diff src/fftw-3.3.5/dft/simd/common/n2fv_12.c @ 42:2cd0e3b3e1fd

Current fftw source

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/src/fftw-3.3.5/dft/simd/common/n2fv_12.c	Tue Oct 18 13:40:26 2016 +0100
@@ -0,0 +1,304 @@
+/*
+ * Copyright (c) 2003, 2007-14 Matteo Frigo
+ * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+/* This file was automatically generated --- DO NOT EDIT */
+/* Generated on Sat Jul 30 16:40:09 EDT 2016 */
+
+#include "codelet-dft.h"
+
+#ifdef HAVE_FMA
+
+/* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n2fv_12 -with-ostride 2 -include n2f.h -store-multiple 2 */
+
+/*
+ * This function contains 48 FP additions, 20 FP multiplications,
+ * (or, 30 additions, 2 multiplications, 18 fused multiply/add),
+ * 61 stack variables, 2 constants, and 30 memory accesses
+ */
+#include "n2f.h"
+
+static void n2fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
+{
+     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
+     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
+     {
+	  INT i;
+	  const R *xi;
+	  R *xo;
+	  xi = ri;
+	  xo = ro;
+	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) {
+	       V T1, T6, Tk, Tn, Tc, Td, Tf, Tr, T4, Ts, T9, Tg, Te, Tl;
+	       {
+		    V T2, T3, T7, T8;
+		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
+		    T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
+		    T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
+		    T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
+		    T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
+		    T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
+		    Tk = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
+		    Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
+		    Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
+		    Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
+		    Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
+		    Tr = VSUB(T3, T2);
+		    T4 = VADD(T2, T3);
+		    Ts = VSUB(T8, T7);
+		    T9 = VADD(T7, T8);
+		    Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
+	       }
+	       Te = VSUB(Tc, Td);
+	       Tl = VADD(Td, Tc);
+	       {
+		    V T5, TF, TB, Tt, Ta, TG, Th, To, Tm, TI;
+		    T5 = VFNMS(LDK(KP500000000), T4, T1);
+		    TF = VADD(T1, T4);
+		    TB = VADD(Tr, Ts);
+		    Tt = VSUB(Tr, Ts);
+		    Ta = VFNMS(LDK(KP500000000), T9, T6);
+		    TG = VADD(T6, T9);
+		    Th = VSUB(Tf, Tg);
+		    To = VADD(Tf, Tg);
+		    Tm = VFNMS(LDK(KP500000000), Tl, Tk);
+		    TI = VADD(Tk, Tl);
+		    {
+			 V TH, TL, Tb, Tx, TJ, Tp, Ti, TA;
+			 TH = VSUB(TF, TG);
+			 TL = VADD(TF, TG);
+			 Tb = VSUB(T5, Ta);
+			 Tx = VADD(T5, Ta);
+			 TJ = VADD(Tn, To);
+			 Tp = VFNMS(LDK(KP500000000), To, Tn);
+			 Ti = VADD(Te, Th);
+			 TA = VSUB(Te, Th);
+			 {
+			      V Tq, Ty, TK, TM;
+			      Tq = VSUB(Tm, Tp);
+			      Ty = VADD(Tm, Tp);
+			      TK = VSUB(TI, TJ);
+			      TM = VADD(TI, TJ);
+			      {
+				   V TC, TE, Tj, Tv;
+				   TC = VMUL(LDK(KP866025403), VSUB(TA, TB));
+				   TE = VMUL(LDK(KP866025403), VADD(TB, TA));
