当前位置: 首页>>代码示例>>C++>>正文


C++ VSUB函数代码示例

本文整理汇总了C++中VSUB函数的典型用法代码示例。如果您正苦于以下问题:C++ VSUB函数的具体用法?C++ VSUB怎么用?C++ VSUB使用的例子?那么, 这里精选的函数代码示例或许可以为您提供帮助。


在下文中一共展示了VSUB函数的15个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的C++代码示例。

示例1: buildPhotonMap


//.........这里部分代码省略.........
      printf("Reading %lu at %lu... ", pmap -> heapBufSize, ftell(pmap->heap));
#endif      
      pmap -> heapBufLen = fread(pmap -> heapBuf, sizeof(Photon), 
                                 pmap -> heapBufSize, pmap -> heap);
#ifdef DEBUG_PMAP                                 
      printf("Got %lu\n", pmap -> heapBufLen);
#endif      

      if (ferror(pmap -> heap))
         error(SYSTEM, "failed to read photon heap in buildPhotonMap");

      for (n = pmap -> heapBufLen, p = pmap -> heapBuf; n; n--, p++) {
         /* Update min and max pos and set photon flux */
         for (i = 0; i <= 2; i++) {
            if (p -> pos [i] < pmap -> minPos [i]) 
               pmap -> minPos [i] = p -> pos [i];
            else if (p -> pos [i] > pmap -> maxPos [i]) 
               pmap -> maxPos [i] = p -> pos [i];   

            /* Update centre of gravity with photon position */                 
            CoG [i] += p -> pos [i];                  
         }  
         
         if (primaryOfs)
            /* Linearise photon primary index from subprocess index using the
             * per-subprocess offsets in primaryOfs */
            p -> primary += primaryOfs [p -> proc];
         
         /* Scale photon's flux (hitherto normalised to 1 over RGB); in
          * case of a contrib photon map, this is done per light source,
          * and photonFlux is assumed to be an array */
         getPhotonFlux(p, flux);            

         if (photonFlux) {
            scalecolor(flux, photonFlux [isContribPmap(pmap) ? 
                                            photonSrcIdx(pmap, p) : 0]);
            setPhotonFlux(p, flux);
         }

         /* Update average photon flux; need a double here */
         addcolor(avgFlux, flux);
      }
         
      /* Write modified photons to new heap */
      fwrite(pmap -> heapBuf, sizeof(Photon), pmap -> heapBufLen, nuHeap);
                
      if (ferror(nuHeap))
         error(SYSTEM, "failed postprocessing photon flux in "
               "buildPhotonMap");
      
      nCheck += pmap -> heapBufLen;
   }

#ifdef DEBUG_PMAP
   if (nCheck < pmap -> numPhotons)
      error(INTERNAL, "truncated photon heap in buildPhotonMap");
#endif
   
   /* Finalise average photon flux */
   scalecolor(avgFlux, 1.0 / pmap -> numPhotons);
   copycolor(pmap -> photonFlux, avgFlux);

   /* Average photon positions to get centre of gravity */
   for (i = 0; i < 3; i++)
      pmap -> CoG [i] = CoG [i] /= pmap -> numPhotons;
      
   rewind(pmap -> heap);
   
   /* Compute average photon distance to centre of gravity */
   while (!feof(pmap -> heap)) {
      pmap -> heapBufLen = fread(pmap -> heapBuf, sizeof(Photon), 
                                 pmap -> heapBufSize, pmap -> heap);
      
      for (n = pmap -> heapBufLen, p = pmap -> heapBuf; n; n--, p++) {
         VSUB(d, p -> pos, CoG);
         CoGdist += DOT(d, d);
      }
   }   

   pmap -> CoGdist = CoGdist /= pmap -> numPhotons;

   /* Swap heaps, discarding unscaled photons */
   fclose(pmap -> heap);
   unlink(pmap -> heapFname);
   pmap -> heap = nuHeap;
   strcpy(pmap -> heapFname, nuHeapFname);
   
#ifdef PMAP_OOC
   OOC_BuildPhotonMap(pmap, nproc);
#else
   kdT_BuildPhotonMap(pmap);
#endif

   /* Trash heap and its buffa */
   free(pmap -> heapBuf);
   fclose(pmap -> heap);
   unlink(pmap -> heapFname);
   pmap -> heap = NULL;
   pmap -> heapBuf = NULL;
}
开发者ID:MITSustainableDesignLab,项目名称:Daysim,代码行数:101,代码来源:pmapdata.c

示例2: n2sv_16

static void n2sv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
     INT i;
     for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
	  V T2p, T2q, T2r, T2s, T2x, T2y, T2z, T2A, T1M, T1N, T1L, T1P, T2F, T2G, T2H;
	  V T2I, T1O, T1Q;
	  {
	       V T1l, T1H, T1R, T7, T1x, TN, TC, T25, T1E, T1b, T1Z, Tt, T2h, T22, T1D;
	       V T1g, T1n, TQ, T11, Ti, Te, T26, T1m, TT, T1S, TJ, TZ, T1V, TW, Tl;
	       V T12, T13;
	       {
		    V Tq, T1c, Tp, T20, T1a, Tr, T1d, T1e;
		    {
			 V T1, T2, Tw, Tx, T4, T5, Tz, TA;
			 T1 = LD(&(ri[0]), ivs, &(ri[0]));
			 T2 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0]));
			 Tw = LD(&(ii[0]), ivs, &(ii[0]));
			 Tx = LD(&(ii[WS(is, 8)]), ivs, &(ii[0]));
			 T4 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0]));
			 T5 = LD(&(ri[WS(is, 12)]), ivs, &(ri[0]));
			 Tz = LD(&(ii[WS(is, 4)]), ivs, &(ii[0]));
			 TA = LD(&(ii[WS(is, 12)]), ivs, &(ii[0]));
			 {
			      V Tn, TL, T3, T1k, Ty, T1j, T6, TM, TB, To, T18, T19;
			      Tn = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)]));
			      TL = VSUB(T1, T2);
			      T3 = VADD(T1, T2);
			      T1k = VSUB(Tw, Tx);
			      Ty = VADD(Tw, Tx);
			      T1j = VSUB(T4, T5);
			      T6 = VADD(T4, T5);
			      TM = VSUB(Tz, TA);
			      TB = VADD(Tz, TA);
			      To = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)]));
			      T18 = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)]));
			      T19 = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)]));
			      Tq = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
			      T1l = VADD(T1j, T1k);
			      T1H = VSUB(T1k, T1j);
			      T1R = VSUB(T3, T6);
			      T7 = VADD(T3, T6);
			      T1x = VADD(TL, TM);
			      TN = VSUB(TL, TM);
			      TC = VADD(Ty, TB);
			      T25 = VSUB(Ty, TB);
			      T1c = VSUB(Tn, To);
			      Tp = VADD(Tn, To);
			      T20 = VADD(T18, T19);
			      T1a = VSUB(T18, T19);
			      Tr = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)]));
			      T1d = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
			      T1e = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)]));
			 }
		    }
		    {
			 V Tb, Ta, TF, Tc, TG, TH, TP, TO;
			 {
			      V T8, T9, TD, TE;
			      T8 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
			      T9 = LD(&(ri[WS(is, 10)]), ivs, &(ri[0]));
			      TD = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
			      TE = LD(&(ii[WS(is, 10)]), ivs, &(ii[0]));
			      Tb = LD(&(ri[WS(is, 14)]), ivs, &(ri[0]));
			      {
				   V T17, Ts, T21, T1f;
				   T17 = VSUB(Tq, Tr);
				   Ts = VADD(Tq, Tr);
				   T21 = VADD(T1d, T1e);
				   T1f = VSUB(T1d, T1e);
				   TP = VSUB(T8, T9);
				   Ta = VADD(T8, T9);
				   TO = VSUB(TD, TE);
				   TF = VADD(TD, TE);
				   T1E = VSUB(T1a, T17);
				   T1b = VADD(T17, T1a);
				   T1Z = VSUB(Tp, Ts);
				   Tt = VADD(Tp, Ts);
				   T2h = VADD(T20, T21);
				   T22 = VSUB(T20, T21);
				   T1D = VADD(T1c, T1f);
				   T1g = VSUB(T1c, T1f);
				   Tc = LD(&(ri[WS(is, 6)]), ivs, &(ri[0]));
			      }
			      TG = LD(&(ii[WS(is, 14)]), ivs, &(ii[0]));
			      TH = LD(&(ii[WS(is, 6)]), ivs, &(ii[0]));
			 }
			 T1n = VADD(TP, TO);
			 TQ = VSUB(TO, TP);
			 {
			      V Tg, Th, TX, TR, Td, TS, TI, TY, Tj, Tk;
			      Tg = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
			      Th = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)]));
			      TX = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
			      TR = VSUB(Tb, Tc);
			      Td = VADD(Tb, Tc);
			      TS = VSUB(TG, TH);
			      TI = VADD(TG, TH);
//.........这里部分代码省略.........
开发者ID:exic,项目名称:last.fm-dbus,代码行数:101,代码来源:n2sv_16.c

