMA(TU, TV, TS); Chris@42: T8u = FNMS(TU, TR, T8t); Chris@42: { Chris@42: E TX, Tl8, T8v, Tk8; Chris@42: TX = TE + TW; Chris@42: Tl8 = TE - TW; Chris@42: T8v = T8s - T8u; Chris@42: Tk8 = T8s + T8u; Chris@42: Tl9 = Tl7 - Tl8; Chris@42: TlD = Tl8 + Tl7; Chris@42: TY = Tq + TX; Chris@42: Tg4 = Tq - TX; Chris@42: T8w = T8q - T8v; Chris@42: TdS = T8q + T8v; Chris@42: TkE = Tkc - Tk8; Chris@42: Tkd = Tk8 + Tkc; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E T2f, T93, T2E, T9d, T2n, T95, T2s, T9b; Chris@42: { Chris@42: E T2a, T2e, T2i, T2m; Chris@42: T2a = ri[WS(rs, 60)]; Chris@42: T2e = ii[WS(rs, 60)]; Chris@42: { Chris@42: E T2z, T2D, T2b, T92, T2A, T9c; Chris@42: T2z = ri[WS(rs, 44)]; Chris@42: T2D = ii[WS(rs, 44)]; Chris@42: T2b = T29 * T2a; Chris@42: T92 = T29 * T2e; Chris@42: T2A = T2y * T2z; Chris@42: T9c = T2y * T2D; Chris@42: T2f = FMA(T2d, T2e, T2b); Chris@42: T93 = FNMS(T2d, T2a, T92); Chris@42: T2E = FMA(T2C, T2D, T2A); Chris@42: T9d = FNMS(T2C, T2z, T9c); Chris@42: } Chris@42: T2i = ri[WS(rs, 28)]; Chris@42: T2m = ii[WS(rs, 28)]; Chris@42: { Chris@42: E T2p, T2r, T2j, T94, T2q, T9a; Chris@42: T2p = ri[WS(rs, 12)]; Chris@42: T2r = ii[WS(rs, 12)]; Chris@42: T2j = T2h * T2i; Chris@42: T94 = T2h * T2m; Chris@42: T2q = TG * T2p; Chris@42: T9a = TG * T2r; Chris@42: T2n = FMA(T2l, T2m, T2j); Chris@42: T95 = FNMS(T2l, T2i, T94); Chris@42: T2s = FMA(TJ, T2r, T2q); Chris@42: T9b = FNMS(TJ, T2p, T9a); Chris@42: } Chris@42: } Chris@42: { Chris@42: E T2o, Tgf, T96, T97, T2F, Tgg; Chris@42: T99 = T2f - T2n; Chris@42: T2o = T2f + T2n; Chris@42: Tgf = T93 + T95; Chris@42: T96 = T93 - T95; Chris@42: T97 = T2s - T2E; Chris@42: T2F = T2s + T2E; Chris@42: Tgg = T9b + T9d; Chris@42: T9e = T9b - T9d; Chris@42: T2G = T2o + T2F; Chris@42: Tge = T2o - T2F; Chris@42: Tgh = Tgf - Tgg; Chris@42: TiK = Tgf + Tgg; Chris@42: Te1 = T96 - T97; Chris@42: T98 = T96 + T97; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T9K, T2T, T9G, T9n, Tgl, T9o, T38, T9I; Chris@42: { Chris@42: E T2M, T9k, T37, T2V, T2S, T2W, T2Y, T9m, T32, T33, T36, T2Z, T9H; Chris@42: { Chris@42: E T2J, T2L, T2K, T9j; Chris@42: T2J = ri[WS(rs, 2)]; Chris@42: T2L = ii[WS(rs, 2)]; Chris@42: T32 = ri[WS(rs, 50)]; Chris@42: Te0 = T99 + T9e; Chris@42: T9f = T99 - T9e; Chris@42: T2K = Tr * T2J; Chris@42: T9j = Tr * T2L; Chris@42: T33 = T31 * T32; Chris@42: T36 = ii[WS(rs, 50)]; Chris@42: T2M = FMA(Tt, T2L, T2K); Chris@42: T9k = FNMS(Tt, T2J, T9j); Chris@42: } Chris@42: { Chris@42: E T2O, T9J, T2R, T2P, T9l; Chris@42: T2O = ri[WS(rs, 34)]; Chris@42: T37 = FMA(T35, T36, T33); Chris@42: T9J = T31 * T36; Chris@42: T2R = ii[WS(rs, 34)]; Chris@42: T2P = T2N * T2O; Chris@42: T2V = ri[WS(rs, 18)]; Chris@42: T9K = FNMS(T35, T32, T9J); Chris@42: T9l = T2N * T2R; Chris@42: T2S = FMA(T2Q, T2R, T2P); Chris@42: T2W = T2U * T2V; Chris@42: T2Y = ii[WS(rs, 18)]; Chris@42: T9m = FNMS(T2Q, T2O, T9l); Chris@42: } Chris@42: T2T = T2M + T2S; Chris@42: T9G = T2M - T2S; Chris@42: T2Z = FMA(T2X, T2Y, T2W); Chris@42: T9H = T2U * T2Y; Chris@42: T9n = T9k - T9m; Chris@42: Tgl = T9k + T9m; Chris@42: T9o = T2Z - T37; Chris@42: T38 = T2Z + T37; Chris@42: T9I = FNMS(T2X, T2V, T9H); Chris@42: } Chris@42: { Chris@42: E T6H, TbD, T6P, T6R, T6T, TbF, T6S, Tcw; Chris@42: { Chris@42: E T6X, T71, T6E, TbC, T6K, TbE; Chris@42: { Chris@42: E T6F, T6G, T9L, Tgm; Chris@42: T6E = ri[WS(rs, 63)]; Chris@42: Te5 = T9n - T9o; Chris@42: T9p = T9n + T9o; Chris@42: Tgq = T2T - T38; Chris@42: T39 = T2T + T38; Chris@42: T9L = T9I - T9K; Chris@42: Tgm = T9I + T9K; Chris@42: T6F = TL * T6E; Chris@42: T6G = ii[WS(rs, 63)]; Chris@42: Te8 = T9G + T9L; Chris@42: T9M = T9G - T9L; Chris@42: TiN = Tgl + Tgm; Chris@42: Tgn = Tgl - Tgm; Chris@42: TbC = TL * T6G; Chris@42: T6H = FMA(TP, T6G, T6F); Chris@42: } Chris@42: T6X = ri[WS(rs, 47)]; Chris@42: T71 = ii[WS(rs, 47)]; Chris@42: TbD = FNMS(TP, T6E, TbC); Chris@42: { Chris@42: E T6O, T6L, T6Y, Tcy; Chris@42: T6K = ri[WS(rs, 31)]; Chris@42: T6Y = T6W * T6X; Chris@42: Tcy = T6W * T71; Chris@42: T6O = ii[WS(rs, 31)]; Chris@42: T6L = T6J * T6K; Chris@42: T72 = FMA(T70, T71, T6Y); Chris@42: Tcz = FNMS(T70, T6X, Tcy); Chris@42: TbE = T6J * T6O; Chris@42: T6P = FMA(T6N, T6O, T6L); Chris@42: } Chris@42: T6R = ri[WS(rs, 15)]; Chris@42: T6T = ii[WS(rs, 15)]; Chris@42: TbF = FNMS(T6N, T6K, TbE); Chris@42: } Chris@42: Tcv = T6H - T6P; Chris@42: T6Q = T6H + T6P; Chris@42: T6S = TK * T6R; Chris@42: Tcw = TK * T6T; Chris@42: Tha = TbD + TbF; Chris@42: TbG = TbD - TbF; Chris@42: T6U = FMA(TO, T6T, T6S); Chris@42: Tcx = FNMS(TO, T6R, Tcw); Chris@42: } Chris@42: } Chris@42: { Chris@42: E T1J, T1G, T1K, T8O, T25, T8Y, T1N, T1S, T1W; Chris@42: { Chris@42: E T1b, T16, T1c, T8y, T1z, T8I, T1f, T1m, T1q; Chris@42: { Chris@42: E T11, T12, T15, T1u, T1y, T8x, T1v, T8H; Chris@42: T11 = ri[WS(rs, 8)]; Chris@42: { Chris@42: E TbH, T73, TcA, Thb; Chris@42: TbH = T6U - T72; Chris@42: T73 = T6U + T72; Chris@42: TcA = Tcx - Tcz; Chris@42: Thb = Tcx + Tcz; Chris@42: TeE = TbG - TbH; Chris@42: TbI = TbG + TbH; Chris@42: Thr = T6Q - T73; Chris@42: T74 = T6Q + T73; Chris@42: TeP = Tcv + TcA; Chris@42: TcB = Tcv - TcA; Chris@42: Tja = Tha + Thb; Chris@42: Thc = Tha - Thb; Chris@42: T12 = T10 * T11; Chris@42: } Chris@42: T15 = ii[WS(rs, 8)]; Chris@42: T1u = ri[WS(rs, 24)]; Chris@42: T1y = ii[WS(rs, 24)]; Chris@42: T1b = ri[WS(rs, 40)]; Chris@42: T16 = FMA(T14, T15, T12); Chris@42: T8x = T10 * T15; Chris@42: T1v = T1t * T1u; Chris@42: T8H = T1t * T1y; Chris@42: T1c = T1a * T1b; Chris@42: T8y = FNMS(T14, T11, T8x); Chris@42: T1z = FMA(T1x, T1y, T1v); Chris@42: T8I = FNMS(T1x, T1u, T8H); Chris@42: T1f = ii[WS(rs, 40)]; Chris@42: T1m = ri[WS(rs, 56)]; Chris@42: T1q = ii[WS(rs, 56)]; Chris@42: } Chris@42: { Chris@42: E T1D, T1E, T1F, T20, T24, T8N, T21, T8X; Chris@42: { Chris@42: E T1h, T8C, T8A, T1r, T8G, Tg5, T8B; Chris@42: T1D = ri[WS(rs, 4)]; Chris@42: { Chris@42: E T1g, T8z, T1n, T8F; Chris@42: T1g = FMA(T1e, T1f, T1c); Chris@42: T8z = T1a * T1f; Chris@42: T1n = T1l * T1m; Chris@42: T8F = T1l * T1q; Chris@42: T1h = T16 + T1g; Chris@42: T8C = T16 - T1g; Chris@42: T8A = FNMS(T1e, T1b, T8z); Chris@42: T1r = FMA(T1p, T1q, T1n); Chris@42: T8G = FNMS(T1p, T1m, T8F); Chris@42: T1E = T7 * T1D; Chris@42: } Chris@42: Tg5 = T8y + T8A; Chris@42: T8B = T8y - T8A; Chris@42: { Chris@42: E T1A, T8E, Tg6, T8J; Chris@42: T1A = T1r + T1z; Chris@42: T8E = T1r - T1z; Chris@42: Tg6 = T8G + T8I; Chris@42: T8J = T8G - T8I; Chris@42: T8D = T8B - T8C; Chris@42: TdT = T8C + T8B; Chris@42: T1B = T1h + T1A; Chris@42: TkD = T1A - T1h; Chris@42: T8K = T8E + T8J; Chris@42: TdU = T8E - T8J; Chris@42: Tg7 = Tg5 - Tg6; Chris@42: Tk7 = Tg5 + Tg6; Chris@42: T1F = ii[WS(rs, 4)]; Chris@42: } Chris@42: } Chris@42: T20 = ri[WS(rs, 52)]; Chris@42: T24 = ii[WS(rs, 52)]; Chris@42: T1J = ri[WS(rs, 36)]; Chris@42: T1G = FMA(Tb, T1F, T1E); Chris@42: T8N = T7 * T1F; Chris@42: T21 = T1Z * T20; Chris@42: T8X = T1Z * T24; Chris@42: T1K = T1I * T1J; Chris@42: T8O = FNMS(Tb, T1D, T8N); Chris@42: T25 = FMA(T23, T24, T21); Chris@42: T8Y = FNMS(T23, T20, T8X); Chris@42: T1N = ii[WS(rs, 36)]; Chris@42: T1S = ri[WS(rs, 20)]; Chris@42: T1W = ii[WS(rs, 20)]; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T3V, T3T, T3W, T9T, T4i, Taj, T3Y, T42, T45; Chris@42: { Chris@42: E T3O, T3P, T3S, T4d, T4h, T9S, T4e, Tai; Chris@42: { Chris@42: E T1P, T8U, T8Q, T1X, T8W, Tga, T8R; Chris@42: T3O = ri[WS(rs, 62)]; Chris@42: { Chris@42: E T1O, T8P, T1T, T8V; Chris@42: T1O = FMA(T1M, T1N, T1K); Chris@42: T8P = T1I * T1N; Chris@42: T1T = T1R * T1S; Chris@42: T8V = T1R * T1W; Chris@42: T1P = T1G + T1O; Chris@42: T8U = T1G - T1O; Chris@42: T8Q = FNMS(T1M, T1J, T8P); Chris@42: T1X = FMA(T1V, T1W, T1T); Chris@42: T8W = FNMS(T1V, T1S, T8V); Chris@42: T3P = T3N * T3O; Chris@42: } Chris@42: Tga = T8O + T8Q; Chris@42: T8R = T8O - T8Q; Chris@42: { Chris@42: E T26, T8S, Tgb, T8Z; Chris@42: T26 = T1X + T25; Chris@42: T8S = T1X - T25; Chris@42: Tgb = T8W + T8Y; Chris@42: T8Z = T8W - T8Y; Chris@42: T8T = T8R + T8S; Chris@42: TdY = T8R - T8S; Chris@42: T27 = T1P + T26; Chris@42: Tg9 = T1P - T26; Chris@42: T90 = T8U - T8Z; Chris@42: TdX = T8U + T8Z; Chris@42: Tgc = Tga - Tgb; Chris@42: TiJ = Tga + Tgb; Chris@42: T3S = ii[WS(rs, 62)]; Chris@42: } Chris@42: } Chris@42: T4d = ri[WS(rs, 46)]; Chris@42: T4h = ii[WS(rs, 46)]; Chris@42: T3V = ri[WS(rs, 30)]; Chris@42: T3T = FMA(T3R, T3S, T3P); Chris@42: T9S = T3N * T3S; Chris@42: T4e = T4c * T4d; Chris@42: Tai = T4c * T4h; Chris@42: T3W = T3U * T3V; Chris@42: T9T = FNMS(T3R, T3O, T9S); Chris@42: T4i = FMA(T4g, T4h, T4e); Chris@42: Taj = FNMS(T4g, T4d, Tai); Chris@42: T3Y = ii[WS(rs, 30)]; Chris@42: T42 = ri[WS(rs, 14)]; Chris@42: T45 = ii[WS(rs, 14)]; Chris@42: } Chris@42: { Chris@42: E T4P, T4Q, T4R, T56, T5a, Tat, T57, Tbp; Chris@42: { Chris@42: E T40, Taf, T9V, T46, Tah, Tgw, T9W; Chris@42: T4P = ri[WS(rs, 1)]; Chris@42: { Chris@42: E T3Z, T9U, T43, Tag; Chris@42: T3Z = FMA(T3X, T3Y, T3W); Chris@42: T9U = T3U * T3Y; Chris@42: T43 = T41 * T42; Chris@42: Tag = T41 * T45; Chris@42: T40 = T3T + T3Z; Chris@42: Taf = T3T - T3Z; Chris@42: T9V = FNMS(T3X, T3V, T9U); Chris@42: T46 = FMA(T44, T45, T43); Chris@42: Tah = FNMS(T44, T42, Tag); Chris@42: T4Q = T2 * T4P; Chris@42: } Chris@42: Tgw = T9T + T9V; Chris@42: T9W = T9T - T9V; Chris@42: { Chris@42: E T4j, T9X, Tgx, Tak; Chris@42: T4j = T46 + T4i; Chris@42: T9X = T46 - T4i; Chris@42: Tgx = Tah + Taj; Chris@42: Tak = Tah - Taj; Chris@42: T9Y = T9W + T9X; Chris@42: Tec = T9W - T9X; Chris@42: T4k = T40 + T4j; Chris@42: TgB = T40 - T4j; Chris@42: Tal = Taf - Tak; Chris@42: Tef = Taf + Tak; Chris@42: Tgy = Tgw - Tgx; Chris@42: TiT = Tgw + Tgx; Chris@42: T4R = ii[WS(rs, 1)]; Chris@42: } Chris@42: } Chris@42: T56 = ri[WS(rs, 49)]; Chris@42: T5a = ii[WS(rs, 49)]; Chris@42: T4U = ri[WS(rs, 33)]; Chris@42: T4S = FMA(T5, T4R, T4Q); Chris@42: Tat = T2 * T4R; Chris@42: T57 = T55 * T56; Chris@42: Tbp = T55 * T5a; Chris@42: T4V = T4T * T4U; Chris@42: Tau = FNMS(T5, T4P, Tat); Chris@42: T5b = FMA(T59, T5a, T57); Chris@42: Tbq = FNMS(T59, T56, Tbp); Chris@42: T4X = ii[WS(rs, 33)]; Chris@42: T50 = ri[WS(rs, 17)]; Chris@42: T52 = ii[WS(rs, 17)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E T7a, T78, T7b, TbL, T7t, TbU, T7d, T7i, T7m; Chris@42: { Chris@42: E T4q, T4o, T4r, Ta1, T4J, Taa, T4t, T4y, T4C; Chris@42: { Chris@42: E T3o, T3f, T3p, T9s, T3I, T9B, T3s, T3x, T3B; Chris@42: { Chris@42: E T3b, T3c, T3e, T3E, T3H, T9r, T3F, T9A; Chris@42: { Chris@42: E T4Z, Tbm, Taw, T53, Tbo, TgJ, Tax; Chris@42: T3b = ri[WS(rs, 10)]; Chris@42: { Chris@42: E T4Y, Tav, T51, Tbn; Chris@42: T4Y = FMA(T4W, T4X, T4V); Chris@42: Tav = T4T * T4X; Chris@42: T51 = T48 * T50; Chris@42: Tbn = T48 * T52; Chris@42: T4Z = T4S + T4Y; Chris@42: Tbm = T4S - T4Y; Chris@42: Taw = FNMS(T4W, T4U, Tav); Chris@42: T53 = FMA(T4b, T52, T51); Chris@42: Tbo = FNMS(T4b, T50, Tbn); Chris@42: T3c = T3a * T3b; Chris@42: } Chris@42: TgJ = Tau + Taw; Chris@42: Tax = Tau - Taw; Chris@42: { Chris@42: E T5c, Tay, TgK, Tbr; Chris@42: T5c = T53 + T5b; Chris@42: Tay = T53 - T5b; Chris@42: TgK = Tbo + Tbq; Chris@42: Tbr = Tbo - Tbq; Chris@42: Taz = Tax + Tay; Chris@42: Tel = Tax - Tay; Chris@42: T5d = T4Z + T5c; Chris@42: Th0 = T4Z - T5c; Chris@42: Tbs = Tbm - Tbr; Chris@42: Tew = Tbm + Tbr; Chris@42: TgL = TgJ - TgK; Chris@42: TiZ = TgJ + TgK; Chris@42: T3e = ii[WS(rs, 10)]; Chris@42: } Chris@42: } Chris@42: T3E = ri[WS(rs, 26)]; Chris@42: T3H = ii[WS(rs, 26)]; Chris@42: T3o = ri[WS(rs, 42)]; Chris@42: T3f = FMA(T3d, T3e, T3c); Chris@42: T9r = T3a * T3e; Chris@42: T3F = T3D * T3E; Chris@42: T9A = T3D * T3H; Chris@42: T3p = T3n * T3o; Chris@42: T9s = FNMS(T3d, T3b, T9r); Chris@42: T3I = FMA(T3G, T3H, T3F); Chris@42: T9B = FNMS(T3G, T3E, T9A); Chris@42: T3s = ii[WS(rs, 42)]; Chris@42: T3x = ri[WS(rs, 58)]; Chris@42: T3B = ii[WS(rs, 58)]; Chris@42: } Chris@42: { Chris@42: E T4l, T4m, T4n, T4F, T4I, Ta0, T4G, Ta9; Chris@42: { Chris@42: E T3u, T9q, T9u, T3C, T9z, Tgr, T9v; Chris@42: T4l = ri[WS(rs, 6)]; Chris@42: { Chris@42: E T3t, T9t, T3y, T9y; Chris@42: T3t = FMA(T3r, T3s, T3p); Chris@42: T9t = T3n * T3s; Chris@42: T3y = T3w * T3x; Chris@42: T9y = T3w * T3B; Chris@42: T3u = T3f + T3t; Chris@42: T9q = T3f - T3t; Chris@42: T9u = FNMS(T3r, T3o, T9t); Chris@42: T3C = FMA(T3A, T3B, T3y); Chris@42: T9z = FNMS(T3A, T3x, T9y); Chris@42: T4m = T3g * T4l; Chris@42: } Chris@42: Tgr = T9s + T9u; Chris@42: T9v = T9s - T9u; Chris@42: { Chris@42: E T3J, T9x, Tgs, T9C; Chris@42: T3J = T3C + T3I; Chris@42: T9x = T3C - T3I; Chris@42: Tgs = T9z + T9B; Chris@42: T9C = T9z - T9B; Chris@42: { Chris@42: E T9w, T9O, T9D, T9N; Chris@42: T9w = T9q + T9v; Chris@42: T9O = T9v - T9q; Chris@42: T3K = T3u + T3J; Chris@42: Tgo = T3J - T3u; Chris@42: T9D = T9x - T9C; Chris@42: T9N = T9x + T9C; Chris@42: Tgt = Tgr - Tgs; Chris@42: TiO = Tgr + Tgs; Chris@42: T9P = T9N - T9O; Chris@42: Te6 = T9O + T9N; Chris@42: T9E = T9w - T9D; Chris@42: Te9 = T9w + T9D; Chris@42: T4n = ii[WS(rs, 6)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: T4F = ri[WS(rs, 22)]; Chris@42: T4I = ii[WS(rs, 22)]; Chris@42: T4q = ri[WS(rs, 38)]; Chris@42: T4o = FMA(T3i, T4n, T4m); Chris@42: Ta0 = T3g * T4n; Chris@42: T4G = T4E * T4F; Chris@42: Ta9 = T4E * T4I; Chris@42: T4r = T4p * T4q; Chris@42: Ta1 = FNMS(T3i, T4l, Ta0); Chris@42: T4J = FMA(T4H, T4I, T4G); Chris@42: Taa = FNMS(T4H, T4F, Ta9); Chris@42: T4t = ii[WS(rs, 38)]; Chris@42: T4y = ri[WS(rs, 54)]; Chris@42: T4C = ii[WS(rs, 54)]; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T5k, T5h, T5l, TaC, T5G, TaL, T5o, T5t, T5x; Chris@42: { Chris@42: E T5e, T5f, T5g, T5B, T5F, TaB, T5C, TaK; Chris@42: { Chris@42: E T4v, T9Z, Ta3, T4D, Ta8, TgC, Ta4; Chris@42: T5e = ri[WS(rs, 9)]; Chris@42: { Chris@42: E T4u, Ta2, T4z, Ta7; Chris@42: T4u = FMA(T4s, T4t, T4r); Chris@42: Ta2 = T4p * T4t; Chris@42: T4z = T4x * T4y; Chris@42: Ta7 = T4x * T4C; Chris@42: T4v = T4o + T4u; Chris@42: T9Z = T4o - T4u; Chris@42: Ta3 = FNMS(T4s, T4q, Ta2); Chris@42: T4D = FMA(T4B, T4C, T4z); Chris@42: Ta8 = FNMS(T4B, T4y, Ta7); Chris@42: T5f = T8 * T5e; Chris@42: } Chris@42: TgC = Ta1 + Ta3; Chris@42: Ta4 = Ta1 - Ta3; Chris@42: { Chris@42: E T4K, Ta6, TgD, Tab; Chris@42: T4K = T4D + T4J; Chris@42: Ta6 = T4D - T4J; Chris@42: TgD = Ta8 + Taa; Chris@42: Tab = Ta8 - Taa; Chris@42: { Chris@42: E Ta5, Tan, Tac, Tam; Chris@42: Ta5 = T9Z + Ta4; Chris@42: Tan = Ta4 - T9Z; Chris@42: T4L = T4v + T4K; Chris@42: Tgz = T4K - T4v; Chris@42: