[MESA]

[reactos.git] / reactos / dll / opengl / mesa / src / mesa / swrast / s_texfilter.c

/*
 * Mesa 3-D graphics library
 * Version:  7.3
 *
 * Copyright (C) 1999-2008  Brian Paul   All Rights Reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included
 * in all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * BRIAN PAUL BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
 * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 */


#include "main/glheader.h"
#include "main/context.h"
#include "main/colormac.h"
#include "main/imports.h"

#include "s_context.h"
#include "s_texfilter.h"


/*
 * Note, the FRAC macro has to work perfectly.  Otherwise you'll sometimes
 * see 1-pixel bands of improperly weighted linear-filtered textures.
 * The tests/texwrap.c demo is a good test.
 * Also note, FRAC(x) doesn't truly return the fractional part of x for x < 0.
 * Instead, if x < 0 then FRAC(x) = 1 - true_frac(x).
 */
#define FRAC(f)  ((f) - IFLOOR(f))



/**
 * Linear interpolation macro
 */
#define LERP(T, A, B)  ( (A) + (T) * ((B) - (A)) )


/**
 * Do 2D/biliner interpolation of float values.
 * v00, v10, v01 and v11 are typically four texture samples in a square/box.
 * a and b are the horizontal and vertical interpolants.
 * It's important that this function is inlined when compiled with
 * optimization!  If we find that's not true on some systems, convert
 * to a macro.
 */
static inline GLfloat
lerp_2d(GLfloat a, GLfloat b,
        GLfloat v00, GLfloat v10, GLfloat v01, GLfloat v11)
{
   const GLfloat temp0 = LERP(a, v00, v10);
   const GLfloat temp1 = LERP(a, v01, v11);
   return LERP(b, temp0, temp1);
}


/**
 * Do 3D/trilinear interpolation of float values.
 * \sa lerp_2d
 */
static inline GLfloat
lerp_3d(GLfloat a, GLfloat b, GLfloat c,
        GLfloat v000, GLfloat v100, GLfloat v010, GLfloat v110,
        GLfloat v001, GLfloat v101, GLfloat v011, GLfloat v111)
{
   const GLfloat temp00 = LERP(a, v000, v100);
   const GLfloat temp10 = LERP(a, v010, v110);
   const GLfloat temp01 = LERP(a, v001, v101);
   const GLfloat temp11 = LERP(a, v011, v111);
   const GLfloat temp0 = LERP(b, temp00, temp10);
   const GLfloat temp1 = LERP(b, temp01, temp11);
   return LERP(c, temp0, temp1);
}


/**
 * Do linear interpolation of colors.
 */
static inline void
lerp_rgba(GLfloat result[4], GLfloat t, const GLfloat a[4], const GLfloat b[4])
{
   result[0] = LERP(t, a[0], b[0]);
   result[1] = LERP(t, a[1], b[1]);
   result[2] = LERP(t, a[2], b[2]);
   result[3] = LERP(t, a[3], b[3]);
}


/**
 * Do bilinear interpolation of colors.
 */
static inline void
lerp_rgba_2d(GLfloat result[4], GLfloat a, GLfloat b,
             const GLfloat t00[4], const GLfloat t10[4],
             const GLfloat t01[4], const GLfloat t11[4])
{
   result[0] = lerp_2d(a, b, t00[0], t10[0], t01[0], t11[0]);
   result[1] = lerp_2d(a, b, t00[1], t10[1], t01[1], t11[1]);
   result[2] = lerp_2d(a, b, t00[2], t10[2], t01[2], t11[2]);
   result[3] = lerp_2d(a, b, t00[3], t10[3], t01[3], t11[3]);
}


/**
 * Do trilinear interpolation of colors.
 */
static inline void
lerp_rgba_3d(GLfloat result[4], GLfloat a, GLfloat b, GLfloat c,
             const GLfloat t000[4], const GLfloat t100[4],
             const GLfloat t010[4], const GLfloat t110[4],
             const GLfloat t001[4], const GLfloat t101[4],
             const GLfloat t011[4], const GLfloat t111[4])
{
   GLuint k;
   /* compiler should unroll these short loops */
   for (k = 0; k < 4; k++) {
      result[k] = lerp_3d(a, b, c, t000[k], t100[k], t010[k], t110[k],
                                   t001[k], t101[k], t011[k], t111[k]);
   }
}