示例3: DVK

static const R *t1fv_8(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     INT i;
     R *x;
     x = ri;
     for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(ios)) {
	  V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2;
	  T1 = LD(&(x[0]), dist, &(x[0]));
	  T2 = LD(&(x[WS(ios, 4)]), dist, &(x[0]));
	  T3 = BYTWJ(&(W[TWVL * 6]), T2);
	  T4 = VSUB(T1, T3);
	  Tq = VADD(T1, T3);
	  {
	       V Tj, Tl, Ti, Tk;
	       Ti = LD(&(x[WS(ios, 2)]), dist, &(x[0]));
	       Tj = BYTWJ(&(W[TWVL * 2]), Ti);
	       Tk = LD(&(x[WS(ios, 6)]), dist, &(x[0]));
	       Tl = BYTWJ(&(W[TWVL * 10]), Tk);
	       Tm = VSUB(Tj, Tl);
	       Tr = VADD(Tj, Tl);
	  }
	  {
	       V T6, T8, T5, T7;
	       T5 = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)]));
	       T6 = BYTWJ(&(W[0]), T5);
	       T7 = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)]));
	       T8 = BYTWJ(&(W[TWVL * 8]), T7);
	       T9 = VSUB(T6, T8);
	       Tt = VADD(T6, T8);
	  }
	  {
	       V Tb, Td, Ta, Tc;
	       Ta = LD(&(x[WS(ios, 7)]), dist, &(x[WS(ios, 1)]));
	       Tb = BYTWJ(&(W[TWVL * 12]), Ta);
	       Tc = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 1)]));
	       Td = BYTWJ(&(W[TWVL * 4]), Tc);
	       Te = VSUB(Tb, Td);
	       Tu = VADD(Tb, Td);
	  }
	  {
	       V Ts, Tv, Tw, Tx;
	       Ts = VADD(Tq, Tr);
	       Tv = VADD(Tt, Tu);
	       ST(&(x[WS(ios, 4)]), VSUB(Ts, Tv), dist, &(x[0]));
	       ST(&(x[0]), VADD(Ts, Tv), dist, &(x[0]));
	       Tw = VSUB(Tq, Tr);
	       Tx = VBYI(VSUB(Tu, Tt));
	       ST(&(x[WS(ios, 6)]), VSUB(Tw, Tx), dist, &(x[0]));
	       ST(&(x[WS(ios, 2)]), VADD(Tw, Tx), dist, &(x[0]));
	       {
		    V Tg, To, Tn, Tp, Tf, Th;
		    Tf = VMUL(LDK(KP707106781), VADD(T9, Te));
		    Tg = VADD(T4, Tf);
		    To = VSUB(T4, Tf);
		    Th = VMUL(LDK(KP707106781), VSUB(Te, T9));
		    Tn = VBYI(VSUB(Th, Tm));
		    Tp = VBYI(VADD(Tm, Th));
		    ST(&(x[WS(ios, 7)]), VSUB(Tg, Tn), dist, &(x[WS(ios, 1)]));
		    ST(&(x[WS(ios, 3)]), VADD(To, Tp), dist, &(x[WS(ios, 1)]));
		    ST(&(x[WS(ios, 1)]), VADD(Tg, Tn), dist, &(x[WS(ios, 1)]));
		    ST(&(x[WS(ios, 5)]), VSUB(To, Tp), dist, &(x[WS(ios, 1)]));
	       }
	  }
     }
     return W;
}
开发者ID:exic,项目名称:last.fm-dbus,代码行数:67,代码来源:t1fv_8.c

