diff Lib/fftw-3.2.1/cell/spu/spu_t1fv_12.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_12.spuc	Thu Jul 09 01:12:16 2015 +0100
@@ -0,0 +1,125 @@
+/*
+ * 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 12 -name X(spu_t1fv_12) */
+
+/*
+ * This function contains 59 FP additions, 42 FP multiplications,
+ * (or, 41 additions, 24 multiplications, 18 fused multiply/add),
+ * 75 stack variables, 2 constants, and 24 memory accesses
+ */
+#include "fftw-spu.h"
+
+void X(spu_t1fv_12) (R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) {
+     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
+     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
+     INT m;
+     R *x;
+     x = ri;
+     for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(rs)) {
+	  V TY, T7, Tf, TZ, TE, TQ, T11, Tw, TA, T12, Tr, TP, T1, T3, Td;
+	  V Tb, T9, T5, T2, Tc, Ta, T8, T4, TC, T6, TD, Te, Ti, Tk, Ty;
+	  V Tu, Tp, Tn, Th, Tj, Tx, Tt, To, Tm, Tl, Tv, Tq, Tz, T16, T17;
+	  V T14, T15, T10, T13, T18, T19, TM, TN, TI, Ts, TF, TJ, Tg, TB, TH;
+	  V TK, TG, TL, TR, TV, TO, TU, TT, TW, TS, TX;
+	  T1 = LD(&(x[0]), ms, &(x[0]));
+	  T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
+	  T3 = BYTWJ(&(W[TWVL * 6]), T2);
+	  Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
+	  Td = BYTWJ(&(W[TWVL * 2]), Tc);
+	  Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
+	  Tb = BYTWJ(&(W[TWVL * 18]), Ta);
+	  T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
+	  T9 = BYTWJ(&(W[TWVL * 10]), T8);
+	  T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
+	  T5 = BYTWJ(&(W[TWVL * 14]), T4);
+	  T6 = VADD(T3, T5);
+	  TC = VSUB(T5, T3);
+	  TY = VADD(T1, T6);
+	  T7 = VFNMS(LDK(KP500000000), T6, T1);
+	  TD = VSUB(Td, Tb);
+	  Te = VADD(Tb, Td);
+	  Tf = VFNMS(LDK(KP500000000), Te, T9);
+	  TZ = VADD(T9, Te);
+	  TE = VSUB(TC, TD);
+	  TQ = VADD(TC, TD);
+	  Th = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
+	  Ti = BYTWJ(&(W[TWVL * 20]), Th);
+	  Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
+	  Tk = BYTWJ(&(W[TWVL * 12]), Tj);
+	  Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
+	  Ty = BYTWJ(&(W[TWVL * 16]), Tx);
+	  Tt = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
+	  Tu = BYTWJ(&(W[TWVL * 4]), Tt);
+	  To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
+	  Tp = BYTWJ(&(W[TWVL * 8]), To);
+	  Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
+	  Tn = BYTWJ(&(W[0]), Tm);
+	  Tv = VADD(Tk, Ti);
+	  Tl = VSUB(Ti, Tk);
+	  T11 = VADD(Tu, Tv);
+	  Tw = VFNMS(LDK(KP500000000), Tv, Tu);
+	  Tq = VSUB(Tn, Tp);
+	  Tz = VADD(Tn, Tp);
+	  TA = VFNMS(LDK(KP500000000), Tz, Ty);
+	  T12 = VADD(Ty, Tz);
+	  Tr = VADD(Tl, Tq);
+	  TP = VSUB(Tl, Tq);
+	  T10 = VSUB(TY, TZ);
+	  T16 = VADD(TY, TZ);
+	  T17 = VADD(T11, T12);
+	  T13 = VSUB(T11, T12);
+	  T14 = VFNMSI(T13, T10);
+	  T15 = VFMAI(T13, T10);
+	  ST(&(x[WS(rs, 9)]), T14, ms, &(x[WS(rs, 1)]));
+	  ST(&(x[WS(rs, 3)]), T15, ms, &(x[WS(rs, 1)]));
+	  T18 = VSUB(T16, T17);
+	  T19 = VADD(T16, T17);
+	  ST(&(x[WS(rs, 6)]), T18, ms, &(x[0]));
+	  ST(&(x[0]), T19, ms, &(x[0]));
+	  Tg = VSUB(T7, Tf);
+	  TM = VADD(T7, Tf);
+	  TI = VFNMS(LDK(KP866025403), Tr, Tg);
+	  Ts = VFMA(LDK(KP866025403), Tr, Tg);
+	  TN = VADD(Tw, TA);
+	  TB = VSUB(Tw, TA);
+	  TF = VFNMS(LDK(KP866025403), TE, TB);
+	  TJ = VFMA(LDK(KP866025403), TE, TB);
+	  TG = VFNMSI(TF, Ts);
+	  TH = VFMAI(TF, Ts);
+	  ST(&(x[WS(rs, 1)]), TG, ms, &(x[WS(rs, 1)]));
+	  TL = VFMAI(TJ, TI);
+	  TK = VFNMSI(TJ, TI);
+	  ST(&(x[WS(rs, 7)]), TL, ms, &(x[WS(rs, 1)]));
+	  ST(&(x[WS(rs, 11)]), TH, ms, &(x[WS(rs, 1)]));
+	  ST(&(x[WS(rs, 5)]), TK, ms, &(x[WS(rs, 1)]));
+	  TR = VMUL(LDK(KP866025403), VSUB(TP, TQ));
+	  TV = VMUL(LDK(KP866025403), VADD(TQ, TP));
+	  TO = VSUB(TM, TN);
+	  TU = VADD(TM, TN);
+	  TS = VFMAI(TR, TO);
+	  TT = VFNMSI(TR, TO);
+	  ST(&(x[WS(rs, 2)]), TS, ms, &(x[0]));
+	  TX = VFNMSI(TV, TU);
+	  TW = VFMAI(TV, TU);
+	  ST(&(x[WS(rs, 8)]), TX, ms, &(x[0]));
+	  ST(&(x[WS(rs, 10)]), TT, ms, &(x[0]));
+	  ST(&(x[WS(rs, 4)]), TW, ms, &(x[0]));
+     }
+}