vbagx/source/vba/bilinear.cpp

// VisualBoyAdvance - Nintendo Gameboy/GameboyAdvance (TM) emulator.
// Copyright (C) 1999-2003 Forgotten
// Copyright (C) 2004 Forgotten and the VBA development team

// 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, 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.

/**     Code adapted from Exult source code by Forgotten
 **	Scale.cc - Trying to scale with bilinear interpolation.
 **
 **	Written: 6/14/00 - JSF
 **/

#include "System.h"

static u8 row_cur[3*322];
static u8 row_next[3*322];

static u8 *rgb_row_cur = row_cur;
static u8 *rgb_row_next = row_next;

#define RGB(r,g,b) ((r)>>3) << systemRedShift |\
  ((g) >> 3) << systemGreenShift |\
  ((b) >> 3) << systemBlueShift\

static void fill_rgb_row_16(u16 *from, int src_width, u8 *row, int width)
{
  u8 *copy_start = row + src_width*3;
  u8 *all_stop = row + width*3;
  while (row < copy_start)
    {
      u16 color = *from++;
      *row++ = ((color >> systemRedShift) & 0x1f) << 3;
      *row++ = ((color >> systemGreenShift) & 0x1f) << 3;
      *row++ = ((color >> systemBlueShift) & 0x1f) << 3;
    }
  // any remaining elements to be written to 'row' are a replica of the
  // preceding pixel
  u8 *p = row-3;
  while (row < all_stop)
    {
      // we're guaranteed three elements per pixel; could unroll the loop
      // further, especially with a Duff's Device, but the gains would be
      // probably limited (judging by profiler output)
      *row++ = *p++;
      *row++ = *p++;
      *row++ = *p++;
    }
}

static void fill_rgb_row_32(u32 *from, int src_width, u8 *row, int width)
{
  u8 *copy_start = row + src_width*3;
  u8 *all_stop = row + width*3;
  while (row < copy_start)
    {
      u32 color = *from++;
      *row++ = ((color >> systemRedShift) & 0x1f) << 3;
      *row++ = ((color >> systemGreenShift) & 0x1f) << 3;
      *row++ = ((color >> systemBlueShift) & 0x1f) << 3;
    }
  // any remaining elements to be written to 'row' are a replica of the
  // preceding pixel
  u8 *p = row-3;
  while (row < all_stop)
    {
      // we're guaranteed three elements per pixel; could unroll the loop
      // further, especially with a Duff's Device, but the gains would be
      // probably limited (judging by profiler output)
      *row++ = *p++;
      *row++ = *p++;
      *row++ = *p++;
    }
}

void Bilinear(u8 *srcPtr, u32 srcPitch, u8 * /* deltaPtr */,
              u8 *dstPtr, u32 dstPitch, int width, int height)
{
  u16 *to = (u16 *)dstPtr;
  u16 *to_odd = (u16 *)(dstPtr + dstPitch);

  int from_width = width;
  u16 *from = (u16 *)srcPtr;
  fill_rgb_row_16(from, from_width, rgb_row_cur, width+1);

  for(int y = 0; y < height; y++)
    {
      u16 *from_orig = from;
      u16 *to_orig = to;

      if (y+1 < height)
        fill_rgb_row_16(from+width+2, from_width, rgb_row_next,
                        width+1);
      else
        fill_rgb_row_16(from, from_width, rgb_row_next, width+1);

      // every pixel in the src region, is extended to 4 pixels in the
      // destination, arranged in a square 'quad'; if the current src
      // pixel is 'a', then in what follows 'b' is the src pixel to the
      // right, 'c' is the src pixel below, and 'd' is the src pixel to
      // the right and down
      u8 *cur_row  = rgb_row_cur;
      u8 *next_row = rgb_row_next;
      u8 *ar = cur_row++;
      u8 *ag = cur_row++;
      u8 *ab = cur_row++;
      u8 *cr = next_row++;
      u8 *cg = next_row++;
      u8 *cb = next_row++;
      for(int x=0; x < width; x++)
        {
          u8 *br = cur_row++;
          u8 *bg = cur_row++;
          u8 *bb = cur_row++;
          u8 *dr = next_row++;
          u8 *dg = next_row++;
          u8 *db = next_row++;