+				   Tj = VFMA(LDK(KP866025403), Ti, Tb);
+				   Tv = VFNMS(LDK(KP866025403), Ti, Tb);
+				   {
+					V Tz, TD, Tu, Tw;
+					Tz = VSUB(Tx, Ty);
+					TD = VADD(Tx, Ty);
+					Tu = VFNMS(LDK(KP866025403), Tt, Tq);
+					Tw = VFMA(LDK(KP866025403), Tt, Tq);
+					{
+					     V TN, TO, TP, TQ;
+					     TN = VADD(TL, TM);
+					     STM2(&(xo[0]), TN, ovs, &(xo[0]));
+					     TO = VSUB(TL, TM);
+					     STM2(&(xo[12]), TO, ovs, &(xo[0]));
+					     TP = VFMAI(TK, TH);
+					     STM2(&(xo[6]), TP, ovs, &(xo[2]));
+					     TQ = VFNMSI(TK, TH);
+					     STM2(&(xo[18]), TQ, ovs, &(xo[2]));
+					     {
+						  V TR, TS, TT, TU;
+						  TR = VFMAI(TE, TD);
+						  STM2(&(xo[8]), TR, ovs, &(xo[0]));
+						  TS = VFNMSI(TE, TD);
+						  STM2(&(xo[16]), TS, ovs, &(xo[0]));
+						  STN2(&(xo[16]), TS, TQ, ovs);
+						  TT = VFNMSI(TC, Tz);
+						  STM2(&(xo[20]), TT, ovs, &(xo[0]));
+						  TU = VFMAI(TC, Tz);
+						  STM2(&(xo[4]), TU, ovs, &(xo[0]));
+						  STN2(&(xo[4]), TU, TP, ovs);
+						  {
+						       V TV, TW, TX, TY;
+						       TV = VFNMSI(Tw, Tv);
+						       STM2(&(xo[10]), TV, ovs, &(xo[2]));
+						       STN2(&(xo[8]), TR, TV, ovs);
+						       TW = VFMAI(Tw, Tv);
+						       STM2(&(xo[14]), TW, ovs, &(xo[2]));
+						       STN2(&(xo[12]), TO, TW, ovs);
+						       TX = VFMAI(Tu, Tj);
+						       STM2(&(xo[22]), TX, ovs, &(xo[2]));
+						       STN2(&(xo[20]), TT, TX, ovs);
+						       TY = VFNMSI(Tu, Tj);
+						       STM2(&(xo[2]), TY, ovs, &(xo[2]));
+						       STN2(&(xo[0]), TN, TY, ovs);
+						  }
+					     }
+					}
+				   }
+			      }
+			 }
+		    }
+	       }
+	  }
+     }
+     VLEAVE();
+}
+
+static const kdft_desc desc = { 12, XSIMD_STRING("n2fv_12"), {30, 2, 18, 0}, &GENUS, 0, 2, 0, 0 };
+
+void XSIMD(codelet_n2fv_12) (planner *p) {
+     X(kdft_register) (p, n2fv_12, &desc);
+}
+
+#else				/* HAVE_FMA */
+
+/* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n2fv_12 -with-ostride 2 -include n2f.h -store-multiple 2 */
+
+/*
+ * This function contains 48 FP additions, 8 FP multiplications,
+ * (or, 44 additions, 4 multiplications, 4 fused multiply/add),
+ * 33 stack variables, 2 constants, and 30 memory accesses
+ */
+#include "n2f.h"
+
+static void n2fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
+{
+     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
+     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
+     {
+	  INT i;
+	  const R *xi;
+	  R *xo;
+	  xi = ri;
+	  xo = ro;
+	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) {
+	       V T5, Ta, TJ, Ty, Tq, Tp, Tg, Tl, TI, TA, Tz, Tu;
+	       {
+		    V T1, T6, T4, Tw, T9, Tx;
+		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
+		    T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
+		    {
+			 V T2, T3, T7, T8;
+			 T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
+			 T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
+			 T4 = VADD(T2, T3);
+			 Tw = VSUB(T3, T2);
+			 T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
+			 T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