Tac = Ta6 - Tab; Chris@42: Tam = Ta6 + Tab; Chris@42: TgE = TgC - TgD; Chris@42: TiU = TgC + TgD; Chris@42: Tao = Tam - Tan; Chris@42: Ted = Tan + Tam; Chris@42: Tad = Ta5 - Tac; Chris@42: Teg = Ta5 + Tac; Chris@42: T5g = ii[WS(rs, 9)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: T5B = ri[WS(rs, 25)]; Chris@42: T5F = ii[WS(rs, 25)]; Chris@42: T5k = ri[WS(rs, 41)]; Chris@42: T5h = FMA(Tc, T5g, T5f); Chris@42: TaB = T8 * T5g; Chris@42: T5C = T5A * T5B; Chris@42: TaK = T5A * T5F; Chris@42: T5l = T5j * T5k; Chris@42: TaC = FNMS(Tc, T5e, TaB); Chris@42: T5G = FMA(T5E, T5F, T5C); Chris@42: TaL = FNMS(T5E, T5B, TaK); Chris@42: T5o = ii[WS(rs, 41)]; Chris@42: T5t = ri[WS(rs, 57)]; Chris@42: T5x = ii[WS(rs, 57)]; Chris@42: } Chris@42: { Chris@42: E T75, T76, T77, T7p, T7s, TbK, T7q, TbT; Chris@42: { Chris@42: E T5q, TaA, TaE, T5y, TaJ, Th1, TaF; Chris@42: T75 = ri[WS(rs, 7)]; Chris@42: { Chris@42: E T5p, TaD, T5u, TaI; Chris@42: T5p = FMA(T5n, T5o, T5l); Chris@42: TaD = T5j * T5o; Chris@42: T5u = T5s * T5t; Chris@42: TaI = T5s * T5x; Chris@42: T5q = T5h + T5p; Chris@42: TaA = T5h - T5p; Chris@42: TaE = FNMS(T5n, T5k, TaD); Chris@42: T5y = FMA(T5w, T5x, T5u); Chris@42: TaJ = FNMS(T5w, T5t, TaI); Chris@42: T76 = T1i * T75; Chris@42: } Chris@42: Th1 = TaC + TaE; Chris@42: TaF = TaC - TaE; Chris@42: { Chris@42: E T5H, TaH, Th2, TaM; Chris@42: T5H = T5y + T5G; Chris@42: TaH = T5y - T5G; Chris@42: Th2 = TaJ + TaL; Chris@42: TaM = TaJ - TaL; Chris@42: { Chris@42: E TaG, Tbu, TaN, Tbt; Chris@42: TaG = TaA + TaF; Chris@42: Tbu = TaF - TaA; Chris@42: T5I = T5q + T5H; Chris@42: TgM = T5H - T5q; Chris@42: TaN = TaH - TaM; Chris@42: Tbt = TaH + TaM; Chris@42: Th3 = Th1 - Th2; Chris@42: Tj0 = Th1 + Th2; Chris@42: Tbv = Tbt - Tbu; Chris@42: Tem = Tbu + Tbt; Chris@42: TaO = TaG - TaN; Chris@42: Tex = TaG + TaN; Chris@42: T77 = ii[WS(rs, 7)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: T7p = ri[WS(rs, 23)]; Chris@42: T7s = ii[WS(rs, 23)]; Chris@42: T7a = ri[WS(rs, 39)]; Chris@42: T78 = FMA(T1k, T77, T76); Chris@42: TbK = T1i * T77; Chris@42: T7q = T7o * T7p; Chris@42: TbT = T7o * T7s; Chris@42: T7b = T79 * T7a; Chris@42: TbL = FNMS(T1k, T75, TbK); Chris@42: T7t = FMA(T7r, T7s, T7q); Chris@42: TbU = FNMS(T7r, T7p, TbT); Chris@42: T7d = ii[WS(rs, 39)]; Chris@42: T7i = ri[WS(rs, 55)]; Chris@42: T7m = ii[WS(rs, 55)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E T6i, T6g, T6j, TaY, T6z, TaU, T6l, T6o, T6q; Chris@42: { Chris@42: E T5P, T5N, T5Q, Tbd, T66, Tb9, T5S, T5V, T5X; Chris@42: { Chris@42: E T5K, T5L, T5M, T61, T65, Tbc, T62, Tb8; Chris@42: { Chris@42: E T7f, TbJ, TbN, T7n, TbS, Ths, TbO; Chris@42: T5K = ri[WS(rs, 5)]; Chris@42: { Chris@42: E T7e, TbM, T7j, TbR; Chris@42: T7e = FMA(T7c, T7d, T7b); Chris@42: TbM = T79 * T7d; Chris@42: T7j = T7h * T7i; Chris@42: TbR = T7h * T7m; Chris@42: T7f = T78 + T7e; Chris@42: TbJ = T78 - T7e; Chris@42: TbN = FNMS(T7c, T7a, TbM); Chris@42: T7n = FMA(T7l, T7m, T7j); Chris@42: TbS = FNMS(T7l, T7i, TbR); Chris@42: T5L = Td * T5K; Chris@42: } Chris@42: Ths = TbL + TbN; Chris@42: TbO = TbL - TbN; Chris@42: { Chris@42: E T7u, TbQ, Tht, TbV; Chris@42: T7u = T7n + T7t; Chris@42: TbQ = T7n - T7t; Chris@42: Tht = TbS + TbU; Chris@42: TbV = TbS - TbU; Chris@42: { Chris@42: E TbP, TcD, TbW, TcC; Chris@42: TbP = TbJ + TbO; Chris@42: TcD = TbO - TbJ; Chris@42: T7v = T7f + T7u; Chris@42: Thd = T7u - T7f; Chris@42: TbW = TbQ - TbV; Chris@42: TcC = TbQ + TbV; Chris@42: Thu = Ths - Tht; Chris@42: Tjb = Ths + Tht; Chris@42: TcE = TcC - TcD; Chris@42: TeF = TcD + TcC; Chris@42: TbX = TbP - TbW; Chris@42: TeQ = TbP + TbW; Chris@42: T5M = ii[WS(rs, 5)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: T61 = ri[WS(rs, 53)]; Chris@42: T65 = ii[WS(rs, 53)]; Chris@42: T5P = ri[WS(rs, 37)]; Chris@42: T5N = FMA(Th, T5M, T5L); Chris@42: Tbc = Td * T5M; Chris@42: T62 = T60 * T61; Chris@42: Tb8 = T60 * T65; Chris@42: T5Q = T5O * T5P; Chris@42: Tbd = FNMS(Th, T5K, Tbc); Chris@42: T66 = FMA(T64, T65, T62); Chris@42: Tb9 = FNMS(T64, T61, Tb8); Chris@42: T5S = ii[WS(rs, 37)]; Chris@42: T5V = ri[WS(rs, 21)]; Chris@42: T5X = ii[WS(rs, 21)]; Chris@42: } Chris@42: { Chris@42: E T6b, T6c, T6f, T6u, T6y, TaX, T6v, TaT; Chris@42: { Chris@42: E T5U, Tb5, Tbf, T5Y, Tb7; Chris@42: T6b = ri[WS(rs, 61)]; Chris@42: { Chris@42: E T5T, Tbe, T5W, Tb6; Chris@42: T5T = FMA(T5R, T5S, T5Q); Chris@42: Tbe = T5O * T5S; Chris@42: T5W = T3j * T5V; Chris@42: Tb6 = T3j * T5X; Chris@42: T5U = T5N + T5T; Chris@42: Tb5 = T5N - T5T; Chris@42: Tbf = FNMS(T5R, T5P, Tbe); Chris@42: T5Y = FMA(T3m, T5X, T5W); Chris@42: Tb7 = FNMS(T3m, T5V, Tb6); Chris@42: T6c = T6a * T6b; Chris@42: } Chris@42: { Chris@42: E TgO, Tbg, T67, Tbh; Chris@42: TgO = Tbd + Tbf; Chris@42: Tbg = Tbd - Tbf; Chris@42: T67 = T5Y + T66; Chris@42: Tbh = T5Y - T66; Chris@42: { Chris@42: E TgP, Tba, Tbi, Teo; Chris@42: TgP = Tb7 + Tb9; Chris@42: Tba = Tb7 - Tb9; Chris@42: Tbi = Tbg + Tbh; Chris@42: Teo = Tbg - Tbh; Chris@42: { Chris@42: E TgR, Tbb, Tep, TgQ; Chris@42: TgR = T5U - T67; Chris@42: T68 = T5U + T67; Chris@42: Tbb = Tb5 - Tba; Chris@42: Tep = Tb5 + Tba; Chris@42: TgQ = TgO - TgP; Chris@42: Tj5 = TgO + TgP; Chris@42: Tez = FMA(KP414213562, Teo, Tep); Chris@42: Teq = FNMS(KP414213562, Tep, Teo); Chris@42: Tbj = FNMS(KP414213562, Tbi, Tbb); Chris@42: Tbx = FMA(KP414213562, Tbb, Tbi); Chris@42: TgS = TgQ - TgR; Chris@42: Th5 = TgR + TgQ; Chris@42: T6f = ii[WS(rs, 61)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: T6u = ri[WS(rs, 45)]; Chris@42: T6y = ii[WS(rs, 45)]; Chris@42: T6i = ri[WS(rs, 29)]; Chris@42: T6g = FMA(T6e, T6f, T6c); Chris@42: TaX = T6a * T6f; Chris@42: T6v = T6t * T6u; Chris@42: TaT = T6t * T6y; Chris@42: T6j = T6h * T6i; Chris@42: TaY = FNMS(T6e, T6b, TaX); Chris@42: T6z = FMA(T6x, T6y, T6v); Chris@42: TaU = FNMS(T6x, T6u, TaT); Chris@42: T6l = ii[WS(rs, 29)]; Chris@42: T6o = ri[WS(rs, 13)]; Chris@42: T6q = ii[WS(rs, 13)]; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T7C, T7A, T7D, Tcm, T7T, Tci, T7F, T7I, T7K; Chris@42: { Chris@42: E T7x, T7y, T7z, T7O, T7S, Tcl, T7P, Tch; Chris@42: { Chris@42: E T6n, TaQ, Tb0, T6r, TaS; Chris@42: T7x = ri[WS(rs, 3)]; Chris@42: { Chris@42: E T6m, TaZ, T6p, TaR; Chris@42: T6m = FMA(T6k, T6l, T6j); Chris@42: TaZ = T6h * T6l; Chris@42: T6p = T17 * T6o; Chris@42: TaR = T17 * T6q; Chris@42: T6n = T6g + T6m; Chris@42: TaQ = T6g - T6m; Chris@42: Tb0 = FNMS(T6k, T6i, TaZ); Chris@42: T6r = FMA(T19, T6q, T6p); Chris@42: TaS = FNMS(T19, T6o, TaR); Chris@42: T7y = T3 * T7x; Chris@42: } Chris@42: { Chris@42: E TgU, Tb1, T6A, Tb2; Chris@42: TgU = TaY + Tb0; Chris@42: Tb1 = TaY - Tb0; Chris@42: T6A = T6r + T6z; Chris@42: Tb2 = T6r - T6z; Chris@42: { Chris@42: E TgV, TaV, Tb3, Ter; Chris@42: TgV = TaS + TaU; Chris@42: TaV = TaS - TaU; Chris@42: Tb3 = Tb1 + Tb2; Chris@42: Ter = Tb1 - Tb2; Chris@42: { Chris@42: E TgT, TaW, Tes, TgW; Chris@42: TgT = T6n - T6A; Chris@42: T6B = T6n + T6A; Chris@42: TaW = TaQ - TaV; Chris@42: Tes = TaQ + TaV; Chris@42: TgW = TgU - TgV; Chris@42: Tj6 = TgU + TgV; Chris@42: TeA = FNMS(KP414213562, Ter, Tes); Chris@42: Tet = FMA(KP414213562, Tes, Ter); Chris@42: Tb4 = FMA(KP414213562, Tb3, TaW); Chris@42: Tby = FNMS(KP414213562, TaW, Tb3); Chris@42: TgX = TgT + TgW; Chris@42: Th6 = TgT - TgW; Chris@42: T7z = ii[WS(rs, 3)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: T7O = ri[WS(rs, 51)]; Chris@42: T7S = ii[WS(rs, 51)]; Chris@42: T7C = ri[WS(rs, 35)]; Chris@42: T7A = FMA(T6, T7z, T7y); Chris@42: Tcl = T3 * T7z; Chris@42: T7P = T7N * T7O; Chris@42: Tch = T7N * T7S; Chris@42: T7D = T7B * T7C; Chris@42: Tcm = FNMS(T6, T7x, Tcl); Chris@42: T7T = FMA(T7R, T7S, T7P); Chris@42: Tci = FNMS(T7R, T7O, Tch); Chris@42: T7F = ii[WS(rs, 35)]; Chris@42: T7I = ri[WS(rs, 19)]; Chris@42: T7K = ii[WS(rs, 19)]; Chris@42: } Chris@42: { Chris@42: E T7Y, T7Z, T82, T8f, T8j, Tc6, T8g, Tc2; Chris@42: { Chris@42: E T7H, Tce, Tco, T7L, Tcg; Chris@42: T7Y = ri[WS(rs, 59)]; Chris@42: { Chris@42: E T7G, Tcn, T7J, Tcf; Chris@42: T7G = FMA(T7E, T7F, T7D); Chris@42: Tcn = T7B * T7F; Chris@42: T7J = T2u * T7I; Chris@42: Tcf = T2u * T7K; Chris@42: T7H = T7A + T7G; Chris@42: Tce = T7A - T7G; Chris@42: Tco = FNMS(T7E, T7C, Tcn); Chris@42: T7L = FMA(T2x, T7K, T7J); Chris@42: Tcg = FNMS(T2x, T7I, Tcf); Chris@42: T7Z = T7X * T7Y; Chris@42: } Chris@42: { Chris@42: E Thf, Tcp, T7U, Tcq; Chris@42: Thf = Tcm + Tco; Chris@42: Tcp = Tcm - Tco; Chris@42: T7U = T7L + T7T; Chris@42: Tcq = T7L - T7T; Chris@42: { Chris@42: E Thg, Tcj, Tcr, TeH; Chris@42: Thg = Tcg + Tci; Chris@42: Tcj = Tcg - Tci; Chris@42: Tcr = Tcp + Tcq; Chris@42: TeH = Tcp - Tcq; Chris@42: { Chris@42: E Thi, Tck, TeI, Thh; Chris@42: Thi = T7H - T7U; Chris@42: T7V = T7H + T7U; Chris@42: Tck = Tce - Tcj; Chris@42: TeI = Tce + Tcj; Chris@42: Thh = Thf - Thg; Chris@42: Tjg = Thf + Thg; Chris@42: TeS = FMA(KP414213562, TeH, TeI); Chris@42: TeJ = FNMS(KP414213562, TeI, TeH); Chris@42: Tcs = FNMS(KP414213562, Tcr, Tck); Chris@42: TcG = FMA(KP414213562, Tck, Tcr); Chris@42: Thj = Thh - Thi; Chris@42: Thw = Thi + Thh; Chris@42: T82 = ii[WS(rs, 59)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: T8f = ri[WS(rs, 43)]; Chris@42: T8j = ii[WS(rs, 43)]; Chris@42: T84 = ri[WS(rs, 27)]; Chris@42: T83 = FMA(T81, T82, T7Z); Chris@42: Tc6 = T7X * T82; Chris@42: T8g = T8e * T8f; Chris@42: Tc2 = T8e * T8j; Chris@42: T85 = Te * T84; Chris@42: Tc7 = FNMS(T81, T7Y, Tc6); Chris@42: T8k = FMA(T8i, T8j, T8g); Chris@42: Tc3 = FNMS(T8i, T8f, Tc2); Chris@42: T86 = ii[WS(rs, 27)]; Chris@42: T89 = ri[WS(rs, 11)]; Chris@42: T8b = ii[WS(rs, 11)]; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E TeT, TeM, Tcd, TcH, Tho, Thx, Tkw, Tkv, Tl6, Tl5; Chris@42: { Chris@42: E TiI, Tkp, TiQ, TiS, TiL, Tkq, TiP, TiV, Tjf, Tjd, Tjc, Tji, Tj4, Tj2, Tj1; Chris@42: E Tj7, Tkh, Tki; Chris@42: { Chris@42: E TjG, T2I, Tkj, T4N, Tkk, Tkf, Tk5, TjJ, T8o, Tk2, TjL, T6D, TjY, TjU, Tk1; Chris@42: E TjO; Chris@42: { Chris@42: E T8m, Tjh, T3L, T4M, Tk6, Tke, TjH, TjI; Chris@42: { Chris@42: E T1C, T88, TbZ, Tc9, T8c, Tc1, T2H; Chris@42: T1C = TY + T1B; Chris@42: TiI = TY - T1B; Chris@42: { Chris@42: E T87, Tc8, T8a, Tc0; Chris@42: T87 = FMA(Ti, T86, T85); Chris@42: Tc8 = Te * T86; Chris@42: T8a = Tu * T89; Chris@42: Tc0 = Tu * T8b; Chris@42: T88 = T83 + T87; Chris@42: TbZ = T83 - T87; Chris@42: Tc9 = FNMS(Ti, T84, Tc8); Chris@42: T8c = FMA(Tx, T8b, T8a); Chris@42: Tc1 = FNMS(Tx, T89, Tc0); Chris@42: T2H = T27 + T2G; Chris@42: Tkp = T2G - T27; Chris@42: } Chris@42: { Chris@42: E Thl, Tca, T8l, Tcb; Chris@42: Thl = Tc7 + Tc9; Chris@42: Tca = Tc7 - Tc9; Chris@42: T8l = T8c + T8k; Chris@42: Tcb = T8c - T8k; Chris@42: { Chris@42: E Thm, Tc4, Tcc, TeK; Chris@42: Thm = Tc1 + Tc3; Chris@42: Tc4 = Tc1 - Tc3; Chris@42: Tcc = Tca + Tcb; Chris@42: TeK = Tca - Tcb; Chris@42: { Chris@42: E Thk, Tc5, TeL, Thn; Chris@42: Thk = T88 - T8l; Chris@42: T8m = T88 + T8l; Chris@42: Tc5 = TbZ - Tc4; Chris@42: TeL = TbZ + Tc4; Chris@42: Thn = Thl - Thm; Chris@42: Tjh = Thl + Thm; Chris@42: TeT = FNMS(KP414213562, TeK, TeL); Chris@42: TeM = FMA(KP414213562, TeL, TeK); Chris@42: Tcd = FMA(KP414213562, Tcc, Tc5); Chris@42: TcH = FNMS(KP414213562, Tc5, Tcc); Chris@42: Tho = Thk + Thn; Chris@42: Thx = Thk - Thn; Chris@42: TjG = T1C - T2H; Chris@42: T2I = T1C + T2H; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: TiQ = T39 - T3K; Chris@42: T3L = T39 + T3K; Chris@42: T4M = T4k + T4L; Chris@42: TiS = T4k - T4L; Chris@42: TiL = TiJ - TiK; Chris@42: Tk6 = TiJ + TiK; Chris@42: Tke = Tk7 + Tkd; Chris@42: Tkq = Tkd - Tk7; Chris@42: TiP = TiN - TiO; Chris@42: TjH = TiN + TiO; Chris@42: Tkj = T4M - T3L; Chris@42: T4N = T3L + T4M; Chris@42: Tkk = Tke - Tk6; Chris@42: Tkf = Tk6 + Tke; Chris@42: TjI = TiT + TiU; Chris@42: TiV = TiT - TiU; Chris@42: { Chris@42: E TjR, TjQ, TjS, T7w, T8n; Chris@42: Tjf = T74 - T7v; Chris@42: T7w = T74 + T7v; Chris@42: T8n = T7V + T8m; Chris@42: Tjd = T8m - T7V; Chris@42: Tjc = Tja - Tjb; Chris@42: TjR = Tja + Tjb; Chris@42: Tk5 = TjH + TjI; Chris@42: TjJ = TjH - TjI; Chris@42: TjQ = T7w - T8n; Chris@42: T8o = T7w + T8n; Chris@42: Tji = Tjg - Tjh; Chris@42: TjS = Tjg + Tjh; Chris@42: { Chris@42: E TjM, TjN, T5J, T6C, TjT; Chris@42: Tj4 = T5d - T5I; Chris@42: T5J = T5d + T5I; Chris@42: T6C = T68 + T6B; Chris@42: Tj2 = T6B - T68; Chris@42: TjT = TjR - TjS; Chris@42: Tk2 = TjR + TjS; Chris@42: Tj1 = TiZ - Tj0; Chris@42: TjM = TiZ + Tj0; Chris@42: TjL = T5J - T6C; Chris@42: T6D = T5J + T6C; Chris@42: Tj7 = Tj5 - Tj6; Chris@42: TjN = Tj5 + Tj6; Chris@42: TjY = TjQ + TjT; Chris@42: TjU = TjQ - TjT; Chris@42: Tk1 = TjM + TjN; Chris@42: TjO = TjM - TjN; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tk0, Tk3, TjW, Tko, Tkn, Tkl, Tkm, TjZ; Chris@42: { Chris@42: E TjP, TjX, Tk4, Tkg, T4O, T8p, TjK, TjV; Chris@42: Tk0 = T2I - T4N; Chris@42: T4O = T2I + T4N; Chris@42: T8p = T6D + T8o; Chris@42: Tkh = T8o - T6D; Chris@42: TjP = TjL + TjO; Chris@42: TjX = TjO - TjL; Chris@42: Tk3 = Tk1 - Tk2; Chris@42: Tk4 = Tk1 + Tk2; Chris@42: ri[0] = T4O + T8p; Chris@42: ri[WS(rs, 32)] = T4O - T8p; Chris@42: Tkg = Tk5 + Tkf; Chris@42: Tki = Tkf - Tk5; Chris@42: TjW = TjG - TjJ; Chris@42: TjK = TjG + TjJ; Chris@42: TjV = TjP + TjU; Chris@42: Tko = TjU - TjP; Chris@42: Tkn = Tkk - Tkj; Chris@42: Tkl = Tkj + Tkk; Chris@42: ii[WS(rs, 32)] = Tkg - Tk4; Chris@42: ii[0] = Tk4 + Tkg; Chris@42: ri[WS(rs, 8)] = FMA(KP707106781, TjV, TjK); Chris@42: ri[WS(rs, 40)] = FNMS(KP707106781, TjV, TjK); Chris@42: Tkm = TjX + TjY; Chris@42: TjZ = TjX - TjY; Chris@42: } Chris@42: ii[WS(rs, 40)] = FNMS(KP707106781, Tkm, Tkl); Chris@42: ii[WS(rs, 8)] = FMA(KP707106781, Tkm, Tkl); Chris@42: ri[WS(rs, 24)] = FMA(KP707106781, TjZ, TjW); Chris@42: ri[WS(rs, 56)] = FNMS(KP707106781, TjZ, TjW); Chris@42: ii[WS(rs, 56)] = FNMS(KP707106781, Tko, Tkn); Chris@42: ii[WS(rs, 24)] = FMA(KP707106781, Tko, Tkn); Chris@42: ri[WS(rs, 16)] = Tk0 + Tk3; Chris@42: ri[WS(rs, 48)] = Tk0 - Tk3; Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tjq, TiM, Tkx, Tkr, Tjt, Tky, Tks, TiX, Tjz, Tje, Tjx, TjD, Tjn, Tj9, Tjr; Chris@42: E TiR; Chris@42: ii[WS(rs, 48)] = Tki - Tkh; Chris@42: ii[WS(rs, 16)] = Tkh + Tki; Chris@42: Tjq = TiI + TiL; Chris@42: TiM = TiI - TiL; Chris@42: Tkx = Tkq - Tkp; Chris@42: Tkr = Tkp + Tkq; Chris@42: Tjr = TiQ + TiP; Chris@42: TiR = TiP - TiQ; Chris@42: { Chris@42: E Tjw, Tj3, Tjs, TiW, Tjv, Tj8; Chris@42: Tjs = TiS - TiV; Chris@42: TiW = TiS + TiV; Chris@42: Tjw = Tj1 + Tj2; Chris@42: Tj3 = Tj1 - Tj2; Chris@42: Tjt = Tjr + Tjs; Chris@42: Tky = Tjs - Tjr; Chris@42: Tks = TiR + TiW; Chris@42: TiX = TiR - TiW; Chris@42: Tjv = Tj4 + Tj7; Chris@42: Tj8 = Tj4 - Tj7; Chris@42: Tjz = Tjc + Tjd; Chris@42: Tje = Tjc - Tjd; Chris@42: Tjx = FMA(KP414213562, Tjw, Tjv); Chris@42: TjD = FNMS(KP414213562, Tjv, Tjw); Chris@42: Tjn = FNMS(KP414213562, Tj3, Tj8); Chris@42: Tj9 = FMA(KP414213562, Tj8, Tj3); Chris@42: } Chris@42: { Chris@42: E Tjm, TiY, Tkz, TkB, Tjy, Tjj; Chris@42: Tjm = FNMS(KP707106781, TiX, TiM); Chris@42: TiY = FMA(KP707106781, TiX, TiM); Chris@42: Tkz = FMA(KP707106781, Tky, Tkx); Chris@42: TkB = FNMS(KP707106781, Tky, Tkx); Chris@42: Tjy = Tjf + Tji; Chris@42: Tjj = Tjf - Tji; Chris@42: { Chris@42: E TjC, Tkt, Tku, TjF; Chris@42: { Chris@42: E Tju, TjE, Tjo, Tjk, TjB, TjA; Chris@42: TjC = FNMS(KP707106781, Tjt, Tjq); Chris@42: Tju = FMA(KP707106781, Tjt, Tjq); Chris@42: TjA = FNMS(KP414213562, Tjz, Tjy); Chris@42: TjE = FMA(KP414213562, Tjy, Tjz); Chris@42: Tjo = FMA(KP414213562, Tje, Tjj); Chris@42: Tjk = FNMS(KP414213562, Tjj, Tje); Chris@42: TjB = Tjx + TjA; Chris@42: Tkw = TjA - Tjx; Chris@42: Tkv = FNMS(KP707106781, Tks, Tkr); Chris@42: Tkt = FMA(KP707106781, Tks, Tkr); Chris@42: { Chris@42: E Tjp, TkA, TkC, Tjl; Chris@42: Tjp = Tjn + Tjo; Chris@42: TkA = Tjo - Tjn; Chris@42: TkC = Tj9 + Tjk; Chris@42: Tjl = Tj9 - Tjk; Chris@42: ri[WS(rs, 4)] = FMA(KP923879532, TjB, Tju); Chris@42: ri[WS(rs, 36)] = FNMS(KP923879532, TjB, Tju); Chris@42: ri[WS(rs, 60)] = FMA(KP923879532, Tjp, Tjm); Chris@42: ri[WS(rs, 28)] = FNMS(KP923879532, Tjp, Tjm); Chris@42: ii[WS(rs, 44)] = FNMS(KP923879532, TkA, Tkz); Chris@42: ii[WS(rs, 12)] = FMA(KP923879532, TkA, Tkz); Chris@42: ii[WS(rs, 60)] = FMA(KP923879532, TkC, TkB); Chris@42: ii[WS(rs, 28)] = FNMS(KP923879532, TkC, TkB); Chris@42: ri[WS(rs, 12)] = FMA(KP923879532, Tjl, TiY); Chris@42: ri[WS(rs, 44)] = FNMS(KP923879532, Tjl, TiY); Chris@42: Tku = TjD + TjE; Chris@42: TjF = TjD - TjE; Chris@42: } Chris@42: } Chris@42: ii[WS(rs, 36)] = FNMS(KP923879532, Tku, Tkt); Chris@42: ii[WS(rs, 4)] = FMA(KP923879532, Tku, Tkt); Chris@42: ri[WS(rs, 20)] = FMA(KP923879532, TjF, TjC); Chris@42: ri[WS(rs, 52)] = FNMS(KP923879532, TjF, TjC); Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E TkV, Tl1, ThG, Tgk, TkH, TkN, Tis, Ti0, Thv, ThJ, TkO, TkI, TgH, Thy, TiC; Chris@42: E TiG, Tiq, Tim, ThN, ThT, ThD, Th9, TkW, Tiv, Tl2, Ti7, ThP, Thq, Tiz, TiF; Chris@42: E Tip, Tif; Chris@42: { Chris@42: E Ti1, Ti2, Ti4, Ti5, Thp, The, Tij, TiB, Tii, Tik; Chris@42: { Chris@42: E ThW, Tg8, TkT, TkF, ThX, ThY, TkU, Tgj, Tgd, Tgi; Chris@42: ThW = Tg4 - Tg7; Chris@42: Tg8 = Tg4 + Tg7; Chris@42: TkT = TkE - TkD; Chris@42: TkF = TkD + TkE; Chris@42: ThX = Tgc - Tg9; Chris@42: Tgd = Tg9 + Tgc; Chris@42: ii[WS(rs, 52)] = FNMS(KP923879532, Tkw, Tkv); Chris@42: ii[WS(rs, 20)] = FMA(KP923879532, Tkw, Tkv); Chris@42: Tgi = Tge - Tgh; Chris@42: ThY = Tge + Tgh; Chris@42: TkU = Tgi - Tgd; Chris@42: Tgj = Tgd + Tgi; Chris@42: { Chris@42: E TgA, ThH, Tgv, TgF; Chris@42: { Chris@42: E Tgp, TkG, ThZ, Tgu; Chris@42: Ti1 = Tgn - Tgo; Chris@42: Tgp = Tgn + Tgo; Chris@42: TkV = FMA(KP707106781, TkU, TkT); Chris@42: Tl1 = FNMS(KP707106781, TkU, TkT); Chris@42: ThG = FMA(KP707106781, Tgj, Tg8); Chris@42: Tgk = FNMS(KP707106781, Tgj, Tg8); Chris@42: TkG = ThX + ThY; Chris@42: ThZ = ThX - ThY; Chris@42: Tgu = Tgq + Tgt; Chris@42: Ti2 = Tgq - Tgt; Chris@42: Ti4 = Tgy - Tgz; Chris@42: TgA = Tgy + Tgz; Chris@42: TkH = FMA(KP707106781, TkG, TkF); Chris@42: TkN = FNMS(KP707106781, TkG, TkF); Chris@42: Tis = FNMS(KP707106781, ThZ, ThW); Chris@42: Ti0 = FMA(KP707106781, ThZ, ThW); Chris@42: ThH = FMA(KP414213562, Tgp, Tgu); Chris@42: Tgv = FNMS(KP414213562, Tgu, Tgp); Chris@42: TgF = TgB + TgE; Chris@42: Ti5 = TgB - TgE; Chris@42: } Chris@42: { Chris@42: E Tig, Tih, ThI, TgG; Chris@42: Thv = Thr + Thu; Chris@42: Tig = Thr - Thu; Chris@42: Tih = Tho - Thj; Chris@42: Thp = Thj + Tho; Chris@42: The = Thc + Thd; Chris@42: Tij = Thc - Thd; Chris@42: ThI = FNMS(KP414213562, TgA, TgF); Chris@42: TgG = FMA(KP414213562, TgF, TgA); Chris@42: TiB = FMA(KP707106781, Tih, Tig); Chris@42: Tii = FNMS(KP707106781, Tih, Tig); Chris@42: ThJ = ThH + ThI; Chris@42: TkO = ThI - ThH; Chris@42: TkI = Tgv + TgG; Chris@42: TgH = Tgv - TgG; Chris@42: Tik = Thw - Thx; Chris@42: Thy = Thw + Thx; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tic, Tia, Ti9, Tid, Tit, Ti3; Chris@42: { Chris@42: E Th4, ThM, TgZ, Th7, ThL, Th8; Chris@42: { Chris@42: E TgN, TgY, TiA, Til; Chris@42: Tic = TgL - TgM; Chris@42: TgN = TgL + TgM; Chris@42: TgY = TgS + TgX; Chris@42: Tia = TgX - TgS; Chris@42: Ti9 = Th0 - Th3; Chris@42: Th4 = Th0 + Th3; Chris@42: TiA = FMA(KP707106781, Tik, Tij); Chris@42: Til = FNMS(KP707106781, Tik, Tij); Chris@42: ThM = FMA(KP707106781, TgY, TgN); Chris@42: TgZ = FNMS(KP707106781, TgY, TgN); Chris@42: TiC = FNMS(KP198912367, TiB, TiA); Chris@42: TiG = FMA(KP198912367, TiA, TiB); Chris@42: Tiq = FMA(KP668178637, Tii, Til); Chris@42: Tim = FNMS(KP668178637, Til, Tii); Chris@42: Th7 = Th5 + Th6; Chris@42: Tid = Th5 - Th6; Chris@42: } Chris@42: ThL = FMA(KP707106781, Th7, Th4); Chris@42: Th8 = FNMS(KP707106781, Th7, Th4); Chris@42: Tit = FNMS(KP414213562, Ti1, Ti2); Chris@42: Ti3 = FMA(KP414213562, Ti2, Ti1); Chris@42: ThN = FMA(KP198912367, ThM, ThL); Chris@42: ThT = FNMS(KP198912367, ThL, ThM); Chris@42: ThD = FNMS(KP668178637, TgZ, Th8); Chris@42: Th9 = FMA(KP668178637, Th8, TgZ); Chris@42: } Chris@42: { Chris@42: E Tiy, Tib, Tiu, Ti6, Tix, Tie; Chris@42: Tiu = FMA(KP414213562, Ti4, Ti5); Chris@42: Ti6 = FNMS(KP414213562, Ti5, Ti4); Chris@42: Tiy = FMA(KP707106781, Tia, Ti9); Chris@42: Tib = FNMS(KP707106781, Tia, Ti9); Chris@42: TkW = Tiu - Tit; Chris@42: Tiv = Tit + Tiu; Chris@42: Tl2 = Ti3 + Ti6; Chris@42: Ti7 = Ti3 - Ti6; Chris@42: Tix = FMA(KP707106781, Tid, Tic); Chris@42: Tie = FNMS(KP707106781, Tid, Tic); Chris@42: ThP = FMA(KP707106781, Thp, The); Chris@42: Thq = FNMS(KP707106781, Thp, The); Chris@42: Tiz = FMA(KP198912367, Tiy, Tix); Chris@42: TiF = FNMS(KP198912367, Tix, Tiy); Chris@42: Tip = FNMS(KP668178637, Tib, Tie); Chris@42: Tif = FMA(KP668178637, Tie, Tib); Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E TkM, TkL, Tl0, TkZ; Chris@42: { Chris@42: E ThC, TgI, TkP, TkR, ThO, Thz; Chris@42: ThC = FNMS(KP923879532, TgH, Tgk); Chris@42: TgI = FMA(KP923879532, TgH, Tgk); Chris@42: TkP = FMA(KP923879532, TkO, TkN); Chris@42: TkR = FNMS(KP923879532, TkO, TkN); Chris@42: ThO = FMA(KP707106781, Thy, Thv); Chris@42: Thz = FNMS(KP707106781, Thy, Thv); Chris@42: { Chris@42: E ThS, TkJ, TkK, ThV; Chris@42: { Chris@42: E ThK, ThU, ThE, ThA, ThR, ThQ; Chris@42: ThS = FNMS(KP923879532, ThJ, ThG); Chris@42: ThK = FMA(KP923879532, ThJ, ThG); Chris@42: ThQ = FNMS(KP198912367, ThP, ThO); Chris@42: ThU = FMA(KP198912367, ThO, ThP); Chris@42: ThE = FMA(KP668178637, Thq, Thz); Chris@42: ThA = FNMS(KP668178637, Thz, Thq); Chris@42: ThR = ThN + ThQ; Chris@42: TkM = ThQ - ThN; Chris@42: TkL = FNMS(KP923879532, TkI, TkH); Chris@42: TkJ = FMA(KP923879532, TkI, TkH); Chris@42: { Chris@42: E ThF, TkQ, TkS, ThB; Chris@42: ThF = ThD + ThE; Chris@42: TkQ = ThE - ThD; Chris@42: TkS = Th9 + ThA; Chris@42: ThB = Th9 - ThA; Chris@42: ri[WS(rs, 2)] = FMA(KP980785280, ThR, ThK); Chris@42: ri[WS(rs, 34)] = FNMS(KP980785280, ThR, ThK); Chris@42: ri[WS(rs, 58)] = FMA(KP831469612, ThF, ThC); Chris@42: ri[WS(rs, 26)] = FNMS(KP831469612, ThF, ThC); Chris@42: ii[WS(rs, 42)] = FNMS(KP831469612, TkQ, TkP); Chris@42: ii[WS(rs, 10)] = FMA(KP831469612, TkQ, TkP); Chris@42: ii[WS(rs, 58)] = FMA(KP831469612, TkS, TkR); Chris@42: ii[WS(rs, 26)] = FNMS(KP831469612, TkS, TkR); Chris@42: ri[WS(rs, 10)] = FMA(KP831469612, ThB, TgI); Chris@42: ri[WS(rs, 42)] = FNMS(KP831469612, ThB, TgI); Chris@42: TkK = ThT + ThU; Chris@42: ThV = ThT - ThU; Chris@42: } Chris@42: } Chris@42: ii[WS(rs, 34)] = FNMS(KP980785280, TkK, TkJ); Chris@42: ii[WS(rs, 2)] = FMA(KP980785280, TkK, TkJ); Chris@42: ri[WS(rs, 18)] = FMA(KP980785280, ThV, ThS); Chris@42: ri[WS(rs, 50)] = FNMS(KP980785280, ThV, ThS); Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tio, TkX, TkY, Tir, Ti8, Tin; Chris@42: Tio = FNMS(KP923879532, Ti7, Ti0); Chris@42: Ti8 = FMA(KP923879532, Ti7, Ti0); Chris@42: Tin = Tif + Tim; Chris@42: Tl0 = Tim - Tif; Chris@42: TkZ = FNMS(KP923879532, TkW, TkV); Chris@42: TkX = FMA(KP923879532, TkW, TkV); Chris@42: ii[WS(rs, 50)] = FNMS(KP980785280, TkM, TkL); Chris@42: ii[WS(rs, 18)] = FMA(KP980785280, TkM, TkL); Chris@42: ri[WS(rs, 6)] = FMA(KP831469612, Tin, Ti8); Chris@42: ri[WS(rs, 38)] = FNMS(KP831469612, Tin, Ti8); Chris@42: TkY = Tip + Tiq; Chris@42: Tir = Tip - Tiq; Chris@42: ii[WS(rs, 38)] = FNMS(KP831469612, TkY, TkX); Chris@42: ii[WS(rs, 6)] = FMA(KP831469612, TkY, TkX); Chris@42: ri[WS(rs, 22)] = FMA(KP831469612, Tir, Tio); Chris@42: ri[WS(rs, 54)] = FNMS(KP831469612, Tir, Tio); Chris@42: } Chris@42: { Chris@42: E TiE, Tl3, Tl4, TiH, Tiw, TiD; Chris@42: TiE = FMA(KP923879532, Tiv, Tis); Chris@42: Tiw = FNMS(KP923879532, Tiv, Tis); Chris@42: TiD = Tiz - TiC; Chris@42: Tl6 = Tiz + TiC; Chris@42: Tl5 = FMA(KP923879532, Tl2, Tl1); Chris@42: Tl3 = FNMS(KP923879532, Tl2, Tl1); Chris@42: ii[WS(rs, 54)] = FNMS(KP831469612, Tl0, TkZ); Chris@42: ii[WS(rs, 22)] = FMA(KP831469612, Tl0, TkZ); Chris@42: ri[WS(rs, 14)] = FMA(KP980785280, TiD, Tiw); Chris@42: ri[WS(rs, 46)] = FNMS(KP980785280, TiD, Tiw); Chris@42: Tl4 = TiG - TiF; Chris@42: TiH = TiF + TiG; Chris@42: ii[WS(rs, 46)] = FNMS(KP980785280, Tl4, Tl3); Chris@42: ii[WS(rs, 14)] = FMA(KP980785280, Tl4, Tl3); Chris@42: ri[WS(rs, 62)] = FMA(KP980785280, TiH, TiE); Chris@42: ri[WS(rs, 30)] = FNMS(KP980785280, TiH, TiE); Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tla, TdV, TdO, Tm6, Tm5, TdR; Chris@42: { Chris@42: E TcT, TlO, TlI, Tar, TcX, Td3, TcN, TbB, TdM, TdQ, TdA, Tdw, TdJ, TdP, Tdz; Chris@42: E Tdp, TlW, TdF, Tm2, Tdh, Td7, T91, Td6, T8M, TlT, TlF, Td0, Td4, TcO, TcK; Chris@42: E T9g, Td8; Chris@42: { Chris@42: E Tdb, Tdc, Tde, Tdf, Tdm, Tdk, Tdj, Tdn, TcF, Tct, TbY, Tdt, TdL, Tds, Tdu; Chris@42: E TcI, TdD, Tdd; Chris@42: { Chris@42: E Tae, TcR, T9R, Tap, T9F, T9Q; Chris@42: Tdb = FMA(KP707106781, T9E, T9p); Chris@42: T9F = FNMS(KP707106781, T9E, T9p); Chris@42: T9Q = FNMS(KP707106781, T9P, T9M); Chris@42: Tdc = FMA(KP707106781, T9P, T9M); Chris@42: Tde = FMA(KP707106781, Tad, T9Y); Chris@42: Tae = FNMS(KP707106781, Tad, T9Y); Chris@42: ii[WS(rs, 62)] = FMA(KP980785280, Tl6, Tl5); Chris@42: ii[WS(rs, 30)] = FNMS(KP980785280, Tl6, Tl5); Chris@42: TcR = FMA(KP668178637, T9F, T9Q); Chris@42: T9R = FNMS(KP668178637, T9Q, T9F); Chris@42: Tap = FNMS(KP707106781, Tao, Tal); Chris@42: Tdf = FMA(KP707106781, Tao, Tal); Chris@42: { Chris@42: E Tbw, TcW, Tbl, Tbz; Chris@42: { Chris@42: E TaP, Tbk, TcS, Taq; Chris@42: Tdm = FMA(KP707106781, TaO, Taz); Chris@42: TaP = FNMS(KP707106781, TaO, Taz); Chris@42: Tbk = Tb4 - Tbj; Chris@42: Tdk = Tbj + Tb4; Chris@42: Tdj = FMA(KP707106781, Tbv, Tbs); Chris@42: Tbw = FNMS(KP707106781, Tbv, Tbs); Chris@42: TcS = FNMS(KP668178637, Tae, Tap); Chris@42: Taq = FMA(KP668178637, Tap, Tae); Chris@42: TcW = FMA(KP923879532, Tbk, TaP); Chris@42: Tbl = FNMS(KP923879532, Tbk, TaP); Chris@42: TcT = TcR + TcS; Chris@42: TlO = TcS - TcR; Chris@42: TlI = T9R + Taq; Chris@42: Tar = T9R - Taq; Chris@42: Tbz = Tbx - Tby; Chris@42: Tdn = Tbx + Tby; Chris@42: } Chris@42: { Chris@42: E Tdq, Tdr, TcV, TbA; Chris@42: TcF = FNMS(KP707106781, TcE, TcB); Chris@42: Tdq = FMA(KP707106781, TcE, TcB); Chris@42: Tdr = Tcs + Tcd; Chris@42: Tct = Tcd - Tcs; Chris@42: TbY = FNMS(KP707106781, TbX, TbI); Chris@42: Tdt = FMA(KP707106781, TbX, TbI); Chris@42: TcV = FMA(KP923879532, Tbz, Tbw); Chris@42: TbA = FNMS(KP923879532, Tbz, Tbw); Chris@42: TdL = FMA(KP923879532, Tdr, Tdq); Chris@42: Tds = FNMS(KP923879532, Tdr, Tdq); Chris@42: TcX = FMA(KP303346683, TcW, TcV); Chris@42: Td3 = FNMS(KP303346683, TcV, TcW); Chris@42: TcN = FNMS(KP534511135, Tbl, TbA); Chris@42: TbB = FMA(KP534511135, TbA, Tbl); Chris@42: Tdu = TcG + TcH; Chris@42: TcI = TcG - TcH; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E TdI, Tdl, TdK, Tdv, TdH, Tdo; Chris@42: TdK = FMA(KP923879532, Tdu, Tdt); Chris@42: Tdv = FNMS(KP923879532, Tdu, Tdt); Chris@42: TdI = FMA(KP923879532, Tdk, Tdj); Chris@42: Tdl = FNMS(KP923879532, Tdk, Tdj); Chris@42: TdM = FNMS(KP098491403, TdL, TdK); Chris@42: TdQ = FMA(KP098491403, TdK, TdL); Chris@42: TdA = FMA(KP820678790, Tds, Tdv); Chris@42: Tdw = FNMS(KP820678790, Tdv, Tds); Chris@42: TdH = FMA(KP923879532, Tdn, Tdm); Chris@42: Tdo = FNMS(KP923879532, Tdn, Tdm); Chris@42: TdD = FNMS(KP198912367, Tdb, Tdc); Chris@42: Tdd = FMA(KP198912367, Tdc, Tdb); Chris@42: TdJ = FMA(KP098491403, TdI, TdH); Chris@42: TdP = FNMS(KP098491403, TdH, TdI); Chris@42: Tdz = FNMS(KP820678790, Tdl, Tdo); Chris@42: Tdp = FMA(KP820678790, Tdo, Tdl); Chris@42: } Chris@42: { Chris@42: E TcZ, Tcu, TdE, Tdg; Chris@42: TdE = FMA(KP198912367, Tde, Tdf); Chris@42: Tdg = FNMS(KP198912367, Tdf, Tde); Chris@42: TcZ = FMA(KP923879532, Tct, TbY); Chris@42: Tcu = FNMS(KP923879532, Tct, TbY); Chris@42: TlW = TdE - TdD; Chris@42: TdF = TdD + TdE; Chris@42: Tm2 = Tdd + Tdg; Chris@42: Tdh = Tdd - Tdg; Chris@42: { Chris@42: E T8L, TlE, TcY, TcJ; Chris@42: Tla = T8D + T8K; Chris@42: T8L = T8D - T8K; Chris@42: TlE = TdU - TdT; Chris@42: TdV = TdT + TdU; Chris@42: Td7 = FNMS(KP414213562, T8T, T90); Chris@42: T91 = FMA(KP414213562, T90, T8T); Chris@42: TcY = FMA(KP923879532, TcI, TcF); Chris@42: TcJ = FNMS(KP923879532, TcI, TcF); Chris@42: Td6 = FNMS(KP707106781, T8L, T8w); Chris@42: T8M = FMA(KP707106781, T8L, T8w); Chris@42: TlT = FNMS(KP707106781, TlE, TlD); Chris@42: TlF = FMA(KP707106781, TlE, TlD); Chris@42: Td0 = FNMS(KP303346683, TcZ, TcY); Chris@42: Td4 = FMA(KP303346683, TcY, TcZ); Chris@42: TcO = FMA(KP534511135, Tcu, TcJ); Chris@42: TcK = FNMS(KP534511135, TcJ, Tcu); Chris@42: T9g = FNMS(KP414213562, T9f, T98); Chris@42: Td8 = FMA(KP414213562, T98, T9f); Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tm1, TlV, TdC, Tda, Td2, TlM, TlL, Td5; Chris@42: { Chris@42: E TlS, TcQ, TlH, TcM, TlR, TcP; Chris@42: { Chris@42: E TcL, Tas, TlP, TlQ, TlN; Chris@42: TlS = TbB + TcK; Chris@42: TcL = TbB - TcK; Chris@42: { Chris@42: E TlU, T9h, TlG, Td9, T9i; Chris@42: TlU = T91 + T9g; Chris@42: T9h = T91 - T9g; Chris@42: TlG = Td8 - Td7; Chris@42: Td9 = Td7 + Td8; Chris@42: Tm1 = FMA(KP923879532, TlU, TlT); Chris@42: TlV = FNMS(KP923879532, TlU, TlT); Chris@42: TcQ = FMA(KP923879532, T9h, T8M); Chris@42: T9i = FNMS(KP923879532, T9h, T8M); Chris@42: TlN = FNMS(KP923879532, TlG, TlF); Chris@42: TlH = FMA(KP923879532, TlG, TlF); Chris@42: TdC = FMA(KP923879532, Td9, Td6); Chris@42: Tda = FNMS(KP923879532, Td9, Td6); Chris@42: Tas = FMA(KP831469612, Tar, T9i); Chris@42: TcM = FNMS(KP831469612, Tar, T9i); Chris@42: } Chris@42: TlR = FNMS(KP831469612, TlO, TlN); Chris@42: TlP = FMA(KP831469612, TlO, TlN); Chris@42: TlQ = TcO - TcN; Chris@42: TcP = TcN + TcO; Chris@42: ri[WS(rs, 11)] = FMA(KP881921264, TcL, Tas); Chris@42: ri[WS(rs, 43)] = FNMS(KP881921264, TcL, Tas); Chris@42: ii[WS(rs, 43)] = FNMS(KP881921264, TlQ, TlP); Chris@42: ii[WS(rs, 11)] = FMA(KP881921264, TlQ, TlP); Chris@42: } Chris@42: { Chris@42: E TcU, Td1, TlJ, TlK; Chris@42: Td2 = FNMS(KP831469612, TcT, TcQ); Chris@42: TcU = FMA(KP831469612, TcT, TcQ); Chris@42: ri[WS(rs, 59)] = FMA(KP881921264, TcP, TcM); Chris@42: ri[WS(rs, 27)] = FNMS(KP881921264, TcP, TcM); Chris@42: ii[WS(rs, 59)] = FMA(KP881921264, TlS, TlR); Chris@42: ii[WS(rs, 27)] = FNMS(KP881921264, TlS, TlR); Chris@42: Td1 = TcX + Td0; Chris@42: TlM = Td0 - TcX; Chris@42: TlL = FNMS(KP831469612, TlI, TlH); Chris@42: TlJ = FMA(KP831469612, TlI, TlH); Chris@42: TlK = Td3 + Td4; Chris@42: Td5 = Td3 - Td4; Chris@42: ri[WS(rs, 3)] = FMA(KP956940335, Td1, TcU); Chris@42: ri[WS(rs, 35)] = FNMS(KP956940335, Td1, TcU); Chris@42: ii[WS(rs, 35)] = FNMS(KP956940335, TlK, TlJ); Chris@42: ii[WS(rs, 3)] = FMA(KP956940335, TlK, TlJ); Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tdy, Tm0, TlZ, TdB; Chris@42: { Chris@42: E Tdi, Tdx, TlX, TlY; Chris@42: Tdy = FNMS(KP980785280, Tdh, Tda); Chris@42: Tdi = FMA(KP980785280, Tdh, Tda); Chris@42: ri[WS(rs, 19)] = FMA(KP956940335, Td5, Td2); Chris@42: ri[WS(rs, 51)] = FNMS(KP956940335, Td5, Td2); Chris@42: ii[WS(rs, 51)] = FNMS(KP956940335, TlM, TlL); Chris@42: ii[WS(rs, 19)] = FMA(KP956940335, TlM, TlL); Chris@42: Tdx = Tdp + Tdw; Chris@42: Tm0 = Tdw - Tdp; Chris@42: TlZ = FNMS(KP980785280, TlW, TlV); Chris@42: TlX = FMA(KP980785280, TlW, TlV); Chris@42: TlY = Tdz + TdA; Chris@42: TdB = Tdz - TdA; Chris@42: ri[WS(rs, 7)] = FMA(KP773010453, Tdx, Tdi); Chris@42: ri[WS(rs, 39)] = FNMS(KP773010453, Tdx, Tdi); Chris@42: ii[WS(rs, 39)] = FNMS(KP773010453, TlY, TlX); Chris@42: ii[WS(rs, 7)] = FMA(KP773010453, TlY, TlX); Chris@42: } Chris@42: { Chris@42: E TdG, TdN, Tm3, Tm4; Chris@42: TdO = FMA(KP980785280, TdF, TdC); Chris@42: TdG = FNMS(KP980785280, TdF, TdC); Chris@42: ri[WS(rs, 23)] = FMA(KP773010453, TdB, Tdy); Chris@42: ri[WS(rs, 55)] = FNMS(KP773010453, TdB, Tdy); Chris@42: ii[WS(rs, 55)] = FNMS(KP773010453, Tm0, TlZ); Chris@42: ii[WS(rs, 23)] = FMA(KP773010453, Tm0, TlZ); Chris@42: TdN = TdJ - TdM; Chris@42: Tm6 = TdJ + TdM; Chris@42: Tm5 = FMA(KP980785280, Tm2, Tm1); Chris@42: Tm3 = FNMS(KP980785280, Tm2, Tm1); Chris@42: Tm4 = TdQ - TdP; Chris@42: TdR = TdP + TdQ; Chris@42: ri[WS(rs, 15)] = FMA(KP995184726, TdN, TdG); Chris@42: ri[WS(rs, 47)] = FNMS(KP995184726, TdN, TdG); Chris@42: ii[WS(rs, 47)] = FNMS(KP995184726, Tm4, Tm3); Chris@42: ii[WS(rs, 15)] = FMA(KP995184726, Tm4, Tm3); Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tf5, Tlk, Tle, Tej, Tf9, Tff, TeZ, TeD, TfY, Tg2, TfM, TfI, TfV, Tg1, TfL; Chris@42: E TfB, Tls, TfR, Tly, Tft, Tfj, TdZ, Tfi, TdW, Tlp, Tlb, Tfc, Tfg, Tf0, TeW; Chris@42: E Te2, Tfk; Chris@42: { Chris@42: E Tfn, Tfo, Tfq, Tfr, Tfy, Tfw, Tfv, Tfz, TeR, TeN, TeG, TfF, TfX, TfE, TfG; Chris@42: E TeU, TfP, Tfp; Chris@42: { Chris@42: E Te7, Tea, Tee, Teh; Chris@42: Tfn = FNMS(KP707106781, Te6, Te5); Chris@42: Te7 = FMA(KP707106781, Te6, Te5); Chris@42: ri[WS(rs, 63)] = FMA(KP995184726, TdR, TdO); Chris@42: ri[WS(rs, 31)] = FNMS(KP995184726, TdR, TdO); Chris@42: ii[WS(rs, 63)] = FMA(KP995184726, Tm6, Tm5); Chris@42: ii[WS(rs, 31)] = FNMS(KP995184726, Tm6, Tm5); Chris@42: Tea = FMA(KP707106781, Te9, Te8); Chris@42: Tfo = FNMS(KP707106781, Te9, Te8); Chris@42: Tfq = FNMS(KP707106781, Ted, Tec); Chris@42: Tee = FMA(KP707106781, Ted, Tec); Chris@42: Teh = FMA(KP707106781, Teg, Tef); Chris@42: Tfr = FNMS(KP707106781, Teg, Tef); Chris@42: { Chris@42: E Tey, Tf8, Tev, TeB; Chris@42: { Chris@42: E Ten, Tf3, Teb, Tf4, Tei, Teu; Chris@42: Tfy = FNMS(KP707106781, Tem, Tel); Chris@42: Ten = FMA(KP707106781, Tem, Tel); Chris@42: Tf3 = FMA(KP198912367, Te7, Tea); Chris@42: Teb = FNMS(KP198912367, Tea, Te7); Chris@42: Tf4 = FNMS(KP198912367, Tee, Teh); Chris@42: Tei = FMA(KP198912367, Teh, Tee); Chris@42: Teu = Teq + Tet; Chris@42: Tfw = Tet - Teq; Chris@42: Tfv = FNMS(KP707106781, Tex, Tew); Chris@42: Tey = FMA(KP707106781, Tex, Tew); Chris@42: Tf5 = Tf3 + Tf4; Chris@42: Tlk = Tf4 - Tf3; Chris@42: Tle = Teb + Tei; Chris@42: Tej = Teb - Tei; Chris@42: Tf8 = FMA(KP923879532, Teu, Ten); Chris@42: Tev = FNMS(KP923879532, Teu, Ten); Chris@42: TeB = Tez + TeA; Chris@42: Tfz = Tez - TeA; Chris@42: } Chris@42: { Chris@42: E TfC, TfD, Tf7, TeC; Chris@42: TeR = FMA(KP707106781, TeQ, TeP); Chris@42: TfC = FNMS(KP707106781, TeQ, TeP); Chris@42: TfD = TeM - TeJ; Chris@42: TeN = TeJ + TeM; Chris@42: TeG = FMA(KP707106781, TeF, TeE); Chris@42: TfF = FNMS(KP707106781, TeF, TeE); Chris@42: Tf7 = FMA(KP923879532, TeB, Tey); Chris@42: TeC = FNMS(KP923879532, TeB, Tey); Chris@42: TfX = FMA(KP923879532, TfD, TfC); Chris@42: TfE = FNMS(KP923879532, TfD, TfC); Chris@42: Tf9 = FMA(KP098491403, Tf8, Tf7); Chris@42: Tff = FNMS(KP098491403, Tf7, Tf8); Chris@42: TeZ = FNMS(KP820678790, Tev, TeC); Chris@42: TeD = FMA(KP820678790, TeC, Tev); Chris@42: TfG = TeS - TeT; Chris@42: TeU = TeS + TeT; Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E TfU, Tfx, TfW, TfH, TfT, TfA; Chris@42: TfW = FMA(KP923879532, TfG, TfF); Chris@42: TfH = FNMS(KP923879532, TfG, TfF); Chris@42: TfU = FMA(KP923879532, Tfw, Tfv); Chris@42: Tfx = FNMS(KP923879532, Tfw, Tfv); Chris@42: TfY = FNMS(KP303346683, TfX, TfW); Chris@42: Tg2 = FMA(KP303346683, TfW, TfX); Chris@42: TfM = FMA(KP534511135, TfE, TfH); Chris@42: TfI = FNMS(KP534511135, TfH, TfE); Chris@42: TfT = FMA(KP923879532, Tfz, Tfy); Chris@42: TfA = FNMS(KP923879532, Tfz, Tfy); Chris@42: TfP = FNMS(KP668178637, Tfn, Tfo); Chris@42: Tfp = FMA(KP668178637, Tfo, Tfn); Chris@42: TfV = FMA(KP303346683, TfU, TfT); Chris@42: Tg1 = FNMS(KP303346683, TfT, TfU); Chris@42: TfL = FNMS(KP534511135, Tfx, TfA); Chris@42: TfB = FMA(KP534511135, TfA, Tfx); Chris@42: } Chris@42: { Chris@42: E Tfb, TeO, TfQ, Tfs, Tfa, TeV; Chris@42: TfQ = FMA(KP668178637, Tfq, Tfr); Chris@42: Tfs = FNMS(KP668178637, Tfr, Tfq); Chris@42: Tfb = FMA(KP923879532, TeN, TeG); Chris@42: TeO = FNMS(KP923879532, TeN, TeG); Chris@42: Tls = TfQ - TfP; Chris@42: TfR = TfP + TfQ; Chris@42: Tly = Tfp + Tfs; Chris@42: Tft = Tfp - Tfs; Chris@42: Tfj = FNMS(KP414213562, TdX, TdY); Chris@42: TdZ = FMA(KP414213562, TdY, TdX); Chris@42: Tfa = FMA(KP923879532, TeU, TeR); Chris@42: TeV = FNMS(KP923879532, TeU, TeR); Chris@42: Tfi = FNMS(KP707106781, TdV, TdS); Chris@42: TdW = FMA(KP707106781, TdV, TdS); Chris@42: Tlp = FNMS(KP707106781, Tla, Tl9); Chris@42: Tlb = FMA(KP707106781, Tla, Tl9); Chris@42: Tfc = FNMS(KP098491403, Tfb, Tfa); Chris@42: Tfg = FMA(KP098491403, Tfa, Tfb); Chris@42: Tf0 = FMA(KP820678790, TeO, TeV); Chris@42: TeW = FNMS(KP820678790, TeV, TeO); Chris@42: Te2 = FNMS(KP414213562, Te1, Te0); Chris@42: Tfk = FMA(KP414213562, Te0, Te1); Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tlx, Tlr, TfO, Tfm, Tfe, Tli, Tlh, Tfh; Chris@42: { Chris@42: E Tlo, Tf2, Tld, TeY, Tln, Tf1; Chris@42: { Chris@42: E TeX, Tek, Tll, Tlm, Tlj; Chris@42: Tlo = TeD + TeW; Chris@42: TeX = TeD - TeW; Chris@42: { Chris@42: E Tlq, Te3, Tlc, Tfl, Te4; Chris@42: Tlq = Te2 - TdZ; Chris@42: Te3 = TdZ + Te2; Chris@42: Tlc = Tfj + Tfk; Chris@42: Tfl = Tfj - Tfk; Chris@42: Tlx = FNMS(KP923879532, Tlq, Tlp); Chris@42: Tlr = FMA(KP923879532, Tlq, Tlp); Chris@42: Tf2 = FMA(KP923879532, Te3, TdW); Chris@42: Te4 = FNMS(KP923879532, Te3, TdW); Chris@42: Tlj = FNMS(KP923879532, Tlc, Tlb); Chris@42: Tld = FMA(KP923879532, Tlc, Tlb); Chris@42: TfO = FNMS(KP923879532, Tfl, Tfi); Chris@42: Tfm = FMA(KP923879532, Tfl, Tfi); Chris@42: Tek = FMA(KP980785280, Tej, Te4); Chris@42: TeY = FNMS(KP980785280, Tej, Te4); Chris@42: } Chris@42: Tln = FNMS(KP980785280, Tlk, Tlj); Chris@42: Tll = FMA(KP980785280, Tlk, Tlj); Chris@42: Tlm = Tf0 - TeZ; Chris@42: Tf1 = TeZ + Tf0; Chris@42: ri[WS(rs, 9)] = FMA(KP773010453, TeX, Tek); Chris@42: ri[WS(rs, 41)] = FNMS(KP773010453, TeX, Tek); Chris@42: ii[WS(rs, 41)] = FNMS(KP773010453, Tlm, Tll); Chris@42: ii[WS(rs, 9)] = FMA(KP773010453, Tlm, Tll); Chris@42: } Chris@42: { Chris@42: E Tf6, Tfd, Tlf, Tlg; Chris@42: Tfe = FNMS(KP980785280, Tf5, Tf2); Chris@42: Tf6 = FMA(KP980785280, Tf5, Tf2); Chris@42: ri[WS(rs, 57)] = FMA(KP773010453, Tf1, TeY); Chris@42: ri[WS(rs, 25)] = FNMS(KP773010453, Tf1, TeY); Chris@42: ii[WS(rs, 57)] = FMA(KP773010453, Tlo, Tln); Chris@42: ii[WS(rs, 25)] = FNMS(KP773010453, Tlo, Tln); Chris@42: Tfd = Tf9 + Tfc; Chris@42: Tli = Tfc - Tf9; Chris@42: Tlh = FNMS(KP980785280, Tle, Tld); Chris@42: Tlf = FMA(KP980785280, Tle, Tld); Chris@42: Tlg = Tff + Tfg; Chris@42: Tfh = Tff - Tfg; Chris@42: ri[WS(rs, 1)] = FMA(KP995184726, Tfd, Tf6); Chris@42: ri[WS(rs, 33)] = FNMS(KP995184726, Tfd, Tf6); Chris@42: ii[WS(rs, 33)] = FNMS(KP995184726, Tlg, Tlf); Chris@42: ii[WS(rs, 1)] = FMA(KP995184726, Tlg, Tlf); Chris@42: } Chris@42: } Chris@42: { Chris@42: E TfK, Tlw, Tlv, TfN; Chris@42: { Chris@42: E Tfu, TfJ, Tlt, Tlu; Chris@42: TfK = FNMS(KP831469612, Tft, Tfm); Chris@42: Tfu = FMA(KP831469612, Tft, Tfm); Chris@42: ri[WS(rs, 17)] = FMA(KP995184726, Tfh, Tfe); Chris@42: ri[WS(rs, 49)] = FNMS(KP995184726, Tfh, Tfe); Chris@42: ii[WS(rs, 49)] = FNMS(KP995184726, Tli, Tlh); Chris@42: ii[WS(rs, 17)] = FMA(KP995184726, Tli, Tlh); Chris@42: TfJ = TfB + TfI; Chris@42: Tlw = TfI - TfB; Chris@42: Tlv = FNMS(KP831469612, Tls, Tlr); Chris@42: Tlt = FMA(KP831469612, Tls, Tlr); Chris@42: Tlu = TfL + TfM; Chris@42: TfN = TfL - TfM; Chris@42: ri[WS(rs, 5)] = FMA(KP881921264, TfJ, Tfu); Chris@42: ri[WS(rs, 37)] = FNMS(KP881921264, TfJ, Tfu); Chris@42: ii[WS(rs, 37)] = FNMS(KP881921264, Tlu, Tlt); Chris@42: ii[WS(rs, 5)] = FMA(KP881921264, Tlu, Tlt); Chris@42: } Chris@42: { Chris@42: E TfS, TfZ, Tlz, TlA; Chris@42: Tg0 = FMA(KP831469612, TfR, TfO); Chris@42: TfS = FNMS(KP831469612, TfR, TfO); Chris@42: ri[WS(rs, 21)] = FMA(KP881921264, TfN, TfK); Chris@42: ri[WS(rs, 53)] = FNMS(KP881921264, TfN, TfK); Chris@42: ii[WS(rs, 53)] = FNMS(KP881921264, Tlw, Tlv); Chris@42: ii[WS(rs, 21)] = FMA(KP881921264, Tlw, Tlv); Chris@42: TfZ = TfV - TfY; Chris@42: TlC = TfV + TfY; Chris@42: TlB = FMA(KP831469612, Tly, Tlx); Chris@42: Tlz = FNMS(KP831469612, Tly, Tlx); Chris@42: TlA = Tg2 - Tg1; Chris@42: Tg3 = Tg1 + Tg2; Chris@42: ri[WS(rs, 13)] = FMA(KP956940335, TfZ, TfS); Chris@42: ri[WS(rs, 45)] = FNMS(KP956940335, TfZ, TfS); Chris@42: ii[WS(rs, 45)] = FNMS(KP956940335, TlA, Tlz); Chris@42: ii[WS(rs, 13)] = FMA(KP956940335, TlA, Tlz); Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: ri[WS(rs, 61)] = FMA(KP956940335, Tg3, Tg0); Chris@42: ri[WS(rs, 29)] = FNMS(KP956940335, Tg3, Tg0); Chris@42: ii[WS(rs, 61)] = FMA(KP956940335, TlC, TlB); Chris@42: ii[WS(rs, 29)] = FNMS(KP956940335, TlC, TlB); Chris@42: } Chris@42: } Chris@42: } Chris@42: Chris@42: static const tw_instr twinstr[] = { Chris@42: {TW_CEXP, 0, 1}, Chris@42: {TW_CEXP, 0, 3}, Chris@42: {TW_CEXP, 0, 9}, Chris@42: {TW_CEXP, 0, 27}, Chris@42: {TW_CEXP, 0, 63}, Chris@42: {TW_NEXT, 1, 0} Chris@42: }; Chris@42: Chris@42: static const ct_desc desc = { 64, "t2_64", twinstr, &GENUS, {520, 206, 634, 0}, 0, 0, 0 }; Chris@42: Chris@42: void X(codelet_t2_64) (planner *p) { Chris@42: X(kdft_dit_register) (p, t2_64, &desc); Chris@42: } Chris@42: #else /* HAVE_FMA */ Chris@42: Chris@42: /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 64 -name t2_64 -include t.h */ Chris@42: Chris@42: /* Chris@42: * This function contains 1154 FP additions, 660 FP multiplications, Chris@42: * (or, 880 additions, 386 multiplications, 274 fused multiply/add), Chris@42: * 302 stack variables, 15 constants, and 256 memory accesses Chris@42: */ Chris@42: #include "t.h" Chris@42: Chris@42: static void t2_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) Chris@42: { Chris@42: DK(KP471396736, +0.471396736825997648556387625905254377657460319); Chris@42: DK(KP881921264, +0.881921264348355029712756863660388349508442621); Chris@42: DK(KP290284677, +0.290284677254462367636192375817395274691476278); Chris@42: DK(KP956940335, +0.956940335732208864935797886980269969482849206); Chris@42: DK(KP634393284, +0.634393284163645498215171613225493370675687095); Chris@42: DK(KP773010453, +0.773010453362736960810906609758469800971041293); Chris@42: DK(KP098017140, +0.098017140329560601994195563888641845861136673); Chris@42: DK(KP995184726, +0.995184726672196886244836953109479921575474869); Chris@42: DK(KP555570233, +0.555570233019602224742830813948532874374937191); Chris@42: DK(KP831469612, +0.831469612302545237078788377617905756738560812); Chris@42: DK(KP980785280, +0.980785280403230449126182236134239036973933731); Chris@42: DK(KP195090322, +0.195090322016128267848284868477022240927691618); Chris@42: DK(KP923879532, +0.923879532511286756128183189396788286822416626); Chris@42: DK(KP382683432, +0.382683432365089771728459984030398866761344562); Chris@42: DK(KP707106781, +0.707106781186547524400844362104849039284835938); Chris@42: { Chris@42: INT m; Chris@42: for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(128, rs)) { Chris@42: E T2, T5, T3, T6, Te, T9, TP, T3e, T1e, T39, T3c, TT, T1a, T37, T8; Chris@42: E Tw, Td, Ty, Tm, Th, T1C, T3K, T1V, T3x, T3I, T1G, T1R, T3v, T2m, T2q; Chris@42: E T5Y, T6u, T53, T5B, T62, T6w, T57, T5D, T2V, T2X, Tg, TE, T3Y, T3V, T3j; Chris@42: E Tl, TA, T3g, T1j, T1t, TV, T2C, T2z, T1u, TZ, T1h, To, T1p, T6j, T6H; Chris@42: E Ts, T1l, T6l, T6F, T2P, T4b, T4x, T5i, T2R, T49, T4z, T5g, TG, T4k, T4m; Chris@42: E TK, T21, T3O, T3Q, T25, TW, T10, T11, T79, T6X, T5M, T6b, T1v, T30, T69; Chris@42: E T77, T13, T2F, T2D, T6p, T6O, T1x, T2a, T2f, T6V, T28, T6r, T2h, T6Q, T32; Chris@42: E T5K, T5w, T4G, T4Q, T3m, T4h, T4I, T5y, T3k, T4f, T41, T4S, T4Y, T3q, T3D; Chris@42: E T3F, T5r, T3s, T4W, T3Z, T5p; Chris@42: { Chris@42: E Ta, Tj, Tx, TC, Tf, Tk, Tz, TD, T1B, T1E, T2o, T2l, T1T, T1Q, T1A; Chris@42: E T1F, T2p, T2k, T1U, T1P; Chris@42: { Chris@42: E T4, T1d, T19, Tb, T1c, T7, Tc, T18, TR, TO, TS, TN; Chris@42: T2 = W[0]; Chris@42: T5 = W[1]; Chris@42: T3 = W[2]; Chris@42: T6 = W[3]; Chris@42: Te = W[5]; Chris@42: T9 = W[4]; Chris@42: T4 = T2 * T3; Chris@42: T1d = T5 * T9; Chris@42: T19 = T5 * Te; Chris@42: Tb = T2 * T6; Chris@42: T1c = T2 * Te; Chris@42: T7 = T5 * T6; Chris@42: Tc = T5 * T3; Chris@42: T18 = T2 * T9; Chris@42: TR = T3 * Te; Chris@42: TO = T6 * Te; Chris@42: TS = T6 * T9; Chris@42: TN = T3 * T9; Chris@42: TP = TN - TO; Chris@42: T3e = TR - TS; Chris@42: T1e = T1c - T1d; Chris@42: T39 = T1c + T1d; Chris@42: T3c = TN + TO; Chris@42: TT = TR + TS; Chris@42: T1a = T18 + T19; Chris@42: T37 = T18 - T19; Chris@42: T8 = T4 - T7; Chris@42: Ta = T8 * T9; Chris@42: Tj = T8 * Te; Chris@42: Tw = T4 + T7; Chris@42: Tx = Tw * T9; Chris@42: TC = Tw * Te; Chris@42: Td = Tb + Tc; Chris@42: Tf = Td * Te; Chris@42: Tk = Td * T9; Chris@42: Ty = Tb - Tc; Chris@42: Tz = Ty * Te; Chris@42: TD = Ty * T9; Chris@42: Tm = W[7]; Chris@42: T1B = T6 * Tm; Chris@42: T1E = T3 * Tm; Chris@42: T2o = T2 * Tm; Chris@42: T2l = T5 * Tm; Chris@42: T1T = T9 * Tm; Chris@42: T1Q = Te * Tm; Chris@42: Th = W[6]; Chris@42: T1A = T3 * Th; Chris@42: T1F = T6 * Th; Chris@42: T2p = T5 * Th; Chris@42: T2k = T2 * Th; Chris@42: T1U = Te * Th; Chris@42: T1P = T9 * Th; Chris@42: } Chris@42: T1C = T1A + T1B; Chris@42: T3K = T1E + T1F; Chris@42: T1V = T1T + T1U; Chris@42: T3x = T2o - T2p; Chris@42: T3I = T1A - T1B; Chris@42: T1G = T1E - T1F; Chris@42: T1R = T1P - T1Q; Chris@42: { Chris@42: E T5W, T5X, T55, T56; Chris@42: T3v = T2k + T2l; Chris@42: T2m = T2k - T2l; Chris@42: T2q = T2o + T2p; Chris@42: T5W = T8 * Th; Chris@42: T5X = Td * Tm; Chris@42: T5Y = T5W - T5X; Chris@42: T6u = T5W + T5X; Chris@42: { Chris@42: E T51, T52, T60, T61; Chris@42: T51 = Tw * Th; Chris@42: T52 = Ty * Tm; Chris@42: T53 = T51 + T52; Chris@42: T5B = T51 - T52; Chris@42: T60 = T8 * Tm; Chris@42: T61 = Td * Th; Chris@42: T62 = T60 + T61; Chris@42: T6w = T60 - T61; Chris@42: } Chris@42: T55 = Tw * Tm; Chris@42: T56 = Ty * Th; Chris@42: T57 = T55 - T56; Chris@42: T5D = T55 + T56; Chris@42: { Chris@42: E Ti, Tq, TF, TJ, T3W, T3X, T3T, T3U, T3h, T3i, Tn, Tr, TB, TI, T3d; Chris@42: E T3f, T1k, T1o, T1Z, T23, TQ, TU, T2A, T2B, T2x, T2y, T20, T24, TX, TY; Chris@42: E T1i, T1n; Chris@42: T2V = T1P + T1Q; Chris@42: T2X = T1T - T1U; Chris@42: Tg = Ta + Tf; Chris@42: Ti = Tg * Th; Chris@42: Tq = Tg * Tm; Chris@42: TE = TC + TD; Chris@42: TF = TE * Tm; Chris@42: TJ = TE * Th; Chris@42: T3W = T37 * Tm; Chris@42: T3X = T39 * Th; Chris@42: T3Y = T3W - T3X; Chris@42: T3T = T37 * Th; Chris@42: T3U = T39 * Tm; Chris@42: T3V = T3T + T3U; Chris@42: T3h = T3c * Tm; Chris@42: T3i = T3e * Th; Chris@42: T3j = T3h - T3i; Chris@42: Tl = Tj - Tk; Chris@42: Tn = Tl * Tm; Chris@42: Tr = Tl * Th; Chris@42: TA = Tx - Tz; Chris@42: TB = TA * Th; Chris@42: TI = TA * Tm; Chris@42: T3d = T3c * Th; Chris@42: T3f = T3e * Tm; Chris@42: T3g = T3d + T3f; Chris@42: T1j = Tj + Tk; Chris@42: T1k = T1j * Tm; Chris@42: T1o = T1j * Th; Chris@42: T1t = Tx + Tz; Chris@42: T1Z = T1t * Th; Chris@42: T23 = T1t * Tm; Chris@42: TQ = TP * Th; Chris@42: TU = TT * Tm; Chris@42: TV = TQ + TU; Chris@42: T2A = T1a * Tm; Chris@42: T2B = T1e * Th; Chris@42: T2C = T2A - T2B; Chris@42: T2x = T1a * Th; Chris@42: T2y = T1e * Tm; Chris@42: T2z = T2x + T2y; Chris@42: T1u = TC - TD; Chris@42: T20 = T1u * Tm; Chris@42: T24 = T1u * Th; Chris@42: TX = TP * Tm; Chris@42: TY = TT * Th; Chris@42: TZ = TX - TY; Chris@42: T1h = Ta - Tf; Chris@42: T1i = T1h * Th; Chris@42: T1n = T1h * Tm; Chris@42: To = Ti - Tn; Chris@42: T1p = T1n + T1o; Chris@42: T6j = TQ - TU; Chris@42: T6H = T2A + T2B; Chris@42: Ts = Tq + Tr; Chris@42: T1l = T1i - T1k; Chris@42: T6l = TX + TY; Chris@42: T6F = T2x - T2y; Chris@42: T2P = T1Z - T20; Chris@42: T4b = TI + TJ; Chris@42: T4x = T3d - T3f; Chris@42: T5i = T3W + T3X; Chris@42: T2R = T23 + T24; Chris@42: T49 = TB - TF; Chris@42: T4z = T3h + T3i; Chris@42: T5g = T3T - T3U; Chris@42: TG = TB + TF; Chris@42: T4k = Ti + Tn; Chris@42: T4m = Tq - Tr; Chris@42: TK = TI - TJ; Chris@42: T21 = T1Z + T20; Chris@42: T3O = T1i + T1k; Chris@42: T3Q = T1n - T1o; Chris@42: T25 = T23 - T24; Chris@42: TW = W[8]; Chris@42: T10 = W[9]; Chris@42: T11 = FMA(TV, TW, TZ * T10); Chris@42: T79 = FNMS(T25, TW, T21 * T10); Chris@42: T6X = FNMS(Td, TW, T8 * T10); Chris@42: T5M = FNMS(T2X, TW, T2V * T10); Chris@42: T6b = FNMS(TK, TW, TG * T10); Chris@42: T1v = FMA(T1t, TW, T1u * T10); Chris@42: T30 = FMA(T1h, TW, T1j * T10); Chris@42: T69 = FMA(TG, TW, TK * T10); Chris@42: T77 = FMA(T21, TW, T25 * T10); Chris@42: T13 = FNMS(TZ, TW, TV * T10); Chris@42: T2F = FNMS(T2C, TW, T2z * T10); Chris@42: T2D = FMA(T2z, TW, T2C * T10); Chris@42: T6p = FMA(T1a, TW, T1e * T10); Chris@42: T6O = FMA(TP, TW, TT * T10); Chris@42: T1x = FNMS(T1u, TW, T1t * T10); Chris@42: T2a = FNMS(TE, TW, TA * T10); Chris@42: T2f = FMA(T3, TW, T6 * T10); Chris@42: T6V = FMA(T8, TW, Td * T10); Chris@42: T28 = FMA(TA, TW, TE * T10); Chris@42: T6r = FNMS(T1e, TW, T1a * T10); Chris@42: T2h = FNMS(T6, TW, T3 * T10); Chris@42: T6Q = FNMS(TT, TW, TP * T10); Chris@42: T32 = FNMS(T1j, TW, T1h * T10); Chris@42: T5K = FMA(T2V, TW, T2X * T10); Chris@42: T5w = FMA(Tw, TW, Ty * T10); Chris@42: T4G = FMA(T3O, TW, T3Q * T10); Chris@42: T4Q = FMA(T4k, TW, T4m * T10); Chris@42: T3m = FNMS(T3j, TW, T3g * T10); Chris@42: T4h = FNMS(Te, TW, T9 * T10); Chris@42: T4I = FNMS(T3Q, TW, T3O * T10); Chris@42: T5y = FNMS(Ty, TW, Tw * T10); Chris@42: T3k = FMA(T3g, TW, T3j * T10); Chris@42: T4f = FMA(T9, TW, Te * T10); Chris@42: T41 = FNMS(T3Y, TW, T3V * T10); Chris@42: T4S = FNMS(T4m, TW, T4k * T10); Chris@42: T4Y = FNMS(T3e, TW, T3c * T10); Chris@42: T3q = FMA(Tg, TW, Tl * T10); Chris@42: T3D = FMA(T2, TW, T5 * T10); Chris@42: T3F = FNMS(T5, TW, T2 * T10); Chris@42: T5r = FNMS(T39, TW, T37 * T10); Chris@42: T3s = FNMS(Tl, TW, Tg * T10); Chris@42: T4W = FMA(T3c, TW, T3e * T10); Chris@42: T3Z = FMA(T3V, TW, T3Y * T10); Chris@42: T5p = FMA(T37, TW, T39 * T10); Chris@42: } Chris@42: } Chris@42: } Chris@42: { Chris@42: E T17, TdV, Tj3, Tjx, T7l, TbJ, Ti3, Tix, T1K, Tiw, TdY, ThY, T7w, Tj0, TbM; Chris@42: E Tjw, T2e, TgA, T7I, TaY, TbQ, Tda, Te4, TfO, T2J, TgB, T7T, TaZ, TbT, Tdb; Chris@42: E Te9, TfP, T36, T3B, TgH, TgE, TgF, TgG, T80, TbW, Tel, TfT, T8b, Tc0, T8k; Chris@42: E TbX, Teg, TfS, T8h, TbZ, T45, T4q, TgJ, TgK, TgL, TgM, T8r, Tc6, Tew, TfW; Chris@42: E T8C, Tc4, T8L, Tc7, Ter, TfV, T8I, Tc3, T6B, Th1, Tfm, Tga, Th8, ThI, T9N; Chris@42: E Tcv, T9Y, TcH, Tav, Tcw, Tf5, Tg7, Tas, TcG, T5c, TgV, TeV, Tg0, TgS, ThD; Chris@42: E T8U, Tcc, T95, Tco, T9C, Tcd, TeE, Tg3, T9z, Tcn, T5R, TgT, TeO, TeW, TgY; Chris@42: E ThE, T9h, T9F, T9s, T9E, Tck, Tcq, TeJ, TeX, Tch, Tcr, T7e, Th9, Tff, Tfn; Chris@42: E Th4, ThJ, Taa, Tay, Tal, Tax, TcD, TcJ, Tfa, Tfo, TcA, TcK; Chris@42: { Chris@42: E T1, Ti1, Tu, Ti0, TM, T7i, T15, T7j, Tp, Tt; Chris@42: T1 = ri[0]; Chris@42: Ti1 = ii[0]; Chris@42: Tp = ri[WS(rs, 32)]; Chris@42: Tt = ii[WS(rs, 32)]; Chris@42: Tu = FMA(To, Tp, Ts * Tt); Chris@42: Ti0 = FNMS(Ts, Tp, To * Tt); Chris@42: { Chris@42: E TH, TL, T12, T14; Chris@42: TH = ri[WS(rs, 16)]; Chris@42: TL = ii[WS(rs, 16)]; Chris@42: TM = FMA(TG, TH, TK * TL); Chris@42: T7i = FNMS(TK, TH, TG * TL); Chris@42: T12 = ri[WS(rs, 48)]; Chris@42: T14 = ii[WS(rs, 48)]; Chris@42: T15 = FMA(T11, T12, T13 * T14); Chris@42: T7j = FNMS(T13, T12, T11 * T14); Chris@42: } Chris@42: { Chris@42: E Tv, T16, Tj1, Tj2; Chris@42: Tv = T1 + Tu; Chris@42: T16 = TM + T15; Chris@42: T17 = Tv + T16; Chris@42: TdV = Tv - T16; Chris@42: Tj1 = Ti1 - Ti0; Chris@42: Tj2 = TM - T15; Chris@42: Tj3 = Tj1 - Tj2; Chris@42: Tjx = Tj2 + Tj1; Chris@42: } Chris@42: { Chris@42: E T7h, T7k, ThZ, Ti2; Chris@42: T7h = T1 - Tu; Chris@42: T7k = T7i - T7j; Chris@42: T7l = T7h - T7k; Chris@42: TbJ = T7h + T7k; Chris@42: ThZ = T7i + T7j; Chris@42: Ti2 = Ti0 + Ti1; Chris@42: Ti3 = ThZ + Ti2; Chris@42: Tix = Ti2 - ThZ; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T1g, T7m, T1r, T7n, T7o, T7p, T1z, T7s, T1I, T7t, T7r, T7u; Chris@42: { Chris@42: E T1b, T1f, T1m, T1q; Chris@42: T1b = ri[WS(rs, 8)]; Chris@42: T1f = ii[WS(rs, 8)]; Chris@42: T1g = FMA(T1a, T1b, T1e * T1f); Chris@42: T7m = FNMS(T1e, T1b, T1a * T1f); Chris@42: T1m = ri[WS(rs, 40)]; Chris@42: T1q = ii[WS(rs, 40)]; Chris@42: T1r = FMA(T1l, T1m, T1p * T1q); Chris@42: T7n = FNMS(T1p, T1m, T1l * T1q); Chris@42: } Chris@42: T7o = T7m - T7n; Chris@42: T7p = T1g - T1r; Chris@42: { Chris@42: E T1w, T1y, T1D, T1H; Chris@42: T1w = ri[WS(rs, 56)]; Chris@42: T1y = ii[WS(rs, 56)]; Chris@42: T1z = FMA(T1v, T1w, T1x * T1y); Chris@42: T7s = FNMS(T1x, T1w, T1v * T1y); Chris@42: T1D = ri[WS(rs, 24)]; Chris@42: T1H = ii[WS(rs, 24)]; Chris@42: T1I = FMA(T1C, T1D, T1G * T1H); Chris@42: T7t = FNMS(T1G, T1D, T1C * T1H); Chris@42: } Chris@42: T7r = T1z - T1I; Chris@42: T7u = T7s - T7t; Chris@42: { Chris@42: E T1s, T1J, TdW, TdX; Chris@42: T1s = T1g + T1r; Chris@42: T1J = T1z + T1I; Chris@42: T1K = T1s + T1J; Chris@42: Tiw = T1J - T1s; Chris@42: TdW = T7m + T7n; Chris@42: TdX = T7s + T7t; Chris@42: TdY = TdW - TdX; Chris@42: ThY = TdW + TdX; Chris@42: } Chris@42: { Chris@42: E T7q, T7v, TbK, TbL; Chris@42: T7q = T7o - T7p; Chris@42: T7v = T7r + T7u; Chris@42: T7w = KP707106781 * (T7q - T7v); Chris@42: Tj0 = KP707106781 * (T7q + T7v); Chris@42: TbK = T7p + T7o; Chris@42: TbL = T7r - T7u; Chris@42: TbM = KP707106781 * (TbK + TbL); Chris@42: Tjw = KP707106781 * (TbL - TbK); Chris@42: } Chris@42: } Chris@42: { Chris@42: E T1Y, Te0, T7A, T7D, T2d, Te1, T7B, T7G, T7C, T7H; Chris@42: { Chris@42: E T1O, T7y, T1X, T7z; Chris@42: { Chris@42: E T1M, T1N, T1S, T1W; Chris@42: T1M = ri[WS(rs, 4)]; Chris@42: T1N = ii[WS(rs, 4)]; Chris@42: T1O = FMA(T8, T1M, Td * T1N); Chris@42: T7y = FNMS(Td, T1M, T8 * T1N); Chris@42: T1S = ri[WS(rs, 36)]; Chris@42: T1W = ii[WS(rs, 36)]; Chris@42: T1X = FMA(T1R, T1S, T1V * T1W); Chris@42: T7z = FNMS(T1V, T1S, T1R * T1W); Chris@42: } Chris@42: T1Y = T1O + T1X; Chris@42: Te0 = T7y + T7z; Chris@42: T7A = T7y - T7z; Chris@42: T7D = T1O - T1X; Chris@42: } Chris@42: { Chris@42: E T27, T7E, T2c, T7F; Chris@42: { Chris@42: E T22, T26, T29, T2b; Chris@42: T22 = ri[WS(rs, 20)]; Chris@42: T26 = ii[WS(rs, 20)]; Chris@42: T27 = FMA(T21, T22, T25 * T26); Chris@42: T7E = FNMS(T25, T22, T21 * T26); Chris@42: T29 = ri[WS(rs, 52)]; Chris@42: T2b = ii[WS(rs, 52)]; Chris@42: T2c = FMA(T28, T29, T2a * T2b); Chris@42: T7F = FNMS(T2a, T29, T28 * T2b); Chris@42: } Chris@42: T2d = T27 + T2c; Chris@42: Te1 = T7E + T7F; Chris@42: T7B = T27 - T2c; Chris@42: T7G = T7E - T7F; Chris@42: } Chris@42: T2e = T1Y + T2d; Chris@42: TgA = Te0 + Te1; Chris@42: T7C = T7A + T7B; Chris@42: T7H = T7D - T7G; Chris@42: T7I = FNMS(KP923879532, T7H, KP382683432 * T7C); Chris@42: TaY = FMA(KP923879532, T7C, KP382683432 * T7H); Chris@42: { Chris@42: E TbO, TbP, Te2, Te3; Chris@42: TbO = T7A - T7B; Chris@42: TbP = T7D + T7G; Chris@42: TbQ = FNMS(KP382683432, TbP, KP923879532 * TbO); Chris@42: Tda = FMA(KP382683432, TbO, KP923879532 * TbP); Chris@42: Te2 = Te0 - Te1; Chris@42: Te3 = T1Y - T2d; Chris@42: Te4 = Te2 - Te3; Chris@42: TfO = Te3 + Te2; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T2t, Te6, T7L, T7O, T2I, Te7, T7M, T7R, T7N, T7S; Chris@42: { Chris@42: E T2j, T7J, T2s, T7K; Chris@42: { Chris@42: E T2g, T2i, T2n, T2r; Chris@42: T2g = ri[WS(rs, 60)]; Chris@42: T2i = ii[WS(rs, 60)]; Chris@42: T2j = FMA(T2f, T2g, T2h * T2i); Chris@42: T7J = FNMS(T2h, T2g, T2f * T2i); Chris@42: T2n = ri[WS(rs, 28)]; Chris@42: T2r = ii[WS(rs, 28)]; Chris@42: T2s = FMA(T2m, T2n, T2q * T2r); Chris@42: T7K = FNMS(T2q, T2n, T2m * T2r); Chris@42: } Chris@42: T2t = T2j + T2s; Chris@42: Te6 = T7J + T7K; Chris@42: T7L = T7J - T7K; Chris@42: T7O = T2j - T2s; Chris@42: } Chris@42: { Chris@42: E T2w, T7P, T2H, T7Q; Chris@42: { Chris@42: E T2u, T2v, T2E, T2G; Chris@42: T2u = ri[WS(rs, 12)]; Chris@42: T2v = ii[WS(rs, 12)]; Chris@42: T2w = FMA(TP, T2u, TT * T2v); Chris@42: T7P = FNMS(TT, T2u, TP * T2v); Chris@42: T2E = ri[WS(rs, 44)]; Chris@42: T2G = ii[WS(rs, 44)]; Chris@42: T2H = FMA(T2D, T2E, T2F * T2G); Chris@42: T7Q = FNMS(T2F, T2E, T2D * T2G); Chris@42: } Chris@42: T2I = T2w + T2H; Chris@42: Te7 = T7P + T7Q; Chris@42: T7M = T2w - T2H; Chris@42: T7R = T7P - T7Q; Chris@42: } Chris@42: T2J = T2t + T2I; Chris@42: TgB = Te6 + Te7; Chris@42: T7N = T7L + T7M; Chris@42: T7S = T7O - T7R; Chris@42: T7T = FMA(KP382683432, T7N, KP923879532 * T7S); Chris@42: TaZ = FNMS(KP923879532, T7N, KP382683432 * T7S); Chris@42: { Chris@42: E TbR, TbS, Te5, Te8; Chris@42: TbR = T7L - T7M; Chris@42: TbS = T7O + T7R; Chris@42: TbT = FMA(KP923879532, TbR, KP382683432 * TbS); Chris@42: Tdb = FNMS(KP382683432, TbR, KP923879532 * TbS); Chris@42: Te5 = T2t - T2I; Chris@42: Te8 = Te6 - Te7; Chris@42: Te9 = Te5 + Te8; Chris@42: TfP = Te5 - Te8; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T2O, T7W, T2T, T7X, T2U, Tec, T2Z, T8e, T34, T8f, T35, Ted, T3p, Tei, T86; Chris@42: E T89, T3A, Tej, T81, T84; Chris@42: { Chris@42: E T2M, T2N, T2Q, T2S; Chris@42: T2M = ri[WS(rs, 2)]; Chris@42: T2N = ii[WS(rs, 2)]; Chris@42: T2O = FMA(Tw, T2M, Ty * T2N); Chris@42: T7W = FNMS(Ty, T2M, Tw * T2N); Chris@42: T2Q = ri[WS(rs, 34)]; Chris@42: T2S = ii[WS(rs, 34)]; Chris@42: T2T = FMA(T2P, T2Q, T2R * T2S); Chris@42: T7X = FNMS(T2R, T2Q, T2P * T2S); Chris@42: } Chris@42: T2U = T2O + T2T; Chris@42: Tec = T7W + T7X; Chris@42: { Chris@42: E T2W, T2Y, T31, T33; Chris@42: T2W = ri[WS(rs, 18)]; Chris@42: T2Y = ii[WS(rs, 18)]; Chris@42: T2Z = FMA(T2V, T2W, T2X * T2Y); Chris@42: T8e = FNMS(T2X, T2W, T2V * T2Y); Chris@42: T31 = ri[WS(rs, 50)]; Chris@42: T33 = ii[WS(rs, 50)]; Chris@42: T34 = FMA(T30, T31, T32 * T33); Chris@42: T8f = FNMS(T32, T31, T30 * T33); Chris@42: } Chris@42: T35 = T2Z + T34; Chris@42: Ted = T8e + T8f; Chris@42: { Chris@42: E T3b, T87, T3o, T88; Chris@42: { Chris@42: E T38, T3a, T3l, T3n; Chris@42: T38 = ri[WS(rs, 10)]; Chris@42: T3a = ii[WS(rs, 10)]; Chris@42: T3b = FMA(T37, T38, T39 * T3a); Chris@42: T87 = FNMS(T39, T38, T37 * T3a); Chris@42: T3l = ri[WS(rs, 42)]; Chris@42: T3n = ii[WS(rs, 42)]; Chris@42: T3o = FMA(T3k, T3l, T3m * T3n); Chris@42: T88 = FNMS(T3m, T3l, T3k * T3n); Chris@42: } Chris@42: T3p = T3b + T3o; Chris@42: Tei = T87 + T88; Chris@42: T86 = T3b - T3o; Chris@42: T89 = T87 - T88; Chris@42: } Chris@42: { Chris@42: E T3u, T82, T3z, T83; Chris@42: { Chris@42: E T3r, T3t, T3w, T3y; Chris@42: T3r = ri[WS(rs, 58)]; Chris@42: T3t = ii[WS(rs, 58)]; Chris@42: T3u = FMA(T3q, T3r, T3s * T3t); Chris@42: T82 = FNMS(T3s, T3r, T3q * T3t); Chris@42: T3w = ri[WS(rs, 26)]; Chris@42: T3y = ii[WS(rs, 26)]; Chris@42: T3z = FMA(T3v, T3w, T3x * T3y); Chris@42: T83 = FNMS(T3x, T3w, T3v * T3y); Chris@42: } Chris@42: T3A = T3u + T3z; Chris@42: Tej = T82 + T83; Chris@42: T81 = T3u - T3z; Chris@42: T84 = T82 - T83; Chris@42: } Chris@42: T36 = T2U + T35; Chris@42: T3B = T3p + T3A; Chris@42: TgH = T36 - T3B; Chris@42: TgE = Tec + Ted; Chris@42: TgF = Tei + Tej; Chris@42: TgG = TgE - TgF; Chris@42: { Chris@42: E T7Y, T7Z, Teh, Tek; Chris@42: T7Y = T7W - T7X; Chris@42: T7Z = T2Z - T34; Chris@42: T80 = T7Y + T7Z; Chris@42: TbW = T7Y - T7Z; Chris@42: Teh = T2U - T35; Chris@42: Tek = Tei - Tej; Chris@42: Tel = Teh - Tek; Chris@42: TfT = Teh + Tek; Chris@42: } Chris@42: { Chris@42: E T85, T8a, T8i, T8j; Chris@42: T85 = T81 - T84; Chris@42: T8a = T86 + T89; Chris@42: T8b = KP707106781 * (T85 - T8a); Chris@42: Tc0 = KP707106781 * (T8a + T85); Chris@42: T8i = T89 - T86; Chris@42: T8j = T81 + T84; Chris@42: T8k = KP707106781 * (T8i - T8j); Chris@42: TbX = KP707106781 * (T8i + T8j); Chris@42: } Chris@42: { Chris@42: E Tee, Tef, T8d, T8g; Chris@42: Tee = Tec - Ted; Chris@42: Tef = T3A - T3p; Chris@42: Teg = Tee - Tef; Chris@42: TfS = Tee + Tef; Chris@42: T8d = T2O - T2T; Chris@42: T8g = T8e - T8f; Chris@42: T8h = T8d - T8g; Chris@42: TbZ = T8d + T8g; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T3H, T8n, T3M, T8o, T3N, Ten, T3S, T8F, T43, T8G, T44, Teo, T4e, Tet, T8x; Chris@42: E T8A, T4p, Teu, T8s, T8v; Chris@42: { Chris@42: E T3E, T3G, T3J, T3L; Chris@42: T3E = ri[WS(rs, 62)]; Chris@42: T3G = ii[WS(rs, 62)]; Chris@42: T3H = FMA(T3D, T3E, T3F * T3G); Chris@42: T8n = FNMS(T3F, T3E, T3D * T3G); Chris@42: T3J = ri[WS(rs, 30)]; Chris@42: T3L = ii[WS(rs, 30)]; Chris@42: T3M = FMA(T3I, T3J, T3K * T3L); Chris@42: T8o = FNMS(T3K, T3J, T3I * T3L); Chris@42: } Chris@42: T3N = T3H + T3M; Chris@42: Ten = T8n + T8o; Chris@42: { Chris@42: E T3P, T3R, T40, T42; Chris@42: T3P = ri[WS(rs, 14)]; Chris@42: T3R = ii[WS(rs, 14)]; Chris@42: T3S = FMA(T3O, T3P, T3Q * T3R); Chris@42: T8F = FNMS(T3Q, T3P, T3O * T3R); Chris@42: T40 = ri[WS(rs, 46)]; Chris@42: T42 = ii[WS(rs, 46)]; Chris@42: T43 = FMA(T3Z, T40, T41 * T42); Chris@42: T8G = FNMS(T41, T40, T3Z * T42); Chris@42: } Chris@42: T44 = T3S + T43; Chris@42: Teo = T8F + T8G; Chris@42: { Chris@42: E T48, T8y, T4d, T8z; Chris@42: { Chris@42: E T46, T47, T4a, T4c; Chris@42: T46 = ri[WS(rs, 6)]; Chris@42: T47 = ii[WS(rs, 6)]; Chris@42: T48 = FMA(T3c, T46, T3e * T47); Chris@42: T8y = FNMS(T3e, T46, T3c * T47); Chris@42: T4a = ri[WS(rs, 38)]; Chris@42: T4c = ii[WS(rs, 38)]; Chris@42: T4d = FMA(T49, T4a, T4b * T4c); Chris@42: T8z = FNMS(T4b, T4a, T49 * T4c); Chris@42: } Chris@42: T4e = T48 + T4d; Chris@42: Tet = T8y + T8z; Chris@42: T8x = T48 - T4d; Chris@42: T8A = T8y - T8z; Chris@42: } Chris@42: { Chris@42: E T4j, T8t, T4o, T8u; Chris@42: { Chris@42: E T4g, T4i, T4l, T4n; Chris@42: T4g = ri[WS(rs, 54)]; Chris@42: T4i = ii[WS(rs, 54)]; Chris@42: T4j = FMA(T4f, T4g, T4h * T4i); Chris@42: T8t = FNMS(T4h, T4g, T4f * T4i); Chris@42: T4l = ri[WS(rs, 22)]; Chris@42: T4n = ii[WS(rs, 22)]; Chris@42: T4o = FMA(T4k, T4l, T4m * T4n); Chris@42: T8u = FNMS(T4m, T4l, T4k * T4n); Chris@42: } Chris@42: T4p = T4j + T4o; Chris@42: Teu = T8t + T8u; Chris@42: T8s = T4j - T4o; Chris@42: T8v = T8t - T8u; Chris@42: } Chris@42: T45 = T3N + T44; Chris@42: T4q = T4e + T4p; Chris@42: TgJ = T45 - T4q; Chris@42: TgK = Ten + Teo; Chris@42: TgL = Tet + Teu; Chris@42: TgM = TgK - TgL; Chris@42: { Chris@42: E T8p, T8q, Tes, Tev; Chris@42: T8p = T8n - T8o; Chris@42: T8q = T3S - T43; Chris@42: T8r = T8p + T8q; Chris@42: Tc6 = T8p - T8q; Chris@42: Tes = T3N - T44; Chris@42: Tev = Tet - Teu; Chris@42: Tew = Tes - Tev; Chris@42: TfW = Tes + Tev; Chris@42: } Chris@42: { Chris@42: E T8w, T8B, T8J, T8K; Chris@42: T8w = T8s - T8v; Chris@42: T8B = T8x + T8A; Chris@42: T8C = KP707106781 * (T8w - T8B); Chris@42: Tc4 = KP707106781 * (T8B + T8w); Chris@42: T8J = T8A - T8x; Chris@42: T8K = T8s + T8v; Chris@42: T8L = KP707106781 * (T8J - T8K); Chris@42: Tc7 = KP707106781 * (T8J + T8K); Chris@42: } Chris@42: { Chris@42: E Tep, Teq, T8E, T8H; Chris@42: Tep = Ten - Teo; Chris@42: Teq = T4p - T4e; Chris@42: Ter = Tep - Teq; Chris@42: TfV = Tep + Teq; Chris@42: T8E = T3H - T3M; Chris@42: T8H = T8F - T8G; Chris@42: T8I = T8E - T8H; Chris@42: Tc3 = T8E + T8H; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T5V, Tao, T64, Tap, T65, Tfi, T68, T9K, T6d, T9L, T6e, Tfj, T6o, Tf2, T9Q; Chris@42: E T9R, T6z, Tf3, T9T, T9W; Chris@42: { Chris@42: E T5T, T5U, T5Z, T63; Chris@42: T5T = ri[WS(rs, 63)]; Chris@42: T5U = ii[WS(rs, 63)]; Chris@42: T5V = FMA(TW, T5T, T10 * T5U); Chris@42: Tao = FNMS(T10, T5T, TW * T5U); Chris@42: T5Z = ri[WS(rs, 31)]; Chris@42: T63 = ii[WS(rs, 31)]; Chris@42: T64 = FMA(T5Y, T5Z, T62 * T63); Chris@42: Tap = FNMS(T62, T5Z, T5Y * T63); Chris@42: } Chris@42: T65 = T5V + T64; Chris@42: Tfi = Tao + Tap; Chris@42: { Chris@42: E T66, T67, T6a, T6c; Chris@42: T66 = ri[WS(rs, 15)]; Chris@42: T67 = ii[WS(rs, 15)]; Chris@42: T68 = FMA(TV, T66, TZ * T67); Chris@42: T9K = FNMS(TZ, T66, TV * T67); Chris@42: T6a = ri[WS(rs, 47)]; Chris@42: T6c = ii[WS(rs, 47)]; Chris@42: T6d = FMA(T69, T6a, T6b * T6c); Chris@42: T9L = FNMS(T6b, T6a, T69 * T6c); Chris@42: } Chris@42: T6e = T68 + T6d; Chris@42: Tfj = T9K + T9L; Chris@42: { Chris@42: E T6i, T9O, T6n, T9P; Chris@42: { Chris@42: E T6g, T6h, T6k, T6m; Chris@42: T6g = ri[WS(rs, 7)]; Chris@42: T6h = ii[WS(rs, 7)]; Chris@42: T6i = FMA(T1t, T6g, T1u * T6h); Chris@42: T9O = FNMS(T1u, T6g, T1t * T6h); Chris@42: T6k = ri[WS(rs, 39)]; Chris@42: T6m = ii[WS(rs, 39)]; Chris@42: T6n = FMA(T6j, T6k, T6l * T6m); Chris@42: T9P = FNMS(T6l, T6k, T6j * T6m); Chris@42: } Chris@42: T6o = T6i + T6n; Chris@42: Tf2 = T9O + T9P; Chris@42: T9Q = T9O - T9P; Chris@42: T9R = T6i - T6n; Chris@42: } Chris@42: { Chris@42: E T6t, T9U, T6y, T9V; Chris@42: { Chris@42: E T6q, T6s, T6v, T6x; Chris@42: T6q = ri[WS(rs, 55)]; Chris@42: T6s = ii[WS(rs, 55)]; Chris@42: T6t = FMA(T6p, T6q, T6r * T6s); Chris@42: T9U = FNMS(T6r, T6q, T6p * T6s); Chris@42: T6v = ri[WS(rs, 23)]; Chris@42: T6x = ii[WS(rs, 23)]; Chris@42: T6y = FMA(T6u, T6v, T6w * T6x); Chris@42: T9V = FNMS(T6w, T6v, T6u * T6x); Chris@42: } Chris@42: T6z = T6t + T6y; Chris@42: Tf3 = T9U + T9V; Chris@42: T9T = T6t - T6y; Chris@42: T9W = T9U - T9V; Chris@42: } Chris@42: { Chris@42: E T6f, T6A, Tfk, Tfl; Chris@42: T6f = T65 + T6e; Chris@42: T6A = T6o + T6z; Chris@42: T6B = T6f + T6A; Chris@42: Th1 = T6f - T6A; Chris@42: Tfk = Tfi - Tfj; Chris@42: Tfl = T6z - T6o; Chris@42: Tfm = Tfk - Tfl; Chris@42: Tga = Tfk + Tfl; Chris@42: } Chris@42: { Chris@42: E Th6, Th7, T9J, T9M; Chris@42: Th6 = Tfi + Tfj; Chris@42: Th7 = Tf2 + Tf3; Chris@42: Th8 = Th6 - Th7; Chris@42: ThI = Th6 + Th7; Chris@42: T9J = T5V - T64; Chris@42: T9M = T9K - T9L; Chris@42: T9N = T9J - T9M; Chris@42: Tcv = T9J + T9M; Chris@42: } Chris@42: { Chris@42: E T9S, T9X, Tat, Tau; Chris@42: T9S = T9Q - T9R; Chris@42: T9X = T9T + T9W; Chris@42: T9Y = KP707106781 * (T9S - T9X); Chris@42: TcH = KP707106781 * (T9S + T9X); Chris@42: Tat = T9T - T9W; Chris@42: Tau = T9R + T9Q; Chris@42: Tav = KP707106781 * (Tat - Tau); Chris@42: Tcw = KP707106781 * (Tau + Tat); Chris@42: } Chris@42: { Chris@42: E Tf1, Tf4, Taq, Tar; Chris@42: Tf1 = T65 - T6e; Chris@42: Tf4 = Tf2 - Tf3; Chris@42: Tf5 = Tf1 - Tf4; Chris@42: Tg7 = Tf1 + Tf4; Chris@42: Taq = Tao - Tap; Chris@42: Tar = T68 - T6d; Chris@42: Tas = Taq + Tar; Chris@42: TcG = Taq - Tar; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T4w, T8Q, T4B, T8R, T4C, TeA, T4F, T9w, T4K, T9x, T4L, TeB, T4V, TeS, T90; Chris@42: E T93, T5a, TeT, T8V, T8Y; Chris@42: { Chris@42: E T4u, T4v, T4y, T4A; Chris@42: T4u = ri[WS(rs, 1)]; Chris@42: T4v = ii[WS(rs, 1)]; Chris@42: T4w = FMA(T2, T4u, T5 * T4v); Chris@42: T8Q = FNMS(T5, T4u, T2 * T4v); Chris@42: T4y = ri[WS(rs, 33)]; Chris@42: T4A = ii[WS(rs, 33)]; Chris@42: T4B = FMA(T4x, T4y, T4z * T4A); Chris@42: T8R = FNMS(T4z, T4y, T4x * T4A); Chris@42: } Chris@42: T4C = T4w + T4B; Chris@42: TeA = T8Q + T8R; Chris@42: { Chris@42: E T4D, T4E, T4H, T4J; Chris@42: T4D = ri[WS(rs, 17)]; Chris@42: T4E = ii[WS(rs, 17)]; Chris@42: T4F = FMA(T3V, T4D, T3Y * T4E); Chris@42: T9w = FNMS(T3Y, T4D, T3V * T4E); Chris@42: T4H = ri[WS(rs, 49)]; Chris@42: T4J = ii[WS(rs, 49)]; Chris@42: T4K = FMA(T4G, T4H, T4I * T4J); Chris@42: T9x = FNMS(T4I, T4H, T4G * T4J); Chris@42: } Chris@42: T4L = T4F + T4K; Chris@42: TeB = T9w + T9x; Chris@42: { Chris@42: E T4P, T91, T4U, T92; Chris@42: { Chris@42: E T4N, T4O, T4R, T4T; Chris@42: T4N = ri[WS(rs, 9)]; Chris@42: T4O = ii[WS(rs, 9)]; Chris@42: T4P = FMA(T9, T4N, Te * T4O); Chris@42: T91 = FNMS(Te, T4N, T9 * T4O); Chris@42: T4R = ri[WS(rs, 41)]; Chris@42: T4T = ii[WS(rs, 41)]; Chris@42: T4U = FMA(T4Q, T4R, T4S * T4T); Chris@42: T92 = FNMS(T4S, T4R, T4Q * T4T); Chris@42: } Chris@42: T4V = T4P + T4U; Chris@42: TeS = T91 + T92; Chris@42: T90 = T4P - T4U; Chris@42: T93 = T91 - T92; Chris@42: } Chris@42: { Chris@42: E T50, T8W, T59, T8X; Chris@42: { Chris@42: E T4X, T4Z, T54, T58; Chris@42: T4X = ri[WS(rs, 57)]; Chris@42: T4Z = ii[WS(rs, 57)]; Chris@42: T50 = FMA(T4W, T4X, T4Y * T4Z); Chris@42: T8W = FNMS(T4Y, T4X, T4W * T4Z); Chris@42: T54 = ri[WS(rs, 25)]; Chris@42: T58 = ii[WS(rs, 25)]; Chris@42: T59 = FMA(T53, T54, T57 * T58); Chris@42: T8X = FNMS(T57, T54, T53 * T58); Chris@42: } Chris@42: T5a = T50 + T59; Chris@42: TeT = T8W + T8X; Chris@42: T8V = T50 - T59; Chris@42: T8Y = T8W - T8X; Chris@42: } Chris@42: { Chris@42: E T4M, T5b, TeR, TeU; Chris@42: T4M = T4C + T4L; Chris@42: T5b = T4V + T5a; Chris@42: T5c = T4M + T5b; Chris@42: TgV = T4M - T5b; Chris@42: TeR = T4C - T4L; Chris@42: TeU = TeS - TeT; Chris@42: TeV = TeR - TeU; Chris@42: Tg0 = TeR + TeU; Chris@42: } Chris@42: { Chris@42: E TgQ, TgR, T8S, T8T; Chris@42: TgQ = TeA + TeB; Chris@42: TgR = TeS + TeT; Chris@42: TgS = TgQ - TgR; Chris@42: ThD = TgQ + TgR; Chris@42: T8S = T8Q - T8R; Chris@42: T8T = T4F - T4K; Chris@42: T8U = T8S + T8T; Chris@42: Tcc = T8S - T8T; Chris@42: } Chris@42: { Chris@42: E T8Z, T94, T9A, T9B; Chris@42: T8Z = T8V - T8Y; Chris@42: T94 = T90 + T93; Chris@42: T95 = KP707106781 * (T8Z - T94); Chris@42: Tco = KP707106781 * (T94 + T8Z); Chris@42: T9A = T93 - T90; Chris@42: T9B = T8V + T8Y; Chris@42: T9C = KP707106781 * (T9A - T9B); Chris@42: Tcd = KP707106781 * (T9A + T9B); Chris@42: } Chris@42: { Chris@42: E TeC, TeD, T9v, T9y; Chris@42: TeC = TeA - TeB; Chris@42: TeD = T5a - T4V; Chris@42: TeE = TeC - TeD; Chris@42: Tg3 = TeC + TeD; Chris@42: T9v = T4w - T4B; Chris@42: T9y = T9w - T9x; Chris@42: T9z = T9v - T9y; Chris@42: Tcn = T9v + T9y; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T5l, TeL, T9k, T9n, T5P, TeH, T9a, T9f, T5u, TeM, T9l, T9q, T5G, TeG, T97; Chris@42: E T9e; Chris@42: { Chris@42: E T5f, T9i, T5k, T9j; Chris@42: { Chris@42: E T5d, T5e, T5h, T5j; Chris@42: T5d = ri[WS(rs, 5)]; Chris@42: T5e = ii[WS(rs, 5)]; Chris@42: T5f = FMA(Tg, T5d, Tl * T5e); Chris@42: T9i = FNMS(Tl, T5d, Tg * T5e); Chris@42: T5h = ri[WS(rs, 37)]; Chris@42: T5j = ii[WS(rs, 37)]; Chris@42: T5k = FMA(T5g, T5h, T5i * T5j); Chris@42: T9j = FNMS(T5i, T5h, T5g * T5j); Chris@42: } Chris@42: T5l = T5f + T5k; Chris@42: TeL = T9i + T9j; Chris@42: T9k = T9i - T9j; Chris@42: T9n = T5f - T5k; Chris@42: } Chris@42: { Chris@42: E T5J, T98, T5O, T99; Chris@42: { Chris@42: E T5H, T5I, T5L, T5N; Chris@42: T5H = ri[WS(rs, 13)]; Chris@42: T5I = ii[WS(rs, 13)]; Chris@42: T5J = FMA(T1h, T5H, T1j * T5I); Chris@42: T98 = FNMS(T1j, T5H, T1h * T5I); Chris@42: T5L = ri[WS(rs, 45)]; Chris@42: T5N = ii[WS(rs, 45)]; Chris@42: T5O = FMA(T5K, T5L, T5M * T5N); Chris@42: T99 = FNMS(T5M, T5L, T5K * T5N); Chris@42: } Chris@42: T5P = T5J + T5O; Chris@42: TeH = T98 + T99; Chris@42: T9a = T98 - T99; Chris@42: T9f = T5J - T5O; Chris@42: } Chris@42: { Chris@42: E T5o, T9o, T5t, T9p; Chris@42: { Chris@42: E T5m, T5n, T5q, T5s; Chris@42: T5m = ri[WS(rs, 21)]; Chris@42: T5n = ii[WS(rs, 21)]; Chris@42: T5o = FMA(T3g, T5m, T3j * T5n); Chris@42: T9o = FNMS(T3j, T5m, T3g * T5n); Chris@42: T5q = ri[WS(rs, 53)]; Chris@42: T5s = ii[WS(rs, 53)]; Chris@42: T5t = FMA(T5p, T5q, T5r * T5s); Chris@42: T9p = FNMS(T5r, T5q, T5p * T5s); Chris@42: } Chris@42: T5u = T5o + T5t; Chris@42: TeM = T9o + T9p; Chris@42: T9l = T5o - T5t; Chris@42: T9q = T9o - T9p; Chris@42: } Chris@42: { Chris@42: E T5A, T9c, T5F, T9d; Chris@42: { Chris@42: E T5x, T5z, T5C, T5E; Chris@42: T5x = ri[WS(rs, 61)]; Chris@42: T5z = ii[WS(rs, 61)]; Chris@42: T5A = FMA(T5w, T5x, T5y * T5z); Chris@42: T9c = FNMS(T5y, T5x, T5w * T5z); Chris@42: T5C = ri[WS(rs, 29)]; Chris@42: T5E = ii[WS(rs, 29)]; Chris@42: T5F = FMA(T5B, T5C, T5D * T5E); Chris@42: T9d = FNMS(T5D, T5C, T5B * T5E); Chris@42: } Chris@42: T5G = T5A + T5F; Chris@42: TeG = T9c + T9d; Chris@42: T97 = T5A - T5F; Chris@42: T9e = T9c - T9d; Chris@42: } Chris@42: { Chris@42: E T5v, T5Q, TeK, TeN; Chris@42: T5v = T5l + T5u; Chris@42: T5Q = T5G + T5P; Chris@42: T5R = T5v + T5Q; Chris@42: TgT = T5Q - T5v; Chris@42: TeK = T5l - T5u; Chris@42: TeN = TeL - TeM; Chris@42: TeO = TeK + TeN; Chris@42: TeW = TeN - TeK; Chris@42: } Chris@42: { Chris@42: E TgW, TgX, T9b, T9g; Chris@42: TgW = TeL + TeM; Chris@42: TgX = TeG + TeH; Chris@42: TgY = TgW - TgX; Chris@42: ThE = TgW + TgX; Chris@42: T9b = T97 - T9a; Chris@42: T9g = T9e + T9f; Chris@42: T9h = FNMS(KP923879532, T9g, KP382683432 * T9b); Chris@42: T9F = FMA(KP382683432, T9g, KP923879532 * T9b); Chris@42: } Chris@42: { Chris@42: E T9m, T9r, Tci, Tcj; Chris@42: T9m = T9k + T9l; Chris@42: T9r = T9n - T9q; Chris@42: T9s = FMA(KP923879532, T9m, KP382683432 * T9r); Chris@42: T9E = FNMS(KP923879532, T9r, KP382683432 * T9m); Chris@42: Tci = T9k - T9l; Chris@42: Tcj = T9n + T9q; Chris@42: Tck = FMA(KP382683432, Tci, KP923879532 * Tcj); Chris@42: Tcq = FNMS(KP382683432, Tcj, KP923879532 * Tci); Chris@42: } Chris@42: { Chris@42: E TeF, TeI, Tcf, Tcg; Chris@42: TeF = T5G - T5P; Chris@42: TeI = TeG - TeH; Chris@42: TeJ = TeF - TeI; Chris@42: TeX = TeF + TeI; Chris@42: Tcf = T97 + T9a; Chris@42: Tcg = T9e - T9f; Chris@42: Tch = FNMS(KP382683432, Tcg, KP923879532 * Tcf); Chris@42: Tcr = FMA(KP923879532, Tcg, KP382683432 * Tcf); Chris@42: } Chris@42: } Chris@42: { Chris@42: E T6K, Tf6, Ta2, Ta5, T7c, Tfd, Tae, Taj, T6T, Tf7, Ta3, Ta8, T73, Tfc, Tad; Chris@42: E Tag; Chris@42: { Chris@42: E T6E, Ta0, T6J, Ta1; Chris@42: { Chris@42: E T6C, T6D, T6G, T6I; Chris@42: T6C = ri[WS(rs, 3)]; Chris@42: T6D = ii[WS(rs, 3)]; Chris@42: T6E = FMA(T3, T6C, T6 * T6D); Chris@42: Ta0 = FNMS(T6, T6C, T3 * T6D); Chris@42: T6G = ri[WS(rs, 35)]; Chris@42: T6I = ii[WS(rs, 35)]; Chris@42: T6J = FMA(T6F, T6G, T6H * T6I); Chris@42: Ta1 = FNMS(T6H, T6G, T6F * T6I); Chris@42: } Chris@42: T6K = T6E + T6J; Chris@42: Tf6 = Ta0 + Ta1; Chris@42: Ta2 = Ta0 - Ta1; Chris@42: Ta5 = T6E - T6J; Chris@42: } Chris@42: { Chris@42: E T76, Tah, T7b, Tai; Chris@42: { Chris@42: E T74, T75, T78, T7a; Chris@42: T74 = ri[WS(rs, 11)]; Chris@42: T75 = ii[WS(rs, 11)]; Chris@42: T76 = FMA(TA, T74, TE * T75); Chris@42: Tah = FNMS(TE, T74, TA * T75); Chris@42: T78 = ri[WS(rs, 43)]; Chris@42: T7a = ii[WS(rs, 43)]; Chris@42: T7b = FMA(T77, T78, T79 * T7a); Chris@42: Tai = FNMS(T79, T78, T77 * T7a); Chris@42: } Chris@42: T7c = T76 + T7b; Chris@42: Tfd = Tah + Tai; Chris@42: Tae = T76 - T7b; Chris@42: Taj = Tah - Tai; Chris@42: } Chris@42: { Chris@42: E T6N, Ta6, T6S, Ta7; Chris@42: { Chris@42: E T6L, T6M, T6P, T6R; Chris@42: T6L = ri[WS(rs, 19)]; Chris@42: T6M = ii[WS(rs, 19)]; Chris@42: T6N = FMA(T2z, T6L, T2C * T6M); Chris@42: Ta6 = FNMS(T2C, T6L, T2z * T6M); Chris@42: T6P = ri[WS(rs, 51)]; Chris@42: T6R = ii[WS(rs, 51)]; Chris@42: T6S = FMA(T6O, T6P, T6Q * T6R); Chris@42: Ta7 = FNMS(T6Q, T6P, T6O * T6R); Chris@42: } Chris@42: T6T = T6N + T6S; Chris@42: Tf7 = Ta6 + Ta7; Chris@42: Ta3 = T6N - T6S; Chris@42: Ta8 = Ta6 - Ta7; Chris@42: } Chris@42: { Chris@42: E T6Z, Tab, T72, Tac; Chris@42: { Chris@42: E T6W, T6Y, T70, T71; Chris@42: T6W = ri[WS(rs, 59)]; Chris@42: T6Y = ii[WS(rs, 59)]; Chris@42: T6Z = FMA(T6V, T6W, T6X * T6Y); Chris@42: Tab = FNMS(T6X, T6W, T6V * T6Y); Chris@42: T70 = ri[WS(rs, 27)]; Chris@42: T71 = ii[WS(rs, 27)]; Chris@42: T72 = FMA(Th, T70, Tm * T71); Chris@42: Tac = FNMS(Tm, T70, Th * T71); Chris@42: } Chris@42: T73 = T6Z + T72; Chris@42: Tfc = Tab + Tac; Chris@42: Tad = Tab - Tac; Chris@42: Tag = T6Z - T72; Chris@42: } Chris@42: { Chris@42: E T6U, T7d, Tfb, Tfe; Chris@42: T6U = T6K + T6T; Chris@42: T7d = T73 + T7c; Chris@42: T7e = T6U + T7d; Chris@42: Th9 = T7d - T6U; Chris@42: Tfb = T73 - T7c; Chris@42: Tfe = Tfc - Tfd; Chris@42: Tff = Tfb + Tfe; Chris@42: Tfn = Tfb - Tfe; Chris@42: } Chris@42: { Chris@42: E Th2, Th3, Ta4, Ta9; Chris@42: Th2 = Tf6 + Tf7; Chris@42: Th3 = Tfc + Tfd; Chris@42: Th4 = Th2 - Th3; Chris@42: ThJ = Th2 + Th3; Chris@42: Ta4 = Ta2 + Ta3; Chris@42: Ta9 = Ta5 - Ta8; Chris@42: Taa = FNMS(KP923879532, Ta9, KP382683432 * Ta4); Chris@42: Tay = FMA(KP923879532, Ta4, KP382683432 * Ta9); Chris@42: } Chris@42: { Chris@42: E Taf, Tak, TcB, TcC; Chris@42: Taf = Tad + Tae; Chris@42: Tak = Tag - Taj; Chris@42: Tal = FMA(KP382683432, Taf, KP923879532 * Tak); Chris@42: Tax = FNMS(KP923879532, Taf, KP382683432 * Tak); Chris@42: TcB = Tad - Tae; Chris@42: TcC = Tag + Taj; Chris@42: TcD = FMA(KP923879532, TcB, KP382683432 * TcC); Chris@42: TcJ = FNMS(KP382683432, TcB, KP923879532 * TcC); Chris@42: } Chris@42: { Chris@42: E Tf8, Tf9, Tcy, Tcz; Chris@42: Tf8 = Tf6 - Tf7; Chris@42: Tf9 = T6K - T6T; Chris@42: Tfa = Tf8 - Tf9; Chris@42: Tfo = Tf9 + Tf8; Chris@42: Tcy = Ta2 - Ta3; Chris@42: Tcz = Ta5 + Ta8; Chris@42: TcA = FNMS(KP382683432, Tcz, KP923879532 * Tcy); Chris@42: TcK = FMA(KP382683432, Tcy, KP923879532 * Tcz); Chris@42: } Chris@42: } Chris@42: { Chris@42: E T2L, Thx, ThU, ThV, Ti5, Tib, T4s, Tia, T7g, Ti7, ThG, ThO, ThL, ThP, ThA; Chris@42: E ThW; Chris@42: { Chris@42: E T1L, T2K, ThS, ThT; Chris@42: T1L = T17 + T1K; Chris@42: T2K = T2e + T2J; Chris@42: T2L = T1L + T2K; Chris@42: Thx = T1L - T2K; Chris@42: ThS = ThD + ThE; Chris@42: ThT = ThI + ThJ; Chris@42: ThU = ThS - ThT; Chris@42: ThV = ThS + ThT; Chris@42: } Chris@42: { Chris@42: E ThX, Ti4, T3C, T4r; Chris@42: ThX = TgA + TgB; Chris@42: Ti4 = ThY + Ti3; Chris@42: Ti5 = ThX + Ti4; Chris@42: Tib = Ti4 - ThX; Chris@42: T3C = T36 + T3B; Chris@42: T4r = T45 + T4q; Chris@42: T4s = T3C + T4r; Chris@42: Tia = T4r - T3C; Chris@42: } Chris@42: { Chris@42: E T5S, T7f, ThC, ThF; Chris@42: T5S = T5c + T5R; Chris@42: T7f = T6B + T7e; Chris@42: T7g = T5S + T7f; Chris@42: Ti7 = T7f - T5S; Chris@42: ThC = T5c - T5R; Chris@42: ThF = ThD - ThE; Chris@42: ThG = ThC + ThF; Chris@42: ThO = ThF - ThC; Chris@42: } Chris@42: { Chris@42: E ThH, ThK, Thy, Thz; Chris@42: ThH = T6B - T7e; Chris@42: ThK = ThI - ThJ; Chris@42: ThL = ThH - ThK; Chris@42: ThP = ThH + ThK; Chris@42: Thy = TgE + TgF; Chris@42: Thz = TgK + TgL; Chris@42: ThA = Thy - Thz; Chris@42: ThW = Thy + Thz; Chris@42: } Chris@42: { Chris@42: E T4t, Ti6, ThR, Ti8; Chris@42: T4t = T2L + T4s; Chris@42: ri[WS(rs, 32)] = T4t - T7g; Chris@42: ri[0] = T4t + T7g; Chris@42: Ti6 = ThW + Ti5; Chris@42: ii[0] = ThV + Ti6; Chris@42: ii[WS(rs, 32)] = Ti6 - ThV; Chris@42: ThR = T2L - T4s; Chris@42: ri[WS(rs, 48)] = ThR - ThU; Chris@42: ri[WS(rs, 16)] = ThR + ThU; Chris@42: Ti8 = Ti5 - ThW; Chris@42: ii[WS(rs, 16)] = Ti7 + Ti8; Chris@42: ii[WS(rs, 48)] = Ti8 - Ti7; Chris@42: } Chris@42: { Chris@42: E ThB, ThM, Ti9, Tic; Chris@42: ThB = Thx + ThA; Chris@42: ThM = KP707106781 * (ThG + ThL); Chris@42: ri[WS(rs, 40)] = ThB - ThM; Chris@42: ri[WS(rs, 8)] = ThB + ThM; Chris@42: Ti9 = KP707106781 * (ThO + ThP); Chris@42: Tic = Tia + Tib; Chris@42: ii[WS(rs, 8)] = Ti9 + Tic; Chris@42: ii[WS(rs, 40)] = Tic - Ti9; Chris@42: } Chris@42: { Chris@42: E ThN, ThQ, Tid, Tie; Chris@42: ThN = Thx - ThA; Chris@42: ThQ = KP707106781 * (ThO - ThP); Chris@42: ri[WS(rs, 56)] = ThN - ThQ; Chris@42: ri[WS(rs, 24)] = ThN + ThQ; Chris@42: Tid = KP707106781 * (ThL - ThG); Chris@42: Tie = Tib - Tia; Chris@42: ii[WS(rs, 24)] = Tid + Tie; Chris@42: ii[WS(rs, 56)] = Tie - Tid; Chris@42: } Chris@42: } Chris@42: { Chris@42: E TgD, Thh, Thr, Thv, Tij, Tip, TgO, Tig, Th0, The, Thk, Tio, Tho, Thu, Thb; Chris@42: E Thf; Chris@42: { Chris@42: E Tgz, TgC, Thp, Thq; Chris@42: Tgz = T17 - T1K; Chris@42: TgC = TgA - TgB; Chris@42: TgD = Tgz - TgC; Chris@42: Thh = Tgz + TgC; Chris@42: Thp = Th1 + Th4; Chris@42: Thq = Th8 + Th9; Chris@42: Thr = FNMS(KP382683432, Thq, KP923879532 * Thp); Chris@42: Thv = FMA(KP923879532, Thq, KP382683432 * Thp); Chris@42: } Chris@42: { Chris@42: E Tih, Tii, TgI, TgN; Chris@42: Tih = T2J - T2e; Chris@42: Tii = Ti3 - ThY; Chris@42: Tij = Tih + Tii; Chris@42: Tip = Tii - Tih; Chris@42: TgI = TgG - TgH; Chris@42: TgN = TgJ + TgM; Chris@42: TgO = KP707106781 * (TgI - TgN); Chris@42: Tig = KP707106781 * (TgI + TgN); Chris@42: } Chris@42: { Chris@42: E TgU, TgZ, Thi, Thj; Chris@42: TgU = TgS - TgT; Chris@42: TgZ = TgV - TgY; Chris@42: Th0 = FMA(KP923879532, TgU, KP382683432 * TgZ); Chris@42: The = FNMS(KP923879532, TgZ, KP382683432 * TgU); Chris@42: Thi = TgH + TgG; Chris@42: Thj = TgJ - TgM; Chris@42: Thk = KP707106781 * (Thi + Thj); Chris@42: Tio = KP707106781 * (Thj - Thi); Chris@42: } Chris@42: { Chris@42: E Thm, Thn, Th5, Tha; Chris@42: Thm = TgS + TgT; Chris@42: Thn = TgV + TgY; Chris@42: Tho = FMA(KP382683432, Thm, KP923879532 * Thn); Chris@42: Thu = FNMS(KP382683432, Thn, KP923879532 * Thm); Chris@42: Th5 = Th1 - Th4; Chris@42: Tha = Th8 - Th9; Chris@42: Thb = FNMS(KP923879532, Tha, KP382683432 * Th5); Chris@42: Thf = FMA(KP382683432, Tha, KP923879532 * Th5); Chris@42: } Chris@42: { Chris@42: E TgP, Thc, Tin, Tiq; Chris@42: TgP = TgD + TgO; Chris@42: Thc = Th0 + Thb; Chris@42: ri[WS(rs, 44)] = TgP - Thc; Chris@42: ri[WS(rs, 12)] = TgP + Thc; Chris@42: Tin = The + Thf; Chris@42: Tiq = Tio + Tip; Chris@42: ii[WS(rs, 12)] = Tin + Tiq; Chris@42: ii[WS(rs, 44)] = Tiq - Tin; Chris@42: } Chris@42: { Chris@42: E Thd, Thg, Tir, Tis; Chris@42: Thd = TgD - TgO; Chris@42: Thg = The - Thf; Chris@42: ri[WS(rs, 60)] = Thd - Thg; Chris@42: ri[WS(rs, 28)] = Thd + Thg; Chris@42: Tir = Thb - Th0; Chris@42: Tis = Tip - Tio; Chris@42: ii[WS(rs, 28)] = Tir + Tis; Chris@42: ii[WS(rs, 60)] = Tis - Tir; Chris@42: } Chris@42: { Chris@42: E Thl, Ths, Tif, Tik; Chris@42: Thl = Thh + Thk; Chris@42: Ths = Tho + Thr; Chris@42: ri[WS(rs, 36)] = Thl - Ths; Chris@42: ri[WS(rs, 4)] = Thl + Ths; Chris@42: Tif = Thu + Thv; Chris@42: Tik = Tig + Tij; Chris@42: ii[WS(rs, 4)] = Tif + Tik; Chris@42: ii[WS(rs, 36)] = Tik - Tif; Chris@42: } Chris@42: { Chris@42: E Tht, Thw, Til, Tim; Chris@42: Tht = Thh - Thk; Chris@42: Thw = Thu - Thv; Chris@42: ri[WS(rs, 52)] = Tht - Thw; Chris@42: ri[WS(rs, 20)] = Tht + Thw; Chris@42: Til = Thr - Tho; Chris@42: Tim = Tij - Tig; Chris@42: ii[WS(rs, 20)] = Til + Tim; Chris@42: ii[WS(rs, 52)] = Tim - Til; Chris@42: } Chris@42: } Chris@42: { Chris@42: E Teb, Tfx, Tey, TiK, TiN, TiT, TfA, TiS, Tfr, TfL, Tfv, TfH, Tf0, TfK, Tfu; Chris@42: E TfE; Chris@42: { Chris@42: E TdZ, Tea, Tfy, Tfz; Chris@42: TdZ = TdV - TdY; Chris@42: Tea = KP707106781 * (Te4 - Te9); Chris@42: Teb = TdZ - Tea; Chris@42: Tfx = TdZ + Tea; Chris@42: { Chris@42: E Tem, Tex, TiL, TiM; Chris@42: Tem = FNMS(KP923879532, Tel, KP382683432 * Teg); Chris@42: Tex = FMA(KP382683432, Ter, KP923879532 * Tew); Chris@42: Tey = Tem - Tex; Chris@42: TiK = Tem + Tex; Chris@42: TiL = KP707106781 * (TfP - TfO); Chris@42: TiM = Tix - Tiw; Chris@42: TiN = TiL + TiM; Chris@42: TiT = TiM - TiL; Chris@42: } Chris@42: Tfy = FMA(KP923879532, Teg, KP382683432 * Tel); Chris@42: Tfz = FNMS(KP923879532, Ter, KP382683432 * Tew); Chris@42: TfA = Tfy + Tfz; Chris@42: TiS = Tfz - Tfy; Chris@42: { Chris@42: E Tfh, TfF, Tfq, TfG, Tfg, Tfp; Chris@42: Tfg = KP707106781 * (Tfa - Tff); Chris@42: Tfh = Tf5 - Tfg; Chris@42: TfF = Tf5 + Tfg; Chris@42: Tfp = KP707106781 * (Tfn - Tfo); Chris@42: Tfq = Tfm - Tfp; Chris@42: TfG = Tfm + Tfp; Chris@42: Tfr = FNMS(KP980785280, Tfq, KP195090322 * Tfh); Chris@42: TfL = FMA(KP831469612, TfG, KP555570233 * TfF); Chris@42: Tfv = FMA(KP195090322, Tfq, KP980785280 * Tfh); Chris@42: TfH = FNMS(KP555570233, TfG, KP831469612 * TfF); Chris@42: } Chris@42: { Chris@42: E TeQ, TfC, TeZ, TfD, TeP, TeY; Chris@42: TeP = KP707106781 * (TeJ - TeO); Chris@42: TeQ = TeE - TeP; Chris@42: TfC = TeE + TeP; Chris@42: TeY = KP707106781 * (TeW - TeX); Chris@42: TeZ = TeV - TeY; Chris@42: TfD = TeV + TeY; Chris@42: Tf0 = FMA(KP980785280, TeQ, KP195090322 * TeZ); Chris@42: TfK = FNMS(KP555570233, TfD, KP831469612 * TfC); Chris@42: Tfu = FNMS(KP980785280, TeZ, KP195090322 * TeQ); Chris@42: TfE = FMA(KP555570233, TfC, KP831469612 * TfD); Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tez, Tfs, TiR, TiU; Chris@42: Tez = Teb + Tey; Chris@42: Tfs = Tf0 + Tfr; Chris@42: ri[WS(rs, 46)] = Tez - Tfs; Chris@42: ri[WS(rs, 14)] = Tez + Tfs; Chris@42: TiR = Tfu + Tfv; Chris@42: TiU = TiS + TiT; Chris@42: ii[WS(rs, 14)] = TiR + TiU; Chris@42: ii[WS(rs, 46)] = TiU - TiR; Chris@42: } Chris@42: { Chris@42: E Tft, Tfw, TiV, TiW; Chris@42: Tft = Teb - Tey; Chris@42: Tfw = Tfu - Tfv; Chris@42: ri[WS(rs, 62)] = Tft - Tfw; Chris@42: ri[WS(rs, 30)] = Tft + Tfw; Chris@42: TiV = Tfr - Tf0; Chris@42: TiW = TiT - TiS; Chris@42: ii[WS(rs, 30)] = TiV + TiW; Chris@42: ii[WS(rs, 62)] = TiW - TiV; Chris@42: } Chris@42: { Chris@42: E TfB, TfI, TiJ, TiO; Chris@42: TfB = Tfx + TfA; Chris@42: TfI = TfE + TfH; Chris@42: ri[WS(rs, 38)] = TfB - TfI; Chris@42: ri[WS(rs, 6)] = TfB + TfI; Chris@42: TiJ = TfK + TfL; Chris@42: TiO = TiK + TiN; Chris@42: ii[WS(rs, 6)] = TiJ + TiO; Chris@42: ii[WS(rs, 38)] = TiO - TiJ; Chris@42: } Chris@42: { Chris@42: E TfJ, TfM, TiP, TiQ; Chris@42: TfJ = Tfx - TfA; Chris@42: TfM = TfK - TfL; Chris@42: ri[WS(rs, 54)] = TfJ - TfM; Chris@42: ri[WS(rs, 22)] = TfJ + TfM; Chris@42: TiP = TfH - TfE; Chris@42: TiQ = TiN - TiK; Chris@42: ii[WS(rs, 22)] = TiP + TiQ; Chris@42: ii[WS(rs, 54)] = TiQ - TiP; Chris@42: } Chris@42: } Chris@42: { Chris@42: E TfR, Tgj, TfY, Tiu, Tiz, TiF, Tgm, TiE, Tgd, Tgx, Tgh, Tgt, Tg6, Tgw, Tgg; Chris@42: E Tgq; Chris@42: { Chris@42: E TfN, TfQ, Tgk, Tgl; Chris@42: TfN = TdV + TdY; Chris@42: TfQ = KP707106781 * (TfO + TfP); Chris@42: TfR = TfN - TfQ; Chris@42: Tgj = TfN + TfQ; Chris@42: { Chris@42: E TfU, TfX, Tiv, Tiy; Chris@42: TfU = FNMS(KP382683432, TfT, KP923879532 * TfS); Chris@42: TfX = FMA(KP923879532, TfV, KP382683432 * TfW); Chris@42: TfY = TfU - TfX; Chris@42: Tiu = TfU + TfX; Chris@42: Tiv = KP707106781 * (Te4 + Te9); Chris@42: Tiy = Tiw + Tix; Chris@42: Tiz = Tiv + Tiy; Chris@42: TiF = Tiy - Tiv; Chris@42: } Chris@42: Tgk = FMA(KP382683432, TfS, KP923879532 * TfT); Chris@42: Tgl = FNMS(KP382683432, TfV, KP923879532 * TfW); Chris@42: Tgm = Tgk + Tgl; Chris@42: TiE = Tgl - Tgk; Chris@42: { Chris@42: E Tg9, Tgr, Tgc, Tgs, Tg8, Tgb; Chris@42: Tg8 = KP707106781 * (Tfo + Tfn); Chris@42: Tg9 = Tg7 - Tg8; Chris@42: Tgr = Tg7 + Tg8; Chris@42: Tgb = KP707106781 * (Tfa + Tff); Chris@42: Tgc = Tga - Tgb; Chris@42: Tgs = Tga + Tgb; Chris@42: Tgd = FNMS(KP831469612, Tgc, KP555570233 * Tg9); Chris@42: Tgx = FMA(KP195090322, Tgr, KP980785280 * Tgs); Chris@42: Tgh = FMA(KP831469612, Tg9, KP555570233 * Tgc); Chris@42: Tgt = FNMS(KP195090322, Tgs, KP980785280 * Tgr); Chris@42: } Chris@42: { Chris@42: E Tg2, Tgo, Tg5, Tgp, Tg1, Tg4; Chris@42: Tg1 = KP707106781 * (TeO + TeJ); Chris@42: Tg2 = Tg0 - Tg1; Chris@42: Tgo = Tg0 + Tg1; Chris@42: Tg4 = KP707106781 * (TeW + TeX); Chris@42: Tg5 = Tg3 - Tg4; Chris@42: Tgp = Tg3 + Tg4; Chris@42: Tg6 = FMA(KP555570233, Tg2, KP831469612 * Tg5); Chris@42: Tgw = FNMS(KP195090322, Tgo, KP980785280 * Tgp); Chris@42: Tgg = FNMS(KP831469612, Tg2, KP555570233 * Tg5); Chris@42: Tgq = FMA(KP980785280, Tgo, KP195090322 * Tgp); Chris@42: } Chris@42: } Chris@42: { Chris@42: E TfZ, Tge, TiD, TiG; Chris@42: TfZ = TfR + TfY; Chris@42: Tge = Tg6 + Tgd; Chris@42: ri[WS(rs, 42)] = TfZ - Tge; Chris@42: ri[WS(rs, 10)] = TfZ + Tge; Chris@42: TiD = Tgg + Tgh; Chris@42: TiG = TiE + TiF; Chris@42: ii[WS(rs, 10)] = TiD + TiG; Chris@42: ii[WS(rs, 42)] = TiG - TiD; Chris@42: } Chris@42: { Chris@42: E Tgf, Tgi, TiH, TiI; Chris@42: Tgf = TfR - TfY; Chris@42: Tgi = Tgg - Tgh; Chris@42: ri[WS(rs, 58)] = Tgf - Tgi; Chris@42: ri[WS(rs, 26)] = Tgf + Tgi; Chris@42: TiH = Tgd - Tg6; Chris@42: TiI = TiF - TiE; Chris@42: ii[WS(rs, 26)] = TiH + TiI; Chris@42: ii[WS(rs, 58)] = TiI - TiH; Chris@42: } Chris@42: { Chris@42: E Tgn, Tgu, Tit, TiA; Chris@42: Tgn = Tgj + Tgm; Chris@42: Tgu = Tgq + Tgt; Chris@42: ri[WS(rs, 34)] = Tgn - Tgu; Chris@42: ri[WS(rs, 