/**
 * Used for GL_REPEAT wrap mode.  Using A % B doesn't produce the
 * right results for A<0.  Casting to A to be unsigned only works if B
 * is a power of two.  Adding a bias to A (which is a multiple of B)
 * avoids the problems with A < 0 (for reasonable A) without using a
 * conditional.
 */
#define REMAINDER(A, B) (((A) + (B) * 1024) % (B))


/**
 * Used to compute texel locations for linear sampling.
 * Input:
 *    wrapMode = GL_REPEAT, GL_CLAMP, GL_CLAMP_TO_EDGE, GL_CLAMP_TO_BORDER
 *    s = texcoord in [0,1]
 *    size = width (or height or depth) of texture
 * Output:
 *    i0, i1 = returns two nearest texel indexes
 *    weight = returns blend factor between texels
 */
static inline void
linear_texel_locations(GLenum wrapMode,
                       const struct gl_texture_image *img,
                       GLint size, GLfloat s,
                       GLint *i0, GLint *i1, GLfloat *weight)
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   GLfloat u;
   switch (wrapMode) {
   case GL_REPEAT:
      u = s * size - 0.5F;
      if (swImg->_IsPowerOfTwo) {
         *i0 = IFLOOR(u) & (size - 1);
         *i1 = (*i0 + 1) & (size - 1);
      }
      else {
         *i0 = REMAINDER(IFLOOR(u), size);
         *i1 = REMAINDER(*i0 + 1, size);
      }
      break;
   case GL_CLAMP_TO_EDGE:
      if (s <= 0.0F)
         u = 0.0F;
      else if (s >= 1.0F)
         u = (GLfloat) size;
      else
         u = s * size;
      u -= 0.5F;
      *i0 = IFLOOR(u);
      *i1 = *i0 + 1;
      if (*i0 < 0)
         *i0 = 0;
      if (*i1 >= (GLint) size)
         *i1 = size - 1;
      break;
   case GL_CLAMP_TO_BORDER:
      {
         const GLfloat min = -1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         if (s <= min)
            u = min * size;
         else if (s >= max)
            u = max * size;
         else
            u = s * size;
         u -= 0.5F;
         *i0 = IFLOOR(u);
         *i1 = *i0 + 1;
      }
      break;
   case GL_MIRRORED_REPEAT:
      {
         const GLint flr = IFLOOR(s);
         if (flr & 1)
            u = 1.0F - (s - (GLfloat) flr);
         else
            u = s - (GLfloat) flr;
         u = (u * size) - 0.5F;
         *i0 = IFLOOR(u);
         *i1 = *i0 + 1;
         if (*i0 < 0)
            *i0 = 0;
         if (*i1 >= (GLint) size)
            *i1 = size - 1;
      }
      break;
   case GL_MIRROR_CLAMP_EXT:
      u = FABSF(s);
      if (u >= 1.0F)
         u = (GLfloat) size;
      else
         u *= size;
      u -= 0.5F;
      *i0 = IFLOOR(u);
      *i1 = *i0 + 1;
      break;
   case GL_MIRROR_CLAMP_TO_EDGE_EXT:
      u = FABSF(s);
      if (u >= 1.0F)
         u = (GLfloat) size;
      else
         u *= size;
      u -= 0.5F;
      *i0 = IFLOOR(u);
      *i1 = *i0 + 1;
      if (*i0 < 0)
         *i0 = 0;
      if (*i1 >= (GLint) size)
         *i1 = size - 1;
      break;
   case GL_MIRROR_CLAMP_TO_BORDER_EXT:
      {
         const GLfloat min = -1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         u = FABSF(s);
         if (u <= min)
            u = min * size;
         else if (u >= max)
            u = max * size;
         else
            u *= size;
         u -= 0.5F;
         *i0 = IFLOOR(u);
         *i1 = *i0 + 1;
      }
      break;
   case GL_CLAMP:
      if (s <= 0.0F)
         u = 0.0F;
      else if (s >= 1.0F)
         u = (GLfloat) size;
      else
         u = s * size;
      u -= 0.5F;
      *i0 = IFLOOR(u);
      *i1 = *i0 + 1;
      break;
   default:
      _mesa_problem(NULL, "Bad wrap mode");
      u = 0.0F;
      break;
   }
   *weight = FRAC(u);
}