          // upper left pixel in quad: just copy it in
          *to++ = RGB(*ar, *ag, *ab);

          // upper right
          *to++ = RGB((*ar+*br)>>1, (*ag+*bg)>>1, (*ab+*bb)>>1);

          // lower left
          *to_odd++ = RGB((*ar+*cr)>>1, (*ag+*cg)>>1, (*ab+*cb)>>1);

          // lower right
          *to_odd++ = RGB((*ar+*br+*cr+*dr)>>2,
                          (*ag+*bg+*cg+*dg)>>2,
                          (*ab+*bb+*cb+*db)>>2);

          // 'b' becomes 'a', 'd' becomes 'c'
          ar = br;
          ag = bg;
          ab = bb;
          cr = dr;
          cg = dg;
          cb = db;
        }

      // the "next" rgb row becomes the current; the old current rgb row is
      // recycled and serves as the new "next" row
      u8 *temp;
      temp = rgb_row_cur;
      rgb_row_cur = rgb_row_next;
      rgb_row_next = temp;

      // update the pointers for start of next pair of lines
      from = (u16 *)((u8 *)from_orig + srcPitch);
      to = (u16 *)((u8 *)to_orig + (dstPitch << 1));
      to_odd = (u16 *)((u8 *)to + dstPitch);
    }
}

void BilinearPlus(u8 *srcPtr, u32 srcPitch, u8 * /* deltaPtr */,
                  u8 *dstPtr, u32 dstPitch, int width, int height)
{
  u16 *to = (u16 *)dstPtr;
  u16 *to_odd = (u16 *)(dstPtr + dstPitch);

  int from_width = width;
  u16 *from = (u16 *)srcPtr;
  fill_rgb_row_16(from, from_width, rgb_row_cur, width+1);

  for(int y = 0; y < height; y++)
    {
      u16 *from_orig = from;
      u16 *to_orig = to;

      if (y+1 < height)
        fill_rgb_row_16(from+width+2, from_width, rgb_row_next,
                        width+1);
      else
        fill_rgb_row_16(from, from_width, rgb_row_next, width+1);

      // every pixel in the src region, is extended to 4 pixels in the
      // destination, arranged in a square 'quad'; if the current src
      // pixel is 'a', then in what follows 'b' is the src pixel to the
      // right, 'c' is the src pixel below, and 'd' is the src pixel to
      // the right and down
      u8 *cur_row  = rgb_row_cur;
      u8 *next_row = rgb_row_next;
      u8 *ar = cur_row++;
      u8 *ag = cur_row++;
      u8 *ab = cur_row++;
      u8 *cr = next_row++;
      u8 *cg = next_row++;
      u8 *cb = next_row++;
      for(int x=0; x < width; x++)
        {
          u8 *br = cur_row++;
          u8 *bg = cur_row++;
          u8 *bb = cur_row++;
          u8 *dr = next_row++;
          u8 *dg = next_row++;
          u8 *db = next_row++;

          // upper left pixel in quad: just copy it in
          //*to++ = manip.rgb(*ar, *ag, *ab);
#ifdef USE_ORIGINAL_BILINEAR_PLUS
          *to++ = RGB(
                    (((*ar)<<2) +((*ar)) + (*cr+*br+*br) )>> 3,
                    (((*ag)<<2) +((*ag)) + (*cg+*bg+*bg) )>> 3,
                    (((*ab)<<2) +((*ab)) + (*cb+*bb+*bb) )>> 3);
#else
          *to++ = RGB(
                    (((*ar)<<3) +((*ar)<<1) + (*cr+*br+*br+*cr) )>> 4,
                    (((*ag)<<3) +((*ag)<<1) + (*cg+*bg+*bg+*cg) )>> 4,
                    (((*ab)<<3) +((*ab)<<1) + (*cb+*bb+*bb+*cb) )>> 4);
#endif

