diff Lib/fftw-3.2.1/cell/spu/spu_t1fv_7.spuc @ 0:25bf17994ef1

First commit. VS2013, Codeblocks and Mac OSX configuration

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Lib/fftw-3.2.1/cell/spu/spu_t1fv_7.spuc	Thu Jul 09 01:12:16 2015 +0100
@@ -0,0 +1,92 @@
+/*
+ * Copyright (c) 2003, 2007-8 Matteo Frigo
+ * Copyright (c) 2003, 2007-8 Massachusetts Institute of Technology
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+/* Generated by: ../../genfft/gen_twiddle_c -standalone -fma -reorder-insns -simd -compact -variables 100000 -include fftw-spu.h -trivial-stores -n 7 -name X(spu_t1fv_7) */
+
+/*
+ * This function contains 36 FP additions, 36 FP multiplications,
+ * (or, 15 additions, 15 multiplications, 21 fused multiply/add),
+ * 49 stack variables, 6 constants, and 14 memory accesses
+ */
+#include "fftw-spu.h"
+
+void X(spu_t1fv_7) (R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) {
+     DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
+     DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
+     DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
+     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
+     DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
+     DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
+     INT m;
+     R *x;
+     x = ri;
+     for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(rs)) {
+	  V T1, T6, Tl, Tm, Tg, Tb, Tn, Ti, Tz, Ts, To, TC, Tv, T3, T5;
+	  V Tf, Td, Ta, T8, T2, T4, Te, Tc, T9, T7, Th, TD, TB, TA, TF;
+	  V TE, Tp, Tk, Tj, Tr, Tq, Tw, Tu, Tt, Ty, Tx;
+	  T1 = LD(&(x[0]), ms, &(x[0]));
+	  T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
+	  T3 = BYTWJ(&(W[0]), T2);
+	  T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
+	  T5 = BYTWJ(&(W[TWVL * 10]), T4);
+	  Te = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
+	  Tf = BYTWJ(&(W[TWVL * 6]), Te);
+	  Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
+	  Td = BYTWJ(&(W[TWVL * 4]), Tc);
+	  T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
+	  Ta = BYTWJ(&(W[TWVL * 8]), T9);
+	  T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
+	  T8 = BYTWJ(&(W[TWVL * 2]), T7);
+	  T6 = VADD(T3, T5);
+	  Tl = VSUB(T5, T3);
+	  Tm = VSUB(Tf, Td);
+	  Tg = VADD(Td, Tf);
+	  Tb = VADD(T8, Ta);
+	  Tn = VSUB(Ta, T8);
+	  Ti = VFNMS(LDK(KP356895867), T6, Tg);
+	  Tz = VFNMS(LDK(KP356895867), Tg, Tb);
+	  Ts = VFNMS(LDK(KP356895867), Tb, T6);
+	  To = VFMA(LDK(KP554958132), Tn, Tm);
+	  TC = VFNMS(LDK(KP554958132), Tl, Tn);
+	  Tv = VFMA(LDK(KP554958132), Tm, Tl);
+	  Th = VADD(T1, VADD(T6, VADD(Tb, Tg)));
+	  ST(&(x[0]), Th, ms, &(x[0]));
+	  TD = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TC, Tm));
+	  TA = VFNMS(LDK(KP692021471), Tz, T6);
+	  TB = VFNMS(LDK(KP900968867), TA, T1);
+	  TF = VFMAI(TD, TB);
+	  TE = VFNMSI(TD, TB);
+	  ST(&(x[WS(rs, 4)]), TE, ms, &(x[0]));
+	  ST(&(x[WS(rs, 3)]), TF, ms, &(x[WS(rs, 1)]));
+	  Tp = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), To, Tl));
+	  Tj = VFNMS(LDK(KP692021471), Ti, Tb);
+	  Tk = VFNMS(LDK(KP900968867), Tj, T1);
+	  Tq = VFNMSI(Tp, Tk);
+	  Tr = VFMAI(Tp, Tk);
+	  ST(&(x[WS(rs, 5)]), Tq, ms, &(x[WS(rs, 1)]));
+	  ST(&(x[WS(rs, 2)]), Tr, ms, &(x[0]));
+	  Tw = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tv, Tn));
+	  Tt = VFNMS(LDK(KP692021471), Ts, Tg);
+	  Tu = VFNMS(LDK(KP900968867), Tt, T1);
+	  Ty = VFMAI(Tw, Tu);
+	  Tx = VFNMSI(Tw, Tu);
+	  ST(&(x[WS(rs, 6)]), Tx, ms, &(x[0]));
+	  ST(&(x[WS(rs, 1)]), Ty, ms, &(x[WS(rs, 1)]));
+     }
+}