/**
 * Used to compute texel location for nearest sampling.
 */
static inline GLint
nearest_texel_location(GLenum wrapMode,
                       const struct gl_texture_image *img,
                       GLint size, GLfloat s)
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   GLint i;

   switch (wrapMode) {
   case GL_REPEAT:
      /* s limited to [0,1) */
      /* i limited to [0,size-1] */
      i = IFLOOR(s * size);
      if (swImg->_IsPowerOfTwo)
         i &= (size - 1);
      else
         i = REMAINDER(i, size);
      return i;
   case GL_CLAMP_TO_EDGE:
      {
         /* s limited to [min,max] */
         /* i limited to [0, size-1] */
         const GLfloat min = 1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         if (s < min)
            i = 0;
         else if (s > max)
            i = size - 1;
         else
            i = IFLOOR(s * size);
      }
      return i;
   case GL_CLAMP_TO_BORDER:
      {
         /* s limited to [min,max] */
         /* i limited to [-1, size] */
         const GLfloat min = -1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         if (s <= min)
            i = -1;
         else if (s >= max)
            i = size;
         else
            i = IFLOOR(s * size);
      }
      return i;
   case GL_MIRRORED_REPEAT:
      {
         const GLfloat min = 1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         const GLint flr = IFLOOR(s);
         GLfloat u;
         if (flr & 1)
            u = 1.0F - (s - (GLfloat) flr);
         else
            u = s - (GLfloat) flr;
         if (u < min)
            i = 0;
         else if (u > max)
            i = size - 1;
         else
            i = IFLOOR(u * size);
      }
      return i;
   case GL_MIRROR_CLAMP_EXT:
      {
         /* s limited to [0,1] */
         /* i limited to [0,size-1] */
         const GLfloat u = FABSF(s);
         if (u <= 0.0F)
            i = 0;
         else if (u >= 1.0F)
            i = size - 1;
         else
            i = IFLOOR(u * size);
      }
      return i;
   case GL_MIRROR_CLAMP_TO_EDGE_EXT:
      {
         /* s limited to [min,max] */
         /* i limited to [0, size-1] */
         const GLfloat min = 1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         const GLfloat u = FABSF(s);
         if (u < min)
            i = 0;
         else if (u > max)
            i = size - 1;
         else
            i = IFLOOR(u * size);
      }
      return i;
   case GL_MIRROR_CLAMP_TO_BORDER_EXT:
      {
         /* s limited to [min,max] */
         /* i limited to [0, size-1] */
         const GLfloat min = -1.0F / (2.0F * size);
         const GLfloat max = 1.0F - min;
         const GLfloat u = FABSF(s);
         if (u < min)
            i = -1;
         else if (u > max)
            i = size;
         else
            i = IFLOOR(u * size);
      }
      return i;
   case GL_CLAMP:
      /* s limited to [0,1] */
      /* i limited to [0,size-1] */
      if (s <= 0.0F)
         i = 0;
      else if (s >= 1.0F)
         i = size - 1;
      else
         i = IFLOOR(s * size);
      return i;
   default:
      _mesa_problem(NULL, "Bad wrap mode");
      return 0;
   }
}


/* Power of two image sizes only */
static inline void
linear_repeat_texel_location(GLuint size, GLfloat s,
                             GLint *i0, GLint *i1, GLfloat *weight)
{
   GLfloat u = s * size - 0.5F;
   *i0 = IFLOOR(u) & (size - 1);
   *i1 = (*i0 + 1) & (size - 1);
   *weight = FRAC(u);
}


/**
 * Do clamp/wrap for a texture rectangle coord, GL_NEAREST filter mode.
 */
static inline GLint
clamp_rect_coord_nearest(GLenum wrapMode, GLfloat coord, GLint max)
{
   switch (wrapMode) {
   case GL_CLAMP:
      return IFLOOR( CLAMP(coord, 0.0F, max - 1) );
   case GL_CLAMP_TO_EDGE:
      return IFLOOR( CLAMP(coord, 0.5F, max - 0.5F) );
   case GL_CLAMP_TO_BORDER:
      return IFLOOR( CLAMP(coord, -0.5F, max + 0.5F) );
   default:
      _mesa_problem(NULL, "bad wrapMode in clamp_rect_coord_nearest");
      return 0;
   }
}