示例4: n1fv_9

static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP342020143, +0.342020143325668733044099614682259580763083368);
     DVK(KP813797681, +0.813797681349373692844693217248393223289101568);
     DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
     DVK(KP296198132, +0.296198132726023843175338011893050938967728390);
     DVK(KP642787609, +0.642787609686539326322643409907263432907559884);
     DVK(KP663413948, +0.663413948168938396205421319635891297216863310);
     DVK(KP556670399, +0.556670399226419366452912952047023132968291906);
     DVK(KP766044443, +0.766044443118978035202392650555416673935832457);
     DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
     DVK(KP150383733, +0.150383733180435296639271897612501926072238258);
     DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
     DVK(KP173648177, +0.173648177666930348851716626769314796000375677);
     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(18, is), MAKE_VOLATILE_STRIDE(18, os)) {
	       V T5, Ts, Tj, To, Tf, Tn, Tp, Tu, Tl, Ta, Tk, Tm, Tt;
	       {
		    V T1, T2, T3, T4;
		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
		    T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    T4 = VADD(T2, T3);
		    T5 = VADD(T1, T4);
		    Ts = VMUL(LDK(KP866025403), VSUB(T3, T2));
		    Tj = VFNMS(LDK(KP500000000), T4, T1);
	       }
	       {
		    V Tb, Te, Tc, Td;
		    Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
		    Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
		    Te = VADD(Tc, Td);
		    To = VSUB(Td, Tc);
		    Tf = VADD(Tb, Te);
		    Tn = VFNMS(LDK(KP500000000), Te, Tb);
		    Tp = VFMA(LDK(KP173648177), Tn, VMUL(LDK(KP852868531), To));
		    Tu = VFNMS(LDK(KP984807753), Tn, VMUL(LDK(KP150383733), To));
	       }
	       {
		    V T6, T9, T7, T8;
		    T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
		    T9 = VADD(T7, T8);
		    Tl = VSUB(T8, T7);
		    Ta = VADD(T6, T9);
		    Tk = VFNMS(LDK(KP500000000), T9, T6);
		    Tm = VFMA(LDK(KP766044443), Tk, VMUL(LDK(KP556670399), Tl));
		    Tt = VFNMS(LDK(KP642787609), Tk, VMUL(LDK(KP663413948), Tl));
	       }
	       {
		    V Ti, Tg, Th, Tz, TA;
		    Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Ta)));
		    Tg = VADD(Ta, Tf);
		    Th = VFNMS(LDK(KP500000000), Tg, T5);
		    ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0]));
		    ST(&(xo[WS(os, 3)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)]));
		    ST(&(xo[WS(os, 6)]), VSUB(Th, Ti), ovs, &(xo[0]));
		    Tz = VFMA(LDK(KP173648177), Tk, VFNMS(LDK(KP296198132), To, VFNMS(LDK(KP939692620), Tn, VFNMS(LDK(KP852868531), Tl, Tj))));
		    TA = VBYI(VSUB(VFNMS(LDK(KP342020143), Tn, VFNMS(LDK(KP150383733), Tl, VFNMS(LDK(KP984807753), Tk, VMUL(LDK(KP813797681), To)))), Ts));
		    ST(&(xo[WS(os, 7)]), VSUB(Tz, TA), ovs, &(xo[WS(os, 1)]));
		    ST(&(xo[WS(os, 2)]), VADD(Tz, TA), ovs, &(xo[0]));
		    {
			 V Tr, Tx, Tw, Ty, Tq, Tv;
			 Tq = VADD(Tm, Tp);
			 Tr = VADD(Tj, Tq);
			 Tx = VFMA(LDK(KP866025403), VSUB(Tt, Tu), VFNMS(LDK(KP500000000), Tq, Tj));
			 Tv = VADD(Tt, Tu);
			 Tw = VBYI(VADD(Ts, Tv));
			 Ty = VBYI(VADD(Ts, VFNMS(LDK(KP500000000), Tv, VMUL(LDK(KP866025403), VSUB(Tp, Tm)))));
			 ST(&(xo[WS(os, 8)]), VSUB(Tr, Tw), ovs, &(xo[0]));
			 ST(&(xo[WS(os, 4)]), VADD(Tx, Ty), ovs, &(xo[0]));
			 ST(&(xo[WS(os, 1)]), VADD(Tw, Tr), ovs, &(xo[WS(os, 1)]));
			 ST(&(xo[WS(os, 5)]), VSUB(Tx, Ty), ovs, &(xo[WS(os, 1)]));
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:SKA-ScienceDataProcessor,项目名称:FastImaging,代码行数:88,代码来源:n1fv_9.c

示例5: n1bv_11

static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
     DVK(KP959492973, +0.959492973614497389890368057066327699062454848);
     DVK(KP654860733, +0.654860733945285064056925072466293553183791199);
     DVK(KP142314838, +0.142314838273285140443792668616369668791051361);
     DVK(KP415415013, +0.415415013001886425529274149229623203524004910);
     DVK(KP841253532, +0.841253532831181168861811648919367717513292498);
     DVK(KP540640817, +0.540640817455597582107635954318691695431770608);
     DVK(KP909631995, +0.909631995354518371411715383079028460060241051);
     DVK(KP989821441, +0.989821441880932732376092037776718787376519372);
     DVK(KP755749574, +0.755749574354258283774035843972344420179717445);
     DVK(KP281732556, +0.281732556841429697711417915346616899035777899);
     int i;
     const R *xi;
     R *xo;
     xi = ii;
     xo = io;
     BEGIN_SIMD();
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
	  V Th, T3, Tm, Tf, Ti, Tc, Tj, T9, Tk, T6, Tl, Ta, Tb, Ts, Tt;
	  Th = LD(&(xi[0]), ivs, &(xi[0]));
	  {
	       V T1, T2, Td, Te;
	       T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
	       T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
	       T3 = VSUB(T1, T2);
	       Tm = VADD(T1, T2);
	       Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
	       Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
	       Tf = VSUB(Td, Te);
	       Ti = VADD(Td, Te);
	  }
	  Ta = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
	  Tb = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
	  Tc = VSUB(Ta, Tb);
	  Tj = VADD(Ta, Tb);
	  {
	       V T7, T8, T4, T5;
	       T7 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
	       T9 = VSUB(T7, T8);
	       Tk = VADD(T7, T8);
	       T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
	       T5 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
	       T6 = VSUB(T4, T5);
	       Tl = VADD(T4, T5);
	  }
	  ST(&(xo[0]), VADD(Th, VADD(Tm, VADD(Ti, VADD(Tl, VADD(Tj, Tk))))), ovs, &(xo[0]));
	  {
	       V Tg, Tn, Tu, Tv;
	       Tg = VBYI(VFMA(LDK(KP281732556), T3, VFMA(LDK(KP755749574), T6, VFNMS(LDK(KP909631995), Tc, VFNMS(LDK(KP540640817), Tf, VMUL(LDK(KP989821441), T9))))));
	       Tn = VFMA(LDK(KP841253532), Ti, VFMA(LDK(KP415415013), Tj, VFNMS(LDK(KP142314838), Tk, VFNMS(LDK(KP654860733), Tl, VFNMS(LDK(KP959492973), Tm, Th)))));
	       ST(&(xo[WS(os, 5)]), VADD(Tg, Tn), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 6)]), VSUB(Tn, Tg), ovs, &(xo[0]));
	       Tu = VBYI(VFMA(LDK(KP755749574), T3, VFMA(LDK(KP540640817), T6, VFNMS(LDK(KP909631995), T9, VFNMS(LDK(KP989821441), Tf, VMUL(LDK(KP281732556), Tc))))));
	       Tv = VFMA(LDK(KP841253532), Tl, VFMA(LDK(KP415415013), Tk, VFNMS(LDK(KP959492973), Tj, VFNMS(LDK(KP142314838), Ti, VFNMS(LDK(KP654860733), Tm, Th)))));
	       ST(&(xo[WS(os, 4)]), VADD(Tu, Tv), ovs, &(xo[0]));
	       ST(&(xo[WS(os, 7)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)]));
	  }
	  Ts = VBYI(VFMA(LDK(KP909631995), T3, VFNMS(LDK(KP540640817), T9, VFNMS(LDK(KP989821441), Tc, VFNMS(LDK(KP281732556), T6, VMUL(LDK(KP755749574), Tf))))));
	  Tt = VFMA(LDK(KP415415013), Tm, VFMA(LDK(KP841253532), Tk, VFNMS(LDK(KP142314838), Tj, VFNMS(LDK(KP959492973), Tl, VFNMS(LDK(KP654860733), Ti, Th)))));
	  ST(&(xo[WS(os, 2)]), VADD(Ts, Tt), ovs, &(xo[0]));
	  ST(&(xo[WS(os, 9)]), VSUB(Tt, Ts), ovs, &(xo[WS(os, 1)]));
	  {
	       V Tq, Tr, To, Tp;
	       Tq = VBYI(VFMA(LDK(KP540640817), T3, VFMA(LDK(KP909631995), Tf, VFMA(LDK(KP989821441), T6, VFMA(LDK(KP755749574), Tc, VMUL(LDK(KP281732556), T9))))));
	       Tr = VFMA(LDK(KP841253532), Tm, VFMA(LDK(KP415415013), Ti, VFNMS(LDK(KP959492973), Tk, VFNMS(LDK(KP654860733), Tj, VFNMS(LDK(KP142314838), Tl, Th)))));
	       ST(&(xo[WS(os, 1)]), VADD(Tq, Tr), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 10)]), VSUB(Tr, Tq), ovs, &(xo[0]));
	       To = VBYI(VFMA(LDK(KP989821441), T3, VFMA(LDK(KP540640817), Tc, VFNMS(LDK(KP909631995), T6, VFNMS(LDK(KP281732556), Tf, VMUL(LDK(KP755749574), T9))))));
	       Tp = VFMA(LDK(KP415415013), Tl, VFMA(LDK(KP841253532), Tj, VFNMS(LDK(KP654860733), Tk, VFNMS(LDK(KP959492973), Ti, VFNMS(LDK(KP142314838), Tm, Th)))));
	       ST(&(xo[WS(os, 3)]), VADD(To, Tp), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 8)]), VSUB(Tp, To), ovs, &(xo[0]));
	  }
     }
     END_SIMD();
}
开发者ID:abrahamneben,项目名称:orbcomm_beam_mapping,代码行数:77,代码来源:n1bv_11.c

