nel/src/3d/ps_attrib_maker_template.cpp

   1 // NeL - MMORPG Framework <http://dev.ryzom.com/projects/nel/>
   2 // Copyright (C) 2010  Winch Gate Property Limited
   3 //
   4 // This program is free software: you can redistribute it and/or modify
   5 // it under the terms of the GNU Affero General Public License as
   6 // published by the Free Software Foundation, either version 3 of the
   7 // License, or (at your option) any later version.
   8 //
   9 // This program is distributed in the hope that it will be useful,
  10 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 // GNU Affero General Public License for more details.
  13 //
  14 // You should have received a copy of the GNU Affero General Public License
  15 // along with this program.  If not, see <http://www.gnu.org/licenses/>.
  16
  17 #include "std3d.h"
  18 #include "nel/3d/ps_attrib_maker_template.h"
  19 #include "nel/misc/system_info.h"
  20
  21 #ifdef DEBUG_NEW
  22 #define new DEBUG_NEW
  23 #endif
  24
  25 namespace NL3D
  26 {
  27
  28
  29 void computeGradient(const NLMISC::CRGBA *valueTab, uint32 numValues, uint32 nbStages, CPSVector<NLMISC::CRGBA>::V &grad, NLMISC::CRGBA &minValue, NLMISC::CRGBA &maxValue)
  30 {
  31         nlassert(numValues > 1);
  32         nlassert(nbStages > 0);
  33
  34         maxValue = minValue = valueTab[0];
  35         uint nbValues = (numValues - 1) * nbStages;
  36         grad.resize(nbValues + 1);
  37
  38
  39         float step = 1.0f / float(nbStages);
  40         float alpha;
  41
  42         NLMISC::CRGBA *dest = &grad[0];
  43 #if defined(NL_OS_WINDOWS) && !defined(NL_NO_ASM)
  44         float stepX127 = 127.f * step;
  45         if (NLMISC::CSystemInfo::hasMMX())
  46         {
  47                 // do interpolation using mmx
  48                 for (uint32 k = 0; k  < (numValues - 1); ++k)
  49                 {
  50                         uint32 col0 = *(uint32 *) &valueTab[k];
  51                         uint64 stepPacked;
  52                         sint16 *step = (sint16 *) &stepPacked;
  53                         step[0] = (sint16) (stepX127 * (valueTab[k + 1].R - valueTab[k].R));
  54                         step[1] = (sint16) (stepX127 * (valueTab[k + 1].G - valueTab[k].G));
  55                         step[2] = (sint16) (stepX127 * (valueTab[k + 1].B - valueTab[k].B));
  56                         step[3] = (sint16) (stepX127 * (valueTab[k + 1].A - valueTab[k].A));
  57                         __asm
  58                         {
  59                                         pxor mm0, mm0            // mm0 = 0
  60                                         movd mm1, col0
  61                                         punpcklbw mm0, mm1 // store current col in mm0, each component is stored in 7:8 fixed integer
  62                                         psrlw     mm0, 1
  63                                         movq mm2, stepPacked       // store step for gradient
  64                                         mov  edi, dest
  65                                         mov  eax, 4
  66                                         mov  ecx, nbStages
  67                                         sub  edi, eax
  68                                 myLoop:
  69                                         // unpack current color
  70                                         movq        mm1, mm0
  71                                         add                     edi, eax // advance destination
  72                                         psrlw       mm1, 7
  73                                         packuswb    mm1, mm1
  74                                         movd            [edi], mm1 // store result
  75                                         paddsw     mm0, mm2 // increment color
  76                                         dec ecx
  77                                         jne myLoop
  78                                         emms
  79                         }
  80                         dest += nbStages;
  81                 }
  82         }
  83         else
  84 #endif
  85         {
  86                 // copy the tab performing linear interpolation between values given in parameter
  87                 for (uint32 k = 0; k  < (numValues - 1); ++k)
  88                 {
  89                         alpha = 0;
  90
  91                         for(uint32 l = 0; l < nbStages; ++l)
  92                         {
  93                                 // use the right version of the template function PSValueBlend
  94                                 // to do the job
  95                                 *dest++ = PSValueBlend(valueTab[k], valueTab[k + 1], alpha);
  96                                 alpha += step;
  97                         }
  98                 }
  99         }
 100         *dest = valueTab[numValues - 1];
 101 }
 102
 103 }