/**
 * As above, but GL_LINEAR filtering.
 */
static inline void
clamp_rect_coord_linear(GLenum wrapMode, GLfloat coord, GLint max,
                        GLint *i0out, GLint *i1out, GLfloat *weight)
{
   GLfloat fcol;
   GLint i0, i1;
   switch (wrapMode) {
   case GL_CLAMP:
      /* Not exactly what the spec says, but it matches NVIDIA output */
      fcol = CLAMP(coord - 0.5F, 0.0F, max - 1);
      i0 = IFLOOR(fcol);
      i1 = i0 + 1;
      break;
   case GL_CLAMP_TO_EDGE:
      fcol = CLAMP(coord, 0.5F, max - 0.5F);
      fcol -= 0.5F;
      i0 = IFLOOR(fcol);
      i1 = i0 + 1;
      if (i1 > max - 1)
         i1 = max - 1;
      break;
   case GL_CLAMP_TO_BORDER:
      fcol = CLAMP(coord, -0.5F, max + 0.5F);
      fcol -= 0.5F;
      i0 = IFLOOR(fcol);
      i1 = i0 + 1;
      break;
   default:
      _mesa_problem(NULL, "bad wrapMode in clamp_rect_coord_linear");
      i0 = i1 = 0;
      fcol = 0.0F;
      break;
   }
   *i0out = i0;
   *i1out = i1;
   *weight = FRAC(fcol);
}


/**
 * Compute slice/image to use for 1D or 2D array texture.
 */
static inline GLint
tex_array_slice(GLfloat coord, GLsizei size)
{
   GLint slice = IFLOOR(coord + 0.5f);
   slice = CLAMP(slice, 0, size - 1);
   return slice;
}


/**
 * Compute nearest integer texcoords for given texobj and coordinate.
 * NOTE: only used for depth texture sampling.
 */
static inline void
nearest_texcoord(const struct gl_texture_object *texObj,
                 GLuint level,
                 const GLfloat texcoord[4],
                 GLint *i, GLint *j, GLint *k)
{
   const struct gl_texture_image *img = texObj->Image[0][level];
   const GLint width = img->Width;
   const GLint height = img->Height;

   switch (texObj->Target) {
   case GL_TEXTURE_1D:
      *i = nearest_texel_location(texObj->Sampler.WrapS, img, width, texcoord[0]);
      *j = 0;
      *k = 0;
      break;
   case GL_TEXTURE_2D:
      *i = nearest_texel_location(texObj->Sampler.WrapS, img, width, texcoord[0]);
      *j = nearest_texel_location(texObj->Sampler.WrapT, img, height, texcoord[1]);
      *k = 0;
      break;
   default:
      *i = *j = *k = 0;
      break;
   }
}


/**
 * Compute linear integer texcoords for given texobj and coordinate.
 * NOTE: only used for depth texture sampling.
 */
static inline void
linear_texcoord(const struct gl_texture_object *texObj,
                GLuint level,
                const GLfloat texcoord[4],
                GLint *i0, GLint *i1, GLint *j0, GLint *j1, GLint *slice,
                GLfloat *wi, GLfloat *wj)
{
   const struct gl_texture_image *img = texObj->Image[0][level];
   const GLint width = img->Width;
   const GLint height = img->Height;

   switch (texObj->Target) {
   case GL_TEXTURE_1D:
   case GL_TEXTURE_2D:
      linear_texel_locations(texObj->Sampler.WrapS, img, width,
                             texcoord[0], i0, i1, wi);
      linear_texel_locations(texObj->Sampler.WrapT, img, height,
                             texcoord[1], j0, j1, wj);
      *slice = 0;
      break;

   default:
      *slice = 0;
      break;
   }
}



/**
 * For linear interpolation between mipmap levels N and N+1, this function
 * computes N.
 */
static inline GLint
linear_mipmap_level(const struct gl_texture_object *tObj, GLfloat lambda)
{
   if (lambda < 0.0F)
      return tObj->BaseLevel;
   else if (lambda > tObj->_MaxLambda)
      return (GLint) (tObj->BaseLevel + tObj->_MaxLambda);
   else
      return (GLint) (tObj->BaseLevel + lambda);
}