示例6: n1bv_7

static void n1bv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
     DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
     DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
     DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
     DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ii;
	  xo = io;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(14, is), MAKE_VOLATILE_STRIDE(14, os)) {
	       V T1, T2, T3, T8, T9, T5, T6;
	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
	       T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
	       T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
	       T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
	       T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
	       T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       {
		    V Tg, T4, Te, Ta, Tf, T7;
		    Tg = VSUB(T2, T3);
		    T4 = VADD(T2, T3);
		    Te = VSUB(T8, T9);
		    Ta = VADD(T8, T9);
		    Tf = VSUB(T5, T6);
		    T7 = VADD(T5, T6);
		    {
			 V Tr, Tj, Tm, Th, To, Tb;
			 Tr = VFMA(LDK(KP554958132), Te, Tg);
			 Tj = VFNMS(LDK(KP356895867), T4, Ta);
			 Tm = VFMA(LDK(KP554958132), Tf, Te);
			 Th = VFNMS(LDK(KP554958132), Tg, Tf);
			 ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0]));
			 To = VFNMS(LDK(KP356895867), T7, T4);
			 Tb = VFNMS(LDK(KP356895867), Ta, T7);
			 {
			      V Ts, Tk, Tn, Ti;
			      Ts = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tr, Tf));
			      Tk = VFNMS(LDK(KP692021471), Tj, T7);
			      Tn = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tm, Tg));
			      Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te));
			      {
				   V Tp, Tc, Tl, Tq, Td;
				   Tp = VFNMS(LDK(KP692021471), To, Ta);
				   Tc = VFNMS(LDK(KP692021471), Tb, T4);
				   Tl = VFNMS(LDK(KP900968867), Tk, T1);
				   Tq = VFNMS(LDK(KP900968867), Tp, T1);
				   Td = VFNMS(LDK(KP900968867), Tc, T1);
				   ST(&(xo[WS(os, 5)]), VFNMSI(Tn, Tl), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 2)]), VFMAI(Tn, Tl), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 6)]), VFNMSI(Ts, Tq), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 1)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 4)]), VFNMSI(Ti, Td), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 3)]), VFMAI(Ti, Td), ovs, &(xo[WS(os, 1)]));
			      }
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:Enny1991,项目名称:ExApp,代码行数:67,代码来源:n1bv_7.c

示例7: t1fv_20

static void t1fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) {
	       V T4, Tx, T1m, T1K, TZ, T16, T17, T10, Tf, Tq, Tr, T1O, T1P, T1Q, T1w;
	       V T1z, T1A, TI, TT, TU, T1L, T1M, T1N, T1p, T1s, T1t, Ts, TV;
	       {
		    V T1, Tw, T3, Tu, Tv, T2, Tt, T1k, T1l;
		    T1 = LD(&(x[0]), ms, &(x[0]));
		    Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
		    Tw = BYTWJ(&(W[TWVL * 28]), Tv);
		    T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
		    T3 = BYTWJ(&(W[TWVL * 18]), T2);
		    Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
		    Tu = BYTWJ(&(W[TWVL * 8]), Tt);
		    T4 = VSUB(T1, T3);
		    Tx = VSUB(Tu, Tw);
		    T1k = VADD(T1, T3);
		    T1l = VADD(Tu, Tw);
		    T1m = VSUB(T1k, T1l);
		    T1K = VADD(T1k, T1l);
	       }
	       {
		    V T9, T1n, TN, T1v, TS, T1y, Te, T1q, Tk, T1u, TC, T1o, TH, T1r, Tp;
		    V T1x;
		    {
			 V T6, T8, T5, T7;
			 T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
			 T6 = BYTWJ(&(W[TWVL * 6]), T5);
			 T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
			 T8 = BYTWJ(&(W[TWVL * 26]), T7);
			 T9 = VSUB(T6, T8);
			 T1n = VADD(T6, T8);
		    }
		    {
			 V TK, TM, TJ, TL;
			 TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
			 TK = BYTWJ(&(W[TWVL * 24]), TJ);
			 TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 TM = BYTWJ(&(W[TWVL * 4]), TL);
			 TN = VSUB(TK, TM);
			 T1v = VADD(TK, TM);
		    }
		    {
			 V TP, TR, TO, TQ;
			 TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
			 TP = BYTWJ(&(W[TWVL * 32]), TO);
			 TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			 TR = BYTWJ(&(W[TWVL * 12]), TQ);
			 TS = VSUB(TP, TR);
			 T1y = VADD(TP, TR);
		    }
		    {
			 V Tb, Td, Ta, Tc;
			 Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
			 Tb = BYTWJ(&(W[TWVL * 30]), Ta);
			 Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 Td = BYTWJ(&(W[TWVL * 10]), Tc);
			 Te = VSUB(Tb, Td);
			 T1q = VADD(Tb, Td);
		    }
		    {
			 V Th, Tj, Tg, Ti;
			 Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
			 Th = BYTWJ(&(W[TWVL * 14]), Tg);
			 Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
			 Tj = BYTWJ(&(W[TWVL * 34]), Ti);
			 Tk = VSUB(Th, Tj);
			 T1u = VADD(Th, Tj);
		    }
		    {
			 V Tz, TB, Ty, TA;
			 Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
			 Tz = BYTWJ(&(W[TWVL * 16]), Ty);
			 TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
			 TB = BYTWJ(&(W[TWVL * 36]), TA);
			 TC = VSUB(Tz, TB);
			 T1o = VADD(Tz, TB);
		    }
		    {
			 V TE, TG, TD, TF;
			 TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 TE = BYTWJ(&(W[0]), TD);
			 TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
			 TG = BYTWJ(&(W[TWVL * 20]), TF);
			 TH = VSUB(TE, TG);
			 T1r = VADD(TE, TG);
		    }
		    {
			 V Tm, To, Tl, Tn;
			 Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
			 Tm = BYTWJ(&(W[TWVL * 22]), Tl);
			 Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
//.........这里部分代码省略.........
开发者ID:DougBurke,项目名称:sherpa,代码行数:101,代码来源:t1fv_20.c