          // upper right
          *to++ = RGB((*ar+*br)>>1, (*ag+*bg)>>1, (*ab+*bb)>>1);

          // lower left
          *to_odd++ = RGB((*ar+*cr)>>1, (*ag+*cg)>>1, (*ab+*cb)>>1);

          // lower right
          *to_odd++ = RGB((*ar+*br+*cr+*dr)>>2,
                          (*ag+*bg+*cg+*dg)>>2,
                          (*ab+*bb+*cb+*db)>>2);

          // 'b' becomes 'a', 'd' becomes 'c'
          ar = br;
          ag = bg;
          ab = bb;
          cr = dr;
          cg = dg;
          cb = db;
        }

      // the "next" rgb row becomes the current; the old current rgb row is
      // recycled and serves as the new "next" row
      u8 *temp;
      temp = rgb_row_cur;
      rgb_row_cur = rgb_row_next;
      rgb_row_next = temp;

      // update the pointers for start of next pair of lines
      from = (u16 *)((u8 *)from_orig + srcPitch);
      to = (u16 *)((u8 *)to_orig + (dstPitch << 1));
      to_odd = (u16 *)((u8 *)to + dstPitch);
    }
}

void Bilinear32(u8 *srcPtr, u32 srcPitch, u8 * /* deltaPtr */,
                u8 *dstPtr, u32 dstPitch, int width, int height)
{
  u32 *to = (u32 *)dstPtr;
  u32 *to_odd = (u32 *)(dstPtr + dstPitch);

  int from_width = width;
  if(width+1 < from_width)
    from_width = width+1;
  u32 *from = (u32 *)srcPtr;
  fill_rgb_row_32(from, from_width, rgb_row_cur, width+1);

  for(int y = 0; y < height; y++)
    {
      u32 *from_orig = from;
      u32 *to_orig = to;

      if (y+1 < height)
        fill_rgb_row_32(from+width+1, from_width, rgb_row_next,
                        width+1);
      else
        fill_rgb_row_32(from, from_width, rgb_row_next, width+1);

      // every pixel in the src region, is extended to 4 pixels in the
      // destination, arranged in a square 'quad'; if the current src
      // pixel is 'a', then in what follows 'b' is the src pixel to the
      // right, 'c' is the src pixel below, and 'd' is the src pixel to
      // the right and down
      u8 *cur_row  = rgb_row_cur;
      u8 *next_row = rgb_row_next;
      u8 *ar = cur_row++;
      u8 *ag = cur_row++;
      u8 *ab = cur_row++;
      u8 *cr = next_row++;
      u8 *cg = next_row++;
      u8 *cb = next_row++;
      for(int x=0; x < width; x++)
        {
          u8 *br = cur_row++;
          u8 *bg = cur_row++;
          u8 *bb = cur_row++;
          u8 *dr = next_row++;
          u8 *dg = next_row++;
          u8 *db = next_row++;

          // upper left pixel in quad: just copy it in
          *to++ = RGB(*ar, *ag, *ab);

          // upper right
          *to++ = RGB((*ar+*br)>>1, (*ag+*bg)>>1, (*ab+*bb)>>1);

          // lower left
          *to_odd++ = RGB((*ar+*cr)>>1, (*ag+*cg)>>1, (*ab+*cb)>>1);

          // lower right
          *to_odd++ = RGB((*ar+*br+*cr+*dr)>>2,
                          (*ag+*bg+*cg+*dg)>>2,
                          (*ab+*bb+*cb+*db)>>2);

          // 'b' becomes 'a', 'd' becomes 'c'
          ar = br;
          ag = bg;
          ab = bb;
          cr = dr;
          cg = dg;
          cb = db;
        }

      // the "next" rgb row becomes the current; the old current rgb row is
      // recycled and serves as the new "next" row
      u8 *temp;
      temp = rgb_row_cur;
      rgb_row_cur = rgb_row_next;
      rgb_row_next = temp;