/**
 * Compute the nearest mipmap level to take texels from.
 */
static inline GLint
nearest_mipmap_level(const struct gl_texture_object *tObj, GLfloat lambda)
{
   GLfloat l;
   GLint level;
   if (lambda <= 0.5F)
      l = 0.0F;
   else if (lambda > tObj->_MaxLambda + 0.4999F)
      l = tObj->_MaxLambda + 0.4999F;
   else
      l = lambda;
   level = (GLint) (tObj->BaseLevel + l + 0.5F);
   if (level > tObj->_MaxLevel)
      level = tObj->_MaxLevel;
   return level;
}



/*
 * Bitflags for texture border color sampling.
 */
#define I0BIT   1
#define I1BIT   2
#define J0BIT   4
#define J1BIT   8
#define K0BIT  16
#define K1BIT  32



/**
 * The lambda[] array values are always monotonic.  Either the whole span
 * will be minified, magnified, or split between the two.  This function
 * determines the subranges in [0, n-1] that are to be minified or magnified.
 */
static inline void
compute_min_mag_ranges(const struct gl_texture_object *tObj,
                       GLuint n, const GLfloat lambda[],
                       GLuint *minStart, GLuint *minEnd,
                       GLuint *magStart, GLuint *magEnd)
{
   GLfloat minMagThresh;

   /* we shouldn't be here if minfilter == magfilter */
   ASSERT(tObj->Sampler.MinFilter != tObj->Sampler.MagFilter);

   /* This bit comes from the OpenGL spec: */
   if (tObj->Sampler.MagFilter == GL_LINEAR
       && (tObj->Sampler.MinFilter == GL_NEAREST_MIPMAP_NEAREST ||
           tObj->Sampler.MinFilter == GL_NEAREST_MIPMAP_LINEAR)) {
      minMagThresh = 0.5F;
   }
   else {
      minMagThresh = 0.0F;
   }

#if 0
   /* DEBUG CODE: Verify that lambda[] is monotonic.
    * We can't really use this because the inaccuracy in the LOG2 function
    * causes this test to fail, yet the resulting texturing is correct.
    */
   if (n > 1) {
      GLuint i;
      printf("lambda delta = %g\n", lambda[0] - lambda[n-1]);
      if (lambda[0] >= lambda[n-1]) { /* decreasing */
         for (i = 0; i < n - 1; i++) {
            ASSERT((GLint) (lambda[i] * 10) >= (GLint) (lambda[i+1] * 10));
         }
      }
      else { /* increasing */
         for (i = 0; i < n - 1; i++) {
            ASSERT((GLint) (lambda[i] * 10) <= (GLint) (lambda[i+1] * 10));
         }
      }
   }
#endif /* DEBUG */

   if (lambda[0] <= minMagThresh && (n <= 1 || lambda[n-1] <= minMagThresh)) {
      /* magnification for whole span */
      *magStart = 0;
      *magEnd = n;
      *minStart = *minEnd = 0;
   }
   else if (lambda[0] > minMagThresh && (n <=1 || lambda[n-1] > minMagThresh)) {
      /* minification for whole span */
      *minStart = 0;
      *minEnd = n;
      *magStart = *magEnd = 0;
   }
   else {
      /* a mix of minification and magnification */
      GLuint i;
      if (lambda[0] > minMagThresh) {
         /* start with minification */
         for (i = 1; i < n; i++) {
            if (lambda[i] <= minMagThresh)
               break;
         }
         *minStart = 0;
         *minEnd = i;
         *magStart = i;
         *magEnd = n;
      }
      else {
         /* start with magnification */
         for (i = 1; i < n; i++) {
            if (lambda[i] > minMagThresh)
               break;
         }
         *magStart = 0;
         *magEnd = i;
         *minStart = i;
         *minEnd = n;
      }
   }