示例8: t3fv_8

static void t3fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs)) {
	       V T2, T3, Ta, T4, Tb, Tc, Tp;
	       T2 = LDW(&(W[0]));
	       T3 = LDW(&(W[TWVL * 2]));
	       Ta = VZMULJ(T2, T3);
	       T4 = VZMUL(T2, T3);
	       Tb = LDW(&(W[TWVL * 4]));
	       Tc = VZMULJ(Ta, Tb);
	       Tp = VZMULJ(T2, Tb);
	       {
		    V T7, Tx, Ts, Ty, Tf, TA, Tk, TB, T1, T6, T5;
		    T1 = LD(&(x[0]), ms, &(x[0]));
		    T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
		    T6 = VZMULJ(T4, T5);
		    T7 = VSUB(T1, T6);
		    Tx = VADD(T1, T6);
		    {
			 V To, Tr, Tn, Tq;
			 Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			 To = VZMULJ(Ta, Tn);
			 Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 Tr = VZMULJ(Tp, Tq);
			 Ts = VSUB(To, Tr);
			 Ty = VADD(To, Tr);
		    }
		    {
			 V T9, Te, T8, Td;
			 T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 T9 = VZMULJ(T2, T8);
			 Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
			 Te = VZMULJ(Tc, Td);
			 Tf = VSUB(T9, Te);
			 TA = VADD(T9, Te);
		    }
		    {
			 V Th, Tj, Tg, Ti;
			 Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			 Th = VZMULJ(Tb, Tg);
			 Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 Tj = VZMULJ(T3, Ti);
			 Tk = VSUB(Th, Tj);
			 TB = VADD(Th, Tj);
		    }
		    {
			 V Tz, TC, TD, TE;
			 Tz = VADD(Tx, Ty);
			 TC = VADD(TA, TB);
			 ST(&(x[WS(rs, 4)]), VSUB(Tz, TC), ms, &(x[0]));
			 ST(&(x[0]), VADD(Tz, TC), ms, &(x[0]));
			 TD = VSUB(Tx, Ty);
			 TE = VSUB(TB, TA);
			 ST(&(x[WS(rs, 6)]), VFNMSI(TE, TD), ms, &(x[0]));
			 ST(&(x[WS(rs, 2)]), VFMAI(TE, TD), ms, &(x[0]));
			 {
			      V Tm, Tv, Tu, Tw, Tl, Tt;
			      Tl = VADD(Tf, Tk);
			      Tm = VFMA(LDK(KP707106781), Tl, T7);
			      Tv = VFNMS(LDK(KP707106781), Tl, T7);
			      Tt = VSUB(Tk, Tf);
			      Tu = VFNMS(LDK(KP707106781), Tt, Ts);
			      Tw = VFMA(LDK(KP707106781), Tt, Ts);
			      ST(&(x[WS(rs, 1)]), VFNMSI(Tu, Tm), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 3)]), VFMAI(Tw, Tv), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 7)]), VFMAI(Tu, Tm), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 5)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)]));
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:DougBurke,项目名称:sherpa,代码行数:79,代码来源:t3fv_8.c

示例9: n1bv_10

static void n1bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ii;
	  xo = io;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) {
	       V Tl, Ty, T7, Te, Tw, Tt, Tz, TA, TB, Tg, Th, Tm, Tj, Tk;
	       Tj = LD(&(xi[0]), ivs, &(xi[0]));
	       Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       Tl = VSUB(Tj, Tk);
	       Ty = VADD(Tj, Tk);
	       {
		    V T3, Tr, Td, Tv, T6, Ts, Ta, Tu;
		    {
			 V T1, T2, Tb, Tc;
			 T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
			 T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
			 T3 = VSUB(T1, T2);
			 Tr = VADD(T1, T2);
			 Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
			 Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
			 Td = VSUB(Tb, Tc);
			 Tv = VADD(Tb, Tc);
		    }
		    {
			 V T4, T5, T8, T9;
			 T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
			 T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
			 T6 = VSUB(T4, T5);
			 Ts = VADD(T4, T5);
			 T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
			 T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
			 Ta = VSUB(T8, T9);
			 Tu = VADD(T8, T9);
		    }
		    T7 = VSUB(T3, T6);
		    Te = VSUB(Ta, Td);
		    Tw = VSUB(Tu, Tv);
		    Tt = VSUB(Tr, Ts);
		    Tz = VADD(Tr, Ts);
		    TA = VADD(Tu, Tv);
		    TB = VADD(Tz, TA);
		    Tg = VADD(T3, T6);
		    Th = VADD(Ta, Td);
		    Tm = VADD(Tg, Th);
	       }
	       ST(&(xo[WS(os, 5)]), VADD(Tl, Tm), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0]));
	       {
		    V Tf, Tq, To, Tp, Ti, Tn;
		    Tf = VBYI(VFMA(LDK(KP951056516), T7, VMUL(LDK(KP587785252), Te)));
		    Tq = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T7)));
		    Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th));
		    Tn = VFNMS(LDK(KP250000000), Tm, Tl);
		    To = VADD(Ti, Tn);
		    Tp = VSUB(Tn, Ti);
		    ST(&(xo[WS(os, 1)]), VADD(Tf, To), ovs, &(xo[WS(os, 1)]));
		    ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)]));
		    ST(&(xo[WS(os, 9)]), VSUB(To, Tf), ovs, &(xo[WS(os, 1)]));
		    ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)]));
	       }
	       {
		    V Tx, TG, TE, TF, TC, TD;
		    Tx = VBYI(VFNMS(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt)));
		    TG = VBYI(VFMA(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Tw)));
		    TC = VFNMS(LDK(KP250000000), TB, Ty);
		    TD = VMUL(LDK(KP559016994), VSUB(Tz, TA));
		    TE = VSUB(TC, TD);
		    TF = VADD(TD, TC);
		    ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0]));
		    ST(&(xo[WS(os, 6)]), VADD(TG, TF), ovs, &(xo[0]));
		    ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0]));
		    ST(&(xo[WS(os, 4)]), VSUB(TF, TG), ovs, &(xo[0]));
	       }
	  }
     }
     VLEAVE();
}
开发者ID:barnex,项目名称:fftw,代码行数:85,代码来源:dft_simd_sse2_n1bv_10.c