2)] = Tgn + Tgu; Chris@42: Tit = Tgw + Tgx; Chris@42: TiA = Tiu + Tiz; Chris@42: ii[WS(rs, 2)] = Tit + TiA; Chris@42: ii[WS(rs, 34)] = TiA - Tit; Chris@42: } Chris@42: { Chris@42: E Tgv, Tgy, TiB, TiC; Chris@42: Tgv = Tgj - Tgm; Chris@42: Tgy = Tgw - Tgx; Chris@42: ri[WS(rs, 50)] = Tgv - Tgy; Chris@42: ri[WS(rs, 18)] = Tgv + Tgy; Chris@42: TiB = Tgt - Tgq; Chris@42: TiC = Tiz - Tiu; Chris@42: ii[WS(rs, 18)] = TiB + TiC; Chris@42: ii[WS(rs, 50)] = TiC - TiB; Chris@42: } Chris@42: } Chris@42: { Chris@42: E T7V, TaH, TjN, TjT, T8O, TjS, TaK, TjK, T9I, TaU, TaE, TaO, TaB, TaV, TaF; Chris@42: E TaR; Chris@42: { Chris@42: E T7x, T7U, TjL, TjM; Chris@42: T7x = T7l - T7w; Chris@42: T7U = T7I - T7T; Chris@42: T7V = T7x - T7U; Chris@42: TaH = T7x + T7U; Chris@42: TjL = TaZ - TaY; Chris@42: TjM = Tjx - Tjw; Chris@42: TjN = TjL + TjM; Chris@42: TjT = TjM - TjL; Chris@42: } Chris@42: { Chris@42: E T8m, TaI, T8N, TaJ; Chris@42: { Chris@42: E T8c, T8l, T8D, T8M; Chris@42: T8c = T80 - T8b; Chris@42: T8l = T8h - T8k; Chris@42: T8m = FNMS(KP980785280, T8l, KP195090322 * T8c); Chris@42: TaI = FMA(KP980785280, T8c, KP195090322 * T8l); Chris@42: T8D = T8r - T8C; Chris@42: T8M = T8I - T8L; Chris@42: T8N = FMA(KP195090322, T8D, KP980785280 * T8M); Chris@42: TaJ = FNMS(KP980785280, T8D, KP195090322 * T8M); Chris@42: } Chris@42: T8O = T8m - T8N; Chris@42: TjS = TaJ - TaI; Chris@42: TaK = TaI + TaJ; Chris@42: TjK = T8m + T8N; Chris@42: } Chris@42: { Chris@42: E T9u, TaM, T9H, TaN; Chris@42: { Chris@42: E T96, T9t, T9D, T9G; Chris@42: T96 = T8U - T95; Chris@42: T9t = T9h - T9s; Chris@42: T9u = T96 - T9t; Chris@42: TaM = T96 + T9t; Chris@42: T9D = T9z - T9C; Chris@42: T9G = T9E - T9F; Chris@42: T9H = T9D - T9G; Chris@42: TaN = T9D + T9G; Chris@42: } Chris@42: T9I = FMA(KP995184726, T9u, KP098017140 * T9H); Chris@42: TaU = FNMS(KP634393284, TaN, KP773010453 * TaM); Chris@42: TaE = FNMS(KP995184726, T9H, KP098017140 * T9u); Chris@42: TaO = FMA(KP634393284, TaM, KP773010453 * TaN); Chris@42: } Chris@42: { Chris@42: E Tan, TaP, TaA, TaQ; Chris@42: { Chris@42: E T9Z, Tam, Taw, Taz; Chris@42: T9Z = T9N - T9Y; Chris@42: Tam = Taa - Tal; Chris@42: Tan = T9Z - Tam; Chris@42: TaP = T9Z + Tam; Chris@42: Taw = Tas - Tav; Chris@42: Taz = Tax - Tay; Chris@42: TaA = Taw - Taz; Chris@42: TaQ = Taw + Taz; Chris@42: } Chris@42: TaB = FNMS(KP995184726, TaA, KP098017140 * Tan); Chris@42: TaV = FMA(KP773010453, TaQ, KP634393284 * TaP); Chris@42: TaF = FMA(KP098017140, TaA, KP995184726 * Tan); Chris@42: TaR = FNMS(KP634393284, TaQ, KP773010453 * TaP); Chris@42: } Chris@42: { Chris@42: E T8P, TaC, TjR, TjU; Chris@42: T8P = T7V + T8O; Chris@42: TaC = T9I + TaB; Chris@42: ri[WS(rs, 47)] = T8P - TaC; Chris@42: ri[WS(rs, 15)] = T8P + TaC; Chris@42: TjR = TaE + TaF; Chris@42: TjU = TjS + TjT; Chris@42: ii[WS(rs, 15)] = TjR + TjU; Chris@42: ii[WS(rs, 47)] = TjU - TjR; Chris@42: } Chris@42: { Chris@42: E TaD, TaG, TjV, TjW; Chris@42: TaD = T7V - T8O; Chris@42: TaG = TaE - TaF; Chris@42: ri[WS(rs, 63)] = TaD - TaG; Chris@42: ri[WS(rs, 31)] = TaD + TaG; Chris@42: TjV = TaB - T9I; Chris@42: TjW = TjT - TjS; Chris@42: ii[WS(rs, 31)] = TjV + TjW; Chris@42: ii[WS(rs, 63)] = TjW - TjV; Chris@42: } Chris@42: { Chris@42: E TaL, TaS, TjJ, TjO; Chris@42: TaL = TaH + TaK; Chris@42: TaS = TaO + TaR; Chris@42: ri[WS(rs, 39)] = TaL - TaS; Chris@42: ri[WS(rs, 7)] = TaL + TaS; Chris@42: TjJ = TaU + TaV; Chris@42: TjO = TjK + TjN; Chris@42: ii[WS(rs, 7)] = TjJ + TjO; Chris@42: ii[WS(rs, 39)] = TjO - TjJ; Chris@42: } Chris@42: { Chris@42: E TaT, TaW, TjP, TjQ; Chris@42: TaT = TaH - TaK; Chris@42: TaW = TaU - TaV; Chris@42: ri[WS(rs, 55)] = TaT - TaW; Chris@42: ri[WS(rs, 23)] = TaT + TaW; Chris@42: TjP = TaR - TaO; Chris@42: TjQ = TjN - TjK; Chris@42: ii[WS(rs, 23)] = TjP + TjQ; Chris@42: ii[WS(rs, 55)] = TjQ - TjP; Chris@42: } Chris@42: } Chris@42: { Chris@42: E TbV, TcT, Tjj, Tjp, Tca, Tjo, TcW, Tjg, Tcu, Td6, TcQ, Td0, TcN, Td7, TcR; Chris@42: E Td3; Chris@42: { Chris@42: E TbN, TbU, Tjh, Tji; Chris@42: TbN = TbJ - TbM; Chris@42: TbU = TbQ - TbT; Chris@42: TbV = TbN - TbU; Chris@42: TcT = TbN + TbU; Chris@42: Tjh = Tdb - Tda; Chris@42: Tji = Tj3 - Tj0; Chris@42: Tjj = Tjh + Tji; Chris@42: Tjp = Tji - Tjh; Chris@42: } Chris@42: { Chris@42: E Tc2, TcU, Tc9, TcV; Chris@42: { Chris@42: E TbY, Tc1, Tc5, Tc8; Chris@42: TbY = TbW - TbX; Chris@42: Tc1 = TbZ - Tc0; Chris@42: Tc2 = FNMS(KP831469612, Tc1, KP555570233 * TbY); Chris@42: TcU = FMA(KP555570233, Tc1, KP831469612 * TbY); Chris@42: Tc5 = Tc3 - Tc4; Chris@42: Tc8 = Tc6 - Tc7; Chris@42: Tc9 = FMA(KP831469612, Tc5, KP555570233 * Tc8); Chris@42: TcV = FNMS(KP831469612, Tc8, KP555570233 * Tc5); Chris@42: } Chris@42: Tca = Tc2 - Tc9; Chris@42: Tjo = TcV - TcU; Chris@42: TcW = TcU + TcV; Chris@42: Tjg = Tc2 + Tc9; Chris@42: } Chris@42: { Chris@42: E Tcm, TcY, Tct, TcZ; Chris@42: { Chris@42: E Tce, Tcl, Tcp, Tcs; Chris@42: Tce = Tcc - Tcd; Chris@42: Tcl = Tch - Tck; Chris@42: Tcm = Tce - Tcl; Chris@42: TcY = Tce + Tcl; Chris@42: Tcp = Tcn - Tco; Chris@42: Tcs = Tcq - Tcr; Chris@42: Tct = Tcp - Tcs; Chris@42: TcZ = Tcp + Tcs; Chris@42: } Chris@42: Tcu = FMA(KP956940335, Tcm, KP290284677 * Tct); Chris@42: Td6 = FNMS(KP471396736, TcZ, KP881921264 * TcY); Chris@42: TcQ = FNMS(KP956940335, Tct, KP290284677 * Tcm); Chris@42: Td0 = FMA(KP471396736, TcY, KP881921264 * TcZ); Chris@42: } Chris@42: { Chris@42: E TcF, Td1, TcM, Td2; Chris@42: { Chris@42: E Tcx, TcE, TcI, TcL; Chris@42: Tcx = Tcv - Tcw; Chris@42: TcE = TcA - TcD; Chris@42: TcF = Tcx - TcE; Chris@42: Td1 = Tcx + TcE; Chris@42: TcI = TcG - TcH; Chris@42: TcL = TcJ - TcK; Chris@42: TcM = TcI - TcL; Chris@42: Td2 = TcI + TcL; Chris@42: } Chris@42: TcN = FNMS(KP956940335, TcM, KP290284677 * TcF); Chris@42: Td7 = FMA(KP881921264, Td2, KP471396736 * Td1); Chris@42: TcR = FMA(KP290284677, TcM, KP956940335 * TcF); Chris@42: Td3 = FNMS(KP471396736, Td2, KP881921264 * Td1); Chris@42: } Chris@42: { Chris@42: E Tcb, TcO, Tjn, Tjq; Chris@42: Tcb = TbV + Tca; Chris@42: TcO = Tcu + TcN; Chris@42: ri[WS(rs, 45)] = Tcb - TcO; Chris@42: ri[WS(rs, 13)] = Tcb + TcO; Chris@42: Tjn = TcQ + TcR; Chris@42: Tjq = Tjo + Tjp; Chris@42: ii[WS(rs, 13)] = Tjn + Tjq; Chris@42: ii[WS(rs, 45)] = Tjq - Tjn; Chris@42: } Chris@42: { Chris@42: E TcP, TcS, Tjr, Tjs; Chris@42: TcP = TbV - Tca; Chris@42: TcS = TcQ - TcR; Chris@42: ri[WS(rs, 61)] = TcP - TcS; Chris@42: ri[WS(rs, 29)] = TcP + TcS; Chris@42: Tjr = TcN - Tcu; Chris@42: Tjs = Tjp - Tjo; Chris@42: ii[WS(rs, 29)] = Tjr + Tjs; Chris@42: ii[WS(rs, 61)] = Tjs - Tjr; Chris@42: } Chris@42: { Chris@42: E TcX, Td4, Tjf, Tjk; Chris@42: TcX = TcT + TcW; Chris@42: Td4 = Td0 + Td3; Chris@42: ri[WS(rs, 37)] = TcX - Td4; Chris@42: ri[WS(rs, 5)] = TcX + Td4; Chris@42: Tjf = Td6 + Td7; Chris@42: Tjk = Tjg + Tjj; Chris@42: ii[WS(rs, 5)] = Tjf + Tjk; Chris@42: ii[WS(rs, 37)] = Tjk - Tjf; Chris@42: } Chris@42: { Chris@42: E Td5, Td8, Tjl, Tjm; Chris@42: Td5 = TcT - TcW; Chris@42: Td8 = Td6 - Td7; Chris@42: ri[WS(rs, 53)] = Td5 - Td8; Chris@42: ri[WS(rs, 21)] = Td5 + Td8; Chris@42: Tjl = Td3 - Td0; Chris@42: Tjm = Tjj - Tjg; Chris@42: ii[WS(rs, 21)] = Tjl + Tjm; Chris@42: ii[WS(rs, 53)] = Tjm - Tjl; Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tdd, TdF, Tj5, Tjb, Tdk, Tja, TdI, TiY, Tds, TdS, TdC, TdM, Tdz, TdT, TdD; Chris@42: E TdP; Chris@42: { Chris@42: E Td9, Tdc, TiZ, Tj4; Chris@42: Td9 = TbJ + TbM; Chris@42: Tdc = Tda + Tdb; Chris@42: Tdd = Td9 - Tdc; Chris@42: TdF = Td9 + Tdc; Chris@42: TiZ = TbQ + TbT; Chris@42: Tj4 = Tj0 + Tj3; Chris@42: Tj5 = TiZ + Tj4; Chris@42: Tjb = Tj4 - TiZ; Chris@42: } Chris@42: { Chris@42: E Tdg, TdG, Tdj, TdH; Chris@42: { Chris@42: E Tde, Tdf, Tdh, Tdi; Chris@42: Tde = TbW + TbX; Chris@42: Tdf = TbZ + Tc0; Chris@42: Tdg = FNMS(KP195090322, Tdf, KP980785280 * Tde); Chris@42: TdG = FMA(KP980785280, Tdf, KP195090322 * Tde); Chris@42: Tdh = Tc3 + Tc4; Chris@42: Tdi = Tc6 + Tc7; Chris@42: Tdj = FMA(KP195090322, Tdh, KP980785280 * Tdi); Chris@42: TdH = FNMS(KP195090322, Tdi, KP980785280 * Tdh); Chris@42: } Chris@42: Tdk = Tdg - Tdj; Chris@42: Tja = TdH - TdG; Chris@42: TdI = TdG + TdH; Chris@42: TiY = Tdg + Tdj; Chris@42: } Chris@42: { Chris@42: E Tdo, TdK, Tdr, TdL; Chris@42: { Chris@42: E Tdm, Tdn, Tdp, Tdq; Chris@42: Tdm = Tcn + Tco; Chris@42: Tdn = Tck + Tch; Chris@42: Tdo = Tdm - Tdn; Chris@42: TdK = Tdm + Tdn; Chris@42: Tdp = Tcc + Tcd; Chris@42: Tdq = Tcq + Tcr; Chris@42: Tdr = Tdp - Tdq; Chris@42: TdL = Tdp + Tdq; Chris@42: } Chris@42: Tds = FMA(KP634393284, Tdo, KP773010453 * Tdr); Chris@42: TdS = FNMS(KP098017140, TdK, KP995184726 * TdL); Chris@42: TdC = FNMS(KP773010453, Tdo, KP634393284 * Tdr); Chris@42: TdM = FMA(KP995184726, TdK, KP098017140 * TdL); Chris@42: } Chris@42: { Chris@42: E Tdv, TdN, Tdy, TdO; Chris@42: { Chris@42: E Tdt, Tdu, Tdw, Tdx; Chris@42: Tdt = Tcv + Tcw; Chris@42: Tdu = TcK + TcJ; Chris@42: Tdv = Tdt - Tdu; Chris@42: TdN = Tdt + Tdu; Chris@42: Tdw = TcG + TcH; Chris@42: Tdx = TcA + TcD; Chris@42: Tdy = Tdw - Tdx; Chris@42: TdO = Tdw + Tdx; Chris@42: } Chris@42: Tdz = FNMS(KP773010453, Tdy, KP634393284 * Tdv); Chris@42: TdT = FMA(KP098017140, TdN, KP995184726 * TdO); Chris@42: TdD = FMA(KP773010453, Tdv, KP634393284 * Tdy); Chris@42: TdP = FNMS(KP098017140, TdO, KP995184726 * TdN); Chris@42: } Chris@42: { Chris@42: E Tdl, TdA, Tj9, Tjc; Chris@42: Tdl = Tdd + Tdk; Chris@42: TdA = Tds + Tdz; Chris@42: ri[WS(rs, 41)] = Tdl - TdA; Chris@42: ri[WS(rs, 9)] = Tdl + TdA; Chris@42: Tj9 = TdC + TdD; Chris@42: Tjc = Tja + Tjb; Chris@42: ii[WS(rs, 9)] = Tj9 + Tjc; Chris@42: ii[WS(rs, 41)] = Tjc - Tj9; Chris@42: } Chris@42: { Chris@42: E TdB, TdE, Tjd, Tje; Chris@42: TdB = Tdd - Tdk; Chris@42: TdE = TdC - TdD; Chris@42: ri[WS(rs, 57)] = TdB - TdE; Chris@42: ri[WS(rs, 25)] = TdB + TdE; Chris@42: Tjd = Tdz - Tds; Chris@42: Tje = Tjb - Tja; Chris@42: ii[WS(rs, 25)] = Tjd + Tje; Chris@42: ii[WS(rs, 57)] = Tje - Tjd; Chris@42: } Chris@42: { Chris@42: E TdJ, TdQ, TiX, Tj6; Chris@42: TdJ = TdF + TdI; Chris@42: TdQ = TdM + TdP; Chris@42: ri[WS(rs, 33)] = TdJ - TdQ; Chris@42: ri[WS(rs, 1)] = TdJ + TdQ; Chris@42: TiX = TdS + TdT; Chris@42: Tj6 = TiY + Tj5; Chris@42: ii[WS(rs, 1)] = TiX + Tj6; Chris@42: ii[WS(rs, 33)] = Tj6 - TiX; Chris@42: } Chris@42: { Chris@42: E TdR, TdU, Tj7, Tj8; Chris@42: TdR = TdF - TdI; Chris@42: TdU = TdS - TdT; Chris@42: ri[WS(rs, 49)] = TdR - TdU; Chris@42: ri[WS(rs, 17)] = TdR + TdU; Chris@42: Tj7 = TdP - TdM; Chris@42: Tj8 = Tj5 - TiY; Chris@42: ii[WS(rs, 17)] = Tj7 + Tj8; Chris@42: ii[WS(rs, 49)] = Tj8 - Tj7; Chris@42: } Chris@42: } Chris@42: { Chris@42: E Tb1, Tbt, Tjz, TjF, Tb8, TjE, Tbw, Tju, Tbg, TbG, Tbq, TbA, Tbn, TbH, Tbr; Chris@42: E TbD; Chris@42: { Chris@42: E TaX, Tb0, Tjv, Tjy; Chris@42: TaX = T7l + T7w; Chris@42: Tb0 = TaY + TaZ; Chris@42: Tb1 = TaX - Tb0; Chris@42: Tbt = TaX + Tb0; Chris@42: Tjv = T7I + T7T; Chris@42: Tjy = Tjw + Tjx; Chris@42: Tjz = Tjv + Tjy; Chris@42: TjF = Tjy - Tjv; Chris@42: } Chris@42: { Chris@42: E Tb4, Tbu, Tb7, Tbv; Chris@42: { Chris@42: E Tb2, Tb3, Tb5, Tb6; Chris@42: Tb2 = T80 + T8b; Chris@42: Tb3 = T8h + T8k; Chris@42: Tb4 = FNMS(KP555570233, Tb3, KP831469612 * Tb2); Chris@42: Tbu = FMA(KP555570233, Tb2, KP831469612 * Tb3); Chris@42: Tb5 = T8r + T8C; Chris@42: Tb6 = T8I + T8L; Chris@42: Tb7 = FMA(KP831469612, Tb5, KP555570233 * Tb6); Chris@42: Tbv = FNMS(KP555570233, Tb5, KP831469612 * Tb6); Chris@42: } Chris@42: Tb8 = Tb4 - Tb7; Chris@42: TjE = Tbv - Tbu; Chris@42: Tbw = Tbu + Tbv; Chris@42: Tju = Tb4 + Tb7; Chris@42: } Chris@42: { Chris@42: E Tbc, Tby, Tbf, Tbz; Chris@42: { Chris@42: E Tba, Tbb, Tbd, Tbe; Chris@42: Tba = T9z + T9C; Chris@42: Tbb = T9s + T9h; Chris@42: Tbc = Tba - Tbb; Chris@42: Tby = Tba + Tbb; Chris@42: Tbd = T8U + T95; Chris@42: Tbe = T9E + T9F; Chris@42: Tbf = Tbd - Tbe; Chris@42: Tbz = Tbd + Tbe; Chris@42: } Chris@42: Tbg = FMA(KP471396736, Tbc, KP881921264 * Tbf); Chris@42: TbG = FNMS(KP290284677, Tby, KP956940335 * Tbz); Chris@42: Tbq = FNMS(KP881921264, Tbc, KP471396736 * Tbf); Chris@42: TbA = FMA(KP956940335, Tby, KP290284677 * Tbz); Chris@42: } Chris@42: { Chris@42: E Tbj, TbB, Tbm, TbC; Chris@42: { Chris@42: E Tbh, Tbi, Tbk, Tbl; Chris@42: Tbh = T9N + T9Y; Chris@42: Tbi = Tay + Tax; Chris@42: Tbj = Tbh - Tbi; Chris@42: TbB = Tbh + Tbi; Chris@42: Tbk = Tas + Tav; Chris@42: Tbl = Taa + Tal; Chris@42: Tbm = Tbk - Tbl; Chris@42: TbC = Tbk + Tbl; Chris@42: } Chris@42: Tbn = FNMS(KP881921264, Tbm, KP471396736 * Tbj); Chris@42: TbH = FMA(KP290284677, TbB, KP956940335 * TbC); Chris@42: Tbr = FMA(KP881921264, Tbj, KP471396736 * Tbm); Chris@42: TbD = FNMS(KP290284677, TbC, KP956940335 * TbB); Chris@42: } Chris@42: { Chris@42: E Tb9, Tbo, TjD, TjG; Chris@42: Tb9 = Tb1 + Tb8; Chris@42: Tbo = Tbg + Tbn; Chris@42: ri[WS(rs, 43)] = Tb9 - Tbo; Chris@42: ri[WS(rs, 11)] = Tb9 + Tbo; Chris@42: TjD = Tbq + Tbr; Chris@42: TjG = TjE + TjF; Chris@42: ii[WS(rs, 11)] = TjD + TjG; Chris@42: ii[WS(rs, 43)] = TjG - TjD; Chris@42: } Chris@42: { Chris@42: E Tbp, Tbs, TjH, TjI; Chris@42: Tbp = Tb1 - Tb8; Chris@42: Tbs = Tbq - Tbr; Chris@42: ri[WS(rs, 59)] = Tbp - Tbs; Chris@42: ri[WS(rs, 27)] = Tbp + Tbs; Chris@42: TjH = Tbn - Tbg; Chris@42: TjI = TjF - TjE; Chris@42: ii[WS(rs, 27)] = TjH + TjI; Chris@42: ii[WS(rs, 59)] = TjI - TjH; Chris@42: } Chris@42: { Chris@42: E Tbx, TbE, Tjt, TjA; Chris@42: Tbx = Tbt + Tbw; Chris@42: TbE = TbA + TbD; Chris@42: ri[WS(rs, 35)] = Tbx - TbE; Chris@42: ri[WS(rs, 3)] = Tbx + TbE; Chris@42: Tjt = TbG + TbH; Chris@42: TjA = Tju + Tjz; Chris@42: ii[WS(rs, 3)] = Tjt + TjA; Chris@42: ii[WS(rs, 35)] = TjA - Tjt; Chris@42: } Chris@42: { Chris@42: E TbF, TbI, TjB, TjC; Chris@42: TbF = Tbt - Tbw; Chris@42: TbI = TbG - TbH; Chris@42: ri[WS(rs, 51)] = TbF - TbI; Chris@42: ri[WS(rs, 19)] = TbF + TbI; Chris@42: TjB = TbD - TbA; Chris@42: TjC = Tjz - Tju; Chris@42: ii[WS(rs, 19)] = TjB + TjC; Chris@42: ii[WS(rs, 51)] = TjC - TjB; Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: } Chris@42: Chris@42: static const tw_instr twinstr[] = { Chris@42: {TW_CEXP, 0, 1}, Chris@42: {TW_CEXP, 0, 3}, Chris@42: {TW_CEXP, 0, 9}, Chris@42: {TW_CEXP, 0, 27}, Chris@42: {TW_CEXP, 0, 63}, Chris@42: {TW_NEXT, 1, 0} Chris@42: }; Chris@42: Chris@42: static const ct_desc desc = { 64, "t2_64", twinstr, &GENUS, {880, 386, 274, 0}, 0, 0, 0 }; Chris@42: Chris@42: void X(codelet_t2_64) (planner *p) { Chris@42: X(kdft_dit_register) (p, t2_64, &desc); Chris@42: } Chris@42: #endif /* HAVE_FMA */