#if 0
   /* Verify the min/mag Start/End values
    * We don't use this either (see above)
    */
   {
      GLint i;
      for (i = 0; i < n; i++) {
         if (lambda[i] > minMagThresh) {
            /* minification */
            ASSERT(i >= *minStart);
            ASSERT(i < *minEnd);
         }
         else {
            /* magnification */
            ASSERT(i >= *magStart);
            ASSERT(i < *magEnd);
         }
      }
   }
#endif
}


/**
 * When we sample the border color, it must be interpreted according to
 * the base texture format.  Ex: if the texture base format it GL_ALPHA,
 * we return (0,0,0,BorderAlpha).
 */
static inline void
get_border_color(const struct gl_texture_object *tObj,
                 const struct gl_texture_image *img,
                 GLfloat rgba[4])
{
   switch (img->_BaseFormat) {
   case GL_RGB:
      rgba[0] = tObj->Sampler.BorderColor.f[0];
      rgba[1] = tObj->Sampler.BorderColor.f[1];
      rgba[2] = tObj->Sampler.BorderColor.f[2];
      rgba[3] = 1.0F;
      break;
   case GL_ALPHA:
      rgba[0] = rgba[1] = rgba[2] = 0.0;
      rgba[3] = tObj->Sampler.BorderColor.f[3];
      break;
   case GL_LUMINANCE:
      rgba[0] = rgba[1] = rgba[2] = tObj->Sampler.BorderColor.f[0];
      rgba[3] = 1.0;
      break;
   case GL_LUMINANCE_ALPHA:
      rgba[0] = rgba[1] = rgba[2] = tObj->Sampler.BorderColor.f[0];
      rgba[3] = tObj->Sampler.BorderColor.f[3];
      break;
   case GL_INTENSITY:
      rgba[0] = rgba[1] = rgba[2] = rgba[3] = tObj->Sampler.BorderColor.f[0];
      break;
   default:
      COPY_4V(rgba, tObj->Sampler.BorderColor.f);
      break;
   }
}


/**********************************************************************/
/*                    1-D Texture Sampling Functions                  */
/**********************************************************************/

/**
 * Return the texture sample for coordinate (s) using GL_NEAREST filter.
 */
static inline void
sample_1d_nearest(struct gl_context *ctx,
                  const struct gl_texture_object *tObj,
                  const struct gl_texture_image *img,
                  const GLfloat texcoord[4], GLfloat rgba[4])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;  /* without border, power of two */
   GLint i;
   i = nearest_texel_location(tObj->Sampler.WrapS, img, width, texcoord[0]);
   /* skip over the border, if any */
   i += img->Border;
   if (i < 0 || i >= (GLint) img->Width) {
      /* Need this test for GL_CLAMP_TO_BORDER mode */
      get_border_color(tObj, img, rgba);
   }
   else {
      swImg->FetchTexel(swImg, i, 0, 0, rgba);
   }
}


/**
 * Return the texture sample for coordinate (s) using GL_LINEAR filter.
 */
static inline void
sample_1d_linear(struct gl_context *ctx,
                 const struct gl_texture_object *tObj,
                 const struct gl_texture_image *img,
                 const GLfloat texcoord[4], GLfloat rgba[4])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;
   GLint i0, i1;
   GLbitfield useBorderColor = 0x0;
   GLfloat a;
   GLfloat t0[4], t1[4];  /* texels */

   linear_texel_locations(tObj->Sampler.WrapS, img, width, texcoord[0], &i0, &i1, &a);

   if (img->Border) {
      i0 += img->Border;
      i1 += img->Border;
   }
   else {
      if (i0 < 0 || i0 >= width)   useBorderColor |= I0BIT;
      if (i1 < 0 || i1 >= width)   useBorderColor |= I1BIT;
   }