示例10: n2bv_16

static void n2bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
     DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     INT i;
     const R *xi;
     R *xo;
     xi = ii;
     xo = io;
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
	  V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW;
	  V T16;
	  {
	       V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj;
	       {
		    V T1, T2, T4, T5, To, Tp, Tr, Ts;
		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
		    T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
		    T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
		    To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    Tp = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Tr = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
		    Ts = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    {
			 V T8, TI, Tq, TJ, Tt, T9, Tb, Tc, T3, T6;
			 T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
			 TH = VSUB(T1, T2);
			 T3 = VADD(T1, T2);
			 TU = VSUB(T4, T5);
			 T6 = VADD(T4, T5);
			 TI = VSUB(To, Tp);
			 Tq = VADD(To, Tp);
			 TJ = VSUB(Tr, Ts);
			 Tt = VADD(Tr, Ts);
			 T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
			 Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
			 Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
			 T7 = VSUB(T3, T6);
			 Tz = VADD(T3, T6);
			 Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
			 TK = VADD(TI, TJ);
			 TV = VSUB(TI, TJ);
			 TA = VADD(Tq, Tt);
			 Tu = VSUB(Tq, Tt);
			 TM = VSUB(T8, T9);
			 Ta = VADD(T8, T9);
			 TN = VSUB(Tb, Tc);
			 Td = VADD(Tb, Tc);
			 Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
			 Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
			 Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
		    }
	       }
	       TF = VADD(Tz, TA);
	       TB = VSUB(Tz, TA);
	       T13 = VFNMS(LDK(KP707106781), TK, TH);
	       TL = VFMA(LDK(KP707106781), TK, TH);
	       TO = VFNMS(LDK(KP414213562), TN, TM);
	       TX = VFMA(LDK(KP414213562), TM, TN);
	       TC = VADD(Ta, Td);
	       Te = VSUB(Ta, Td);
	       TP = VSUB(Tf, Tg);
	       Th = VADD(Tf, Tg);
	       TQ = VSUB(Tj, Ti);
	       Tk = VADD(Ti, Tj);
	       TW = VFMA(LDK(KP707106781), TV, TU);
	       T16 = VFNMS(LDK(KP707106781), TV, TU);
	  }
	  {
	       V TY, TR, Tl, TD;
	       TY = VFMA(LDK(KP414213562), TP, TQ);
	       TR = VFNMS(LDK(KP414213562), TQ, TP);
	       Tl = VSUB(Th, Tk);
	       TD = VADD(Th, Tk);
	       {
		    V TS, T17, TZ, T14;
		    TS = VADD(TO, TR);
		    T17 = VSUB(TO, TR);
		    TZ = VSUB(TX, TY);
		    T14 = VADD(TX, TY);
		    {
			 V TE, TG, Tm, Tv;
			 TE = VSUB(TC, TD);
			 TG = VADD(TC, TD);
			 Tm = VADD(Te, Tl);
			 Tv = VSUB(Te, Tl);
			 {
			      V T18, T1a, TT, T11;
			      T18 = VFMA(LDK(KP923879532), T17, T16);
			      T1a = VFNMS(LDK(KP923879532), T17, T16);
			      TT = VFNMS(LDK(KP923879532), TS, TL);
			      T11 = VFMA(LDK(KP923879532), TS, TL);
			      {
				   V T15, T19, T10, T12;
				   T15 = VFNMS(LDK(KP923879532), T14, T13);
				   T19 = VFMA(LDK(KP923879532), T14, T13);
				   T10 = VFNMS(LDK(KP923879532), TZ, TW);
				   T12 = VFMA(LDK(KP923879532), TZ, TW);
//.........这里部分代码省略.........
开发者ID:exic,项目名称:last.fm-dbus,代码行数:101,代码来源:n2bv_16.c

示例11: n1fv_13

static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
     DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
     DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
     DVK(KP075902986, +0.075902986037193865983102897245103540356428373);
     DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
     DVK(KP132983124, +0.132983124607418643793760531921092974399165133);
     DVK(KP258260390, +0.258260390311744861420450644284508567852516811);
     DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
     DVK(KP300238635, +0.300238635966332641462884626667381504676006424);
     DVK(KP011599105, +0.011599105605768290721655456654083252189827041);
     DVK(KP156891391, +0.156891391051584611046832726756003269660212636);
     DVK(KP256247671, +0.256247671582936600958684654061725059144125175);
     DVK(KP174138601, +0.174138601152135905005660794929264742616964676);
     DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
     DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
     DVK(KP113854479, +0.113854479055790798974654345867655310534642560);
     DVK(KP265966249, +0.265966249214837287587521063842185948798330267);
     DVK(KP387390585, +0.387390585467617292130675966426762851778775217);
     DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     int i;
     const R *xi;
     R *xo;
     xi = ri;
     xo = ro;
     BEGIN_SIMD();
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
	  V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT;
	  TW = LD(&(xi[0]), ivs, &(xi[0]));
	  {
	       V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1;
	       V T2, Tq, Tt;
	       T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       T3 = VSUB(T1, T2);
	       TH = VADD(T1, T2);
	       {
		    V Th, Ti, Tj, Tk;
		    Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
		    Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    Tk = VADD(Ti, Tj);
		    Tl = VADD(Th, Tk);
		    Tw = VSUB(Ti, Tj);
		    Tp = VFNMS(LDK(KP500000000), Tk, Th);
	       }
	       {
		    V Tc, Td, Te, Tf;
		    Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
		    Tf = VADD(Td, Te);
		    Tg = VADD(Tc, Tf);
		    Tv = VSUB(Td, Te);
		    To = VFNMS(LDK(KP500000000), Tf, Tc);
	       }
	       {
		    V T4, T5, T7, T8;
		    T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
		    T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    T6 = VSUB(T4, T5);
		    Tr = VADD(T4, T5);
		    T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
		    T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    T9 = VSUB(T7, T8);
		    Ts = VADD(T7, T8);
	       }
	       Ta = VADD(T6, T9);
	       TI = VADD(Tr, Ts);
	       Tb = VADD(T3, Ta);
	       Tm = VSUB(Tg, Tl);
	       Tq = VSUB(To, Tp);
	       Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts));
	       Tu = VADD(Tq, Tt);
	       TC = VSUB(Tq, Tt);
	       {
		    V TP, TQ, TG, TJ;
		    TP = VADD(Tg, Tl);
		    TQ = VADD(TH, TI);
		    TR = VMUL(LDK(KP300462606), VSUB(TP, TQ));
		    TX = VADD(TP, TQ);
		    TG = VADD(To, Tp);
		    TJ = VFNMS(LDK(KP500000000), TI, TH);
		    TK = VSUB(TG, TJ);
		    TU = VADD(TG, TJ);
	       }
	       {
		    V Tx, Ty, TL, TM;
		    Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw));
		    Ty = VFNMS(LDK(KP500000000), Ta, T3);
		    Tz = VSUB(Tx, Ty);
		    TB = VADD(Tx, Ty);
		    TL = VADD(Tv, Tw);
		    TM = VSUB(T6, T9);
		    TN = VSUB(TL, TM);
		    TT = VADD(TL, TM);
	       }
	  }
//.........这里部分代码省略.........
开发者ID:OS2World,项目名称:MM-SOUND-PM123,代码行数:101,代码来源:n1fv_13.c