      // update the pointers for start of next pair of lines
      from = (u32 *)((u8 *)from_orig + srcPitch);
      to = (u32 *)((u8 *)to_orig + (dstPitch << 1));
      to_odd = (u32 *)((u8 *)to + dstPitch);
    }
}

void BilinearPlus32(u8 *srcPtr, u32 srcPitch, u8 * /* deltaPtr */,
                    u8 *dstPtr, u32 dstPitch, int width, int height)
{
  u32 *to = (u32 *)dstPtr;
  u32 *to_odd = (u32 *)(dstPtr + dstPitch);

  int from_width = width;
  if(width+1 < from_width)
    from_width = width+1;
  u32 *from = (u32 *)srcPtr;
  fill_rgb_row_32(from, from_width, rgb_row_cur, width+1);

  for(int y = 0; y < height; y++)
    {
      u32 *from_orig = from;
      u32 *to_orig = to;

      if (y+1 < height)
        fill_rgb_row_32(from+width+1, from_width, rgb_row_next,
                        width+1);
      else
        fill_rgb_row_32(from, from_width, rgb_row_next, width+1);

      // every pixel in the src region, is extended to 4 pixels in the
      // destination, arranged in a square 'quad'; if the current src
      // pixel is 'a', then in what follows 'b' is the src pixel to the
      // right, 'c' is the src pixel below, and 'd' is the src pixel to
      // the right and down
      u8 *cur_row  = rgb_row_cur;
      u8 *next_row = rgb_row_next;
      u8 *ar = cur_row++;
      u8 *ag = cur_row++;
      u8 *ab = cur_row++;
      u8 *cr = next_row++;
      u8 *cg = next_row++;
      u8 *cb = next_row++;
      for(int x=0; x < width; x++)
        {
          u8 *br = cur_row++;
          u8 *bg = cur_row++;
          u8 *bb = cur_row++;
          u8 *dr = next_row++;
          u8 *dg = next_row++;
          u8 *db = next_row++;

          // upper left pixel in quad: just copy it in
          //*to++ = manip.rgb(*ar, *ag, *ab);
#ifdef USE_ORIGINAL_BILINEAR_PLUS
          *to++ = RGB(
                    (((*ar)<<2) +((*ar)) + (*cr+*br+*br) )>> 3,
                    (((*ag)<<2) +((*ag)) + (*cg+*bg+*bg) )>> 3,
                    (((*ab)<<2) +((*ab)) + (*cb+*bb+*bb) )>> 3);
#else
          *to++ = RGB(
                    (((*ar)<<3) +((*ar)<<1) + (*cr+*br+*br+*cr) )>> 4,
                    (((*ag)<<3) +((*ag)<<1) + (*cg+*bg+*bg+*cg) )>> 4,
                    (((*ab)<<3) +((*ab)<<1) + (*cb+*bb+*bb+*cb) )>> 4);
#endif

          // upper right
          *to++ = RGB((*ar+*br)>>1, (*ag+*bg)>>1, (*ab+*bb)>>1);

          // lower left
          *to_odd++ = RGB((*ar+*cr)>>1, (*ag+*cg)>>1, (*ab+*cb)>>1);

          // lower right
          *to_odd++ = RGB((*ar+*br+*cr+*dr)>>2,
                          (*ag+*bg+*cg+*dg)>>2,
                          (*ab+*bb+*cb+*db)>>2);

          // 'b' becomes 'a', 'd' becomes 'c'
          ar = br;
          ag = bg;
          ab = bb;
          cr = dr;
          cg = dg;
          cb = db;
        }

      // the "next" rgb row becomes the current; the old current rgb row is
      // recycled and serves as the new "next" row
      u8 *temp;
      temp = rgb_row_cur;
      rgb_row_cur = rgb_row_next;
      rgb_row_next = temp;

      // update the pointers for start of next pair of lines
      from = (u32 *)((u8 *)from_orig + srcPitch);
      to = (u32 *)((u8 *)to_orig + (dstPitch << 1));
      to_odd = (u32 *)((u8 *)to + dstPitch);
    }
}