   /* fetch texel colors */
   if (useBorderColor & I0BIT) {
      get_border_color(tObj, img, t0);
   }
   else {
      swImg->FetchTexel(swImg, i0, 0, 0, t0);
   }
   if (useBorderColor & I1BIT) {
      get_border_color(tObj, img, t1);
   }
   else {
      swImg->FetchTexel(swImg, i1, 0, 0, t1);
   }

   lerp_rgba(rgba, a, t0, t1);
}


static void
sample_1d_nearest_mipmap_nearest(struct gl_context *ctx,
                                 const struct gl_texture_object *tObj,
                                 GLuint n, const GLfloat texcoord[][4],
                                 const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_1d_nearest(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_1d_linear_mipmap_nearest(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_1d_linear(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_1d_nearest_mipmap_linear(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_1d_nearest(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                           texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];
         const GLfloat f = FRAC(lambda[i]);
         sample_1d_nearest(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_1d_nearest(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


static void
sample_1d_linear_mipmap_linear(struct gl_context *ctx,
                               const struct gl_texture_object *tObj,
                               GLuint n, const GLfloat texcoord[][4],
                               const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_1d_linear(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                          texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];
         const GLfloat f = FRAC(lambda[i]);
         sample_1d_linear(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_1d_linear(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


/** Sample 1D texture, nearest filtering for both min/magnification */
static void
sample_nearest_1d( struct gl_context *ctx,
                   const struct gl_texture_object *tObj, GLuint n,
                   const GLfloat texcoords[][4], const GLfloat lambda[],
                   GLfloat rgba[][4] )
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   (void) lambda;
   for (i = 0; i < n; i++) {
      sample_1d_nearest(ctx, tObj, image, texcoords[i], rgba[i]);
   }
}


/** Sample 1D texture, linear filtering for both min/magnification */
static void
sample_linear_1d( struct gl_context *ctx,
                  const struct gl_texture_object *tObj, GLuint n,
                  const GLfloat texcoords[][4], const GLfloat lambda[],
                  GLfloat rgba[][4] )
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   (void) lambda;
   for (i = 0; i < n; i++) {
      sample_1d_linear(ctx, tObj, image, texcoords[i], rgba[i]);
   }
}


/** Sample 1D texture, using lambda to choose between min/magnification */
static void
sample_lambda_1d( struct gl_context *ctx,
                  const struct gl_texture_object *tObj, GLuint n,
                  const GLfloat texcoords[][4],
                  const GLfloat lambda[], GLfloat rgba[][4] )
{
   GLuint minStart, minEnd;  /* texels with minification */
   GLuint magStart, magEnd;  /* texels with magnification */
   GLuint i;

   ASSERT(lambda != NULL);
   compute_min_mag_ranges(tObj, n, lambda,
                          &minStart, &minEnd, &magStart, &magEnd);

   if (minStart < minEnd) {
      /* do the minified texels */
      const GLuint m = minEnd - minStart;
      switch (tObj->Sampler.MinFilter) {
      case GL_NEAREST:
         for (i = minStart; i < minEnd; i++)
            sample_1d_nearest(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                              texcoords[i], rgba[i]);
         break;
      case GL_LINEAR:
         for (i = minStart; i < minEnd; i++)
            sample_1d_linear(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                             texcoords[i], rgba[i]);
         break;
      case GL_NEAREST_MIPMAP_NEAREST:
         sample_1d_nearest_mipmap_nearest(ctx, tObj, m, texcoords + minStart,
                                          lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_NEAREST:
         sample_1d_linear_mipmap_nearest(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_LINEAR:
         sample_1d_nearest_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_LINEAR:
         sample_1d_linear_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                        lambda + minStart, rgba + minStart);
         break;
      default:
         _mesa_problem(ctx, "Bad min filter in sample_1d_texture");
         return;
      }
   }

   if (magStart < magEnd) {
      /* do the magnified texels */
      switch (tObj->Sampler.MagFilter) {
      case GL_NEAREST:
         for (i = magStart; i < magEnd; i++)
            sample_1d_nearest(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                              texcoords[i], rgba[i]);
         break;
      case GL_LINEAR:
         for (i = magStart; i < magEnd; i++)
            sample_1d_linear(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                             texcoords[i], rgba[i]);
         break;
      default:
         _mesa_problem(ctx, "Bad mag filter in sample_1d_texture");
         return;
      }
   }
}


/**********************************************************************/
/*                    2-D Texture Sampling Functions                  */
/**********************************************************************/


/**
 * Return the texture sample for coordinate (s,t) using GL_NEAREST filter.
 */
static inline void
sample_2d_nearest(struct gl