示例12: n1fv_12

static void n1fv_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(is), MAKE_VOLATILE_STRIDE(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);
				   ST(&(xo[0]), VADD(TL, TM), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 6)]), VSUB(TL, TM), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 3)]), VFMAI(TK, TH), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 9)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 4)]), VFMAI(TE, TD), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 10)]), VFNMSI(TC, Tz), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 2)]), VFMAI(TC, Tz), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 5)]), VFNMSI(Tw, Tv), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tv), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 11)]), VFMAI(Tu, Tj), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 1)]), VFNMSI(Tu, Tj), ovs, &(xo[WS(os, 1)]));
			      }
			 }
		    }
	       }
	  }
     }
}
开发者ID:BackupTheBerlios,项目名称:openvsipl,代码行数:91,代码来源:n1fv_12.c

示例13: ext_volume_texture


//.........这里部分代码省略.........
 
 if (ry->d.y == 0.0) {
    if ((ry->o.y < bx->min.y) || (ry->o.y > bx->max.y)) return col;
  }
  else {
    ty1 = (bx->min.y - ry->o.y) / ry->d.y;
    ty2 = (bx->max.y - ry->o.y) / ry->d.y;
    if (ty1 > ty2) { a=ty1; ty1=ty2; ty2=a; }
    if (ty1 > tnear) tnear=ty1;
    if (ty2 < tfar)   tfar=ty2;
  }
  if (tnear > tfar) return col;
  if (tfar < 0.0) return col;
 
  if (ry->d.z == 0.0) {
    if ((ry->o.z < bx->min.z) || (ry->o.z > bx->max.z)) return col;
  }
  else {
    tz1 = (bx->min.z - ry->o.z) / ry->d.z;
    tz2 = (bx->max.z - ry->o.z) / ry->d.z;
    if (tz1 > tz2) { a=tz1; tz1=tz2; tz2=a; }
    if (tz1 > tnear) tnear=tz1;
    if (tz2 < tfar)   tfar=tz2;
  }
  if (tnear > tfar) return col;
  if (tfar < 0.0) return col;
 
  if (tnear < 0.0) tnear=0.0;
 
  tdist = xvol->samples;

  tt = (xvol->opacity / tdist); 

  bln.x=fabs(bx->min.x - bx->max.x);
  bln.y=fabs(bx->min.y - bx->max.y);
  bln.z=fabs(bx->min.z - bx->max.z);
  
     dt = 1.0 / tdist; 
    sum = 0.0;

/* Accumulate color as the ray passes through the voxels */
  for (t=tnear; t<=tfar; t+=dt) {
    if (sum < 1.0) {
      pnt.x=((ry->o.x + (ry->d.x * t)) - bx->min.x) / bln.x;
      pnt.y=((ry->o.y + (ry->d.y * t)) - bx->min.y) / bln.y;
      pnt.z=((ry->o.z + (ry->d.z * t)) - bx->min.z) / bln.z;

      /* call external evaluator assume 0.0 -> 1.0 range.. */ 
      scalar = xvol->evaluator(pnt.x, pnt.y, pnt.z);  

      transval = tt * scalar; 
      sum += transval; 

      col2 = ExtVoxelColor(scalar);

      col.r += transval * col2.r * xvol->ambient;
      col.g += transval * col2.g * xvol->ambient;
      col.b += transval * col2.b * xvol->ambient;

      ddt = dt;

      /* Add in diffuse shaded light sources (no shadows) */
      if (xvol->diffuse > 0.0) {
  
        /* Calculate the Volume gradient at the voxel */
        N.x = (xvol->evaluator(pnt.x - ddt, pnt.y, pnt.z)  -  
              xvol->evaluator(pnt.x + ddt, pnt.y, pnt.z)) *  8.0 * tt; 
  
        N.y = (xvol->evaluator(pnt.x, pnt.y - ddt, pnt.z)  -  
              xvol->evaluator(pnt.x, pnt.y + ddt, pnt.z)) *  8.0 * tt; 
  
        N.z = (xvol->evaluator(pnt.x, pnt.y, pnt.z - ddt)  -  
              xvol->evaluator(pnt.x, pnt.y, pnt.z + ddt)) *  8.0 * tt; 
 
        /* only light surfaces with enough of a normal.. */
        if ((N.x*N.x + N.y*N.y + N.z*N.z) > 0.0) { 
          diffint.r = 0.0; 
          diffint.g = 0.0; 
          diffint.b = 0.0; 
    
          /* add the contribution of each of the lights.. */
          for (i=0; i<numlights; i++) {
            li=lightlist[i];
            VSUB(li->ctr, (*hit), L)
            VNorm(&L);
            VDOT(inten, N, L)
    
            /* only add light if its from the front of the surface */
            /* could add back-lighting if we wanted to later.. */
            if (inten > 0.0) {
              diffint.r += inten*li->tex->col.r;
              diffint.g += inten*li->tex->col.g;
              diffint.b += inten*li->tex->col.b;
            }
          }   
          col.r += col2.r * diffint.r * xvol->diffuse;
          col.g += col2.g * diffint.g * xvol->diffuse;
          col.b += col2.b * diffint.b * xvol->diffuse;
        }
      }
开发者ID:Chalenko,项目名称:PCM,代码行数:101,代码来源:extvol.cpp

示例14: n2bv_14

static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
     DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
     DVK(KP222520933, +0.222520933956314404288902564496794759466355569);
     DVK(KP623489801, +0.623489801858733530525004884004239810632274731);
     DVK(KP781831482, +0.781831482468029808708444526674057750232334519);
     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
     DVK(KP433883739, +0.433883739117558120475768332848358754609990728);
     int i;
     const R *xi;
     R *xo;
     xi = ii;
     xo = io;
     BEGIN_SIMD();
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
	  V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn;
	  V To;
	  Tn = LD(&(xi[0]), ivs, &(xi[0]));
	  To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
	  Tp = VSUB(Tn, To);
	  Ty = VADD(Tn, To);
	  {
	       V Th, TC, Tk, TD;
	       {
		    V Tf, Tg, Ti, Tj;
		    Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
		    Th = VSUB(Tf, Tg);
		    TC = VADD(Tf, Tg);
		    Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    Tk = VSUB(Ti, Tj);
		    TD = VADD(Ti, Tj);
	       }
	       Tl = VSUB(Th, Tk);
	       TL = VSUB(TD, TC);
	       Tq = VADD(Th, Tk);
	       TE = VADD(TC, TD);
	  }
	  {
	       V T3, Tz, T6, TA;
	       {
		    V T1, T2, T4, T5;
		    T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
		    T3 = VSUB(T1, T2);
		    Tz = VADD(T1, T2);
		    T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
		    T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
		    T6 = VSUB(T4, T5);
		    TA = VADD(T4, T5);
	       }
	       T7 = VSUB(T3, T6);
	       TJ = VSUB(Tz, TA);
	       Ts = VADD(T3, T6);
	       TB = VADD(Tz, TA);
	  }
	  {
	       V Ta, TF, Td, TG;
	       {
		    V T8, T9, Tb, Tc;
		    T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
		    Ta = VSUB(T8, T9);
		    TF = VADD(T8, T9);
		    Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
		    Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    Td = VSUB(Tb, Tc);
		    TG = VADD(Tb, Tc);
	       }
	       Te = VSUB(Ta, Td);
	       TK = VSUB(TG, TF);
	       Tr = VADD(Ta, Td);
	       TH = VADD(TF, TG);
	  }
	  ST(&(xo[14]), VADD(Tp, VADD(Ts, VADD(Tq, Tr))), ovs, &(xo[2]));
	  ST(&(xo[0]), VADD(Ty, VADD(TB, VADD(TE, TH))), ovs, &(xo[0]));
	  {
	       V Tm, Tt, TQ, TP;
	       Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te))));
	       Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp)));
	       ST(&(xo[6]), VADD(Tm, Tt), ovs, &(xo[2]));
	       ST(&(xo[22]), VSUB(Tt, Tm), ovs, &(xo[2]));
	       TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK))));
	       TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty)));
	       ST(&(xo[24]), VSUB(TP, TQ), ovs, &(xo[0]));
	       ST(&(xo[4]), VADD(TP, TQ), ovs, &(xo[0]));
	  }
	  {
	       V Tu, Tv, TM, TI;
	       Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te))));
	       Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp)));
	       ST(&(xo[2]), VADD(Tu, Tv), ovs, &(xo[2]));
	       ST(&(xo[26]), VSUB(Tv, Tu), ovs, &(xo[2]));
	       TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ))));
	       TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty)));
	       ST(&(xo[12]), VSUB(TI, TM), ovs, &(xo[0]));
	       ST(&(xo[16]), VADD(TI, TM), ovs, &(xo[0]));
	  }
	  {
//.........这里部分代码省略.........
开发者ID:OS2World,项目名称:MM-SOUND-PM123,代码行数:101,代码来源:n2bv_14.c