+			 T9 = VADD(T7, T8);
+			 Tx = VSUB(T8, T7);
+		    }
+		    T5 = VADD(T1, T4);
+		    Ta = VADD(T6, T9);
+		    TJ = VADD(Tw, Tx);
+		    Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx));
+		    Tq = VFNMS(LDK(KP500000000), T9, T6);
+		    Tp = VFNMS(LDK(KP500000000), T4, T1);
+	       }
+	       {
+		    V Tc, Th, Tf, Ts, Tk, Tt;
+		    Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
+		    Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
+		    {
+			 V Td, Te, Ti, Tj;
+			 Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
+			 Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
+			 Tf = VADD(Td, Te);
+			 Ts = VSUB(Te, Td);
+			 Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
+			 Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
+			 Tk = VADD(Ti, Tj);
+			 Tt = VSUB(Tj, Ti);
+		    }
+		    Tg = VADD(Tc, Tf);
+		    Tl = VADD(Th, Tk);
+		    TI = VADD(Ts, Tt);
+		    TA = VFNMS(LDK(KP500000000), Tk, Th);
+		    Tz = VFNMS(LDK(KP500000000), Tf, Tc);
+		    Tu = VMUL(LDK(KP866025403), VSUB(Ts, Tt));
+	       }
+	       {
+		    V TN, TO, TP, TQ, TR, TS;
+		    {
+			 V Tb, Tm, Tn, To;
+			 Tb = VSUB(T5, Ta);
+			 Tm = VBYI(VSUB(Tg, Tl));
+			 TN = VSUB(Tb, Tm);
+			 STM2(&(xo[18]), TN, ovs, &(xo[2]));
+			 TO = VADD(Tb, Tm);
+			 STM2(&(xo[6]), TO, ovs, &(xo[2]));
+			 Tn = VADD(T5, Ta);
+			 To = VADD(Tg, Tl);
+			 TP = VSUB(Tn, To);
+			 STM2(&(xo[12]), TP, ovs, &(xo[0]));
+			 TQ = VADD(Tn, To);
+			 STM2(&(xo[0]), TQ, ovs, &(xo[0]));
+		    }
+		    {
+			 V Tv, TE, TC, TD, Tr, TB, TT, TU;
+			 Tr = VSUB(Tp, Tq);
+			 Tv = VSUB(Tr, Tu);
+			 TE = VADD(Tr, Tu);
+			 TB = VSUB(Tz, TA);
+			 TC = VBYI(VADD(Ty, TB));
+			 TD = VBYI(VSUB(Ty, TB));
+			 TR = VSUB(Tv, TC);
+			 STM2(&(xo[10]), TR, ovs, &(xo[2]));
+			 TS = VSUB(TE, TD);
+			 STM2(&(xo[22]), TS, ovs, &(xo[2]));
+			 TT = VADD(TC, Tv);
+			 STM2(&(xo[14]), TT, ovs, &(xo[2]));
+			 STN2(&(xo[12]), TP, TT, ovs);
+			 TU = VADD(TD, TE);
+			 STM2(&(xo[2]), TU, ovs, &(xo[2]));
+			 STN2(&(xo[0]), TQ, TU, ovs);
+		    }
+		    {
+			 V TK, TM, TH, TL, TF, TG;
+			 TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ)));
+			 TM = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI)));
+			 TF = VADD(Tp, Tq);
+			 TG = VADD(Tz, TA);
+			 TH = VSUB(TF, TG);
+			 TL = VADD(TF, TG);
+			 {
+			      V TV, TW, TX, TY;
+			      TV = VSUB(TH, TK);
+			      STM2(&(xo[20]), TV, ovs, &(xo[0]));
+			      STN2(&(xo[20]), TV, TS, ovs);
+			      TW = VADD(TL, TM);
+			      STM2(&(xo[8]), TW, ovs, &(xo[0]));
+			      STN2(&(xo[8]), TW, TR, ovs);
+			      TX = VADD(TH, TK);
+			      STM2(&(xo[4]), TX, ovs, &(xo[0]));
+			      STN2(&(xo[4]), TX, TO, ovs);
+			      TY = VSUB(TL, TM);
+			      STM2(&(xo[16]), TY, ovs, &(xo[0]));
+			      STN2(&(xo[16]), TY, TN, ovs);
+			 }
+		    }
+	       }
+	  }
+     }
+     VLEAVE();
+}
+
+static const kdft_desc desc = { 12, XSIMD_STRING("n2fv_12"), {44, 4, 4, 0}, &GENUS, 0, 2, 0, 0 };
+
+void XSIMD(codelet_n2fv_12) (planner *p) {
+     X(kdft_register) (p, n2fv_12, &desc);
+}
+
+#endif				/* HAVE_FMA */