示例15: n1bv_11

static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP959492973, +0.959492973614497389890368057066327699062454848);
     DVK(KP876768831, +0.876768831002589333891339807079336796764054852);
     DVK(KP918985947, +0.918985947228994779780736114132655398124909697);
     DVK(KP989821441, +0.989821441880932732376092037776718787376519372);
     DVK(KP778434453, +0.778434453334651800608337670740821884709317477);
     DVK(KP830830026, +0.830830026003772851058548298459246407048009821);
     DVK(KP372785597, +0.372785597771792209609773152906148328659002598);
     DVK(KP634356270, +0.634356270682424498893150776899916060542806975);
     DVK(KP715370323, +0.715370323453429719112414662767260662417897278);
     DVK(KP342584725, +0.342584725681637509502641509861112333758894680);
     DVK(KP521108558, +0.521108558113202722944698153526659300680427422);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ii;
	  xo = io;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
	       V T1, Tb, T4, Tq, Tg, Tm, T7, Tp, Ta, To, Tc, T11;
	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
	       {
		    V T2, T3, Te, Tf;
		    T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
		    Tf = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    {
			 V T5, T6, T8, T9;
			 T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
			 T6 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
			 T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
			 T9 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
			 Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
			 T4 = VADD(T2, T3);
			 Tq = VSUB(T2, T3);
			 Tg = VADD(Te, Tf);
			 Tm = VSUB(Te, Tf);
			 T7 = VADD(T5, T6);
			 Tp = VSUB(T5, T6);
			 Ta = VADD(T8, T9);
			 To = VSUB(T8, T9);
			 Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
		    }
	       }
	       T11 = VFMA(LDK(KP521108558), Tm, Tq);
	       {
		    V TA, TS, TE, TW, Td, Tn, Ts, Tw, Tr, Tv, TT, TF;
		    Tr = VFNMS(LDK(KP521108558), Tq, Tp);
		    Tv = VFNMS(LDK(KP342584725), T7, Tg);
		    TA = VFMA(LDK(KP715370323), To, Tq);
		    TS = VFMA(LDK(KP521108558), To, Tm);
		    TE = VFNMS(LDK(KP342584725), T4, Ta);
		    TW = VFNMS(LDK(KP342584725), Ta, T7);
		    Td = VADD(Tb, Tc);
		    Tn = VSUB(Tb, Tc);
		    Ts = VFNMS(LDK(KP715370323), Tr, To);
		    Tw = VFNMS(LDK(KP634356270), Tv, T4);
		    TT = VFNMS(LDK(KP715370323), TS, Tp);
		    TF = VFNMS(LDK(KP634356270), TE, Tg);
		    {
			 V Tu, TV, TD, TL, T14, TP, TZ, Tj, Tz, TI, TB, TJ, TM;
			 TB = VFMA(LDK(KP372785597), Tn, TA);
			 TJ = VFNMS(LDK(KP521108558), Tp, Tn);
			 {
			      V T12, TN, TX, Th;
			      T12 = VFMA(LDK(KP715370323), T11, Tn);
			      ST(&(xo[0]), VADD(Tg, VADD(Td, VADD(Ta, VADD(T7, VADD(T4, T1))))), ovs, &(xo[0]));
			      TN = VFNMS(LDK(KP342584725), Td, T4);
			      TX = VFNMS(LDK(KP634356270), TW, Td);
			      Th = VFNMS(LDK(KP342584725), Tg, Td);
			      {
				   V Tt, Tx, TU, TG;
				   Tt = VFNMS(LDK(KP830830026), Ts, Tn);
				   Tx = VFNMS(LDK(KP778434453), Tw, Ta);
				   TU = VFMA(LDK(KP830830026), TT, Tq);
				   TG = VFNMS(LDK(KP778434453), TF, Td);
				   {
					V TC, TK, T13, TO;
					TC = VFNMS(LDK(KP830830026), TB, Tm);
					TK = VFMA(LDK(KP715370323), TJ, Tm);
					T13 = VFMA(LDK(KP830830026), T12, Tp);
					TO = VFNMS(LDK(KP634356270), TN, T7);
					{
					     V TY, Ti, Ty, TH;
					     TY = VFNMS(LDK(KP778434453), TX, T4);
					     Ti = VFNMS(LDK(KP634356270), Th, Ta);
					     Tu = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), Tt, Tm));
					     Ty = VFNMS(LDK(KP876768831), Tx, Td);
					     TV = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TU, Tn));
					     TH = VFNMS(LDK(KP876768831), TG, T7);
					     TD = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), TC, Tp));
					     TL = VFNMS(LDK(KP830830026), TK, To);
					     T14 = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), T13, To));
					     TP = VFNMS(LDK(KP778434453), TO, Tg);
					     TZ = VFNMS(LDK(KP876768831), TY, Tg);
					     Tj = VFNMS(LDK(KP778434453), Ti, T7);
					     Tz = VFNMS(LDK(KP959492973), Ty, T1);
					     TI = VFNMS(LDK(KP959492973), TH, T1);
//.........这里部分代码省略.........
开发者ID:dstuck,项目名称:tinker_integrated_PIMC,代码行数:101,代码来源:n1bv_11.c


注:本文中的VSUB函数示例由纯净天空整理自Github/MSDocs等开源代码及文档管理平台,相关代码片段筛选自各路编程大神贡献的开源项目,源码版权归原作者所有,传播和使用请参考对应项目的License;未经允许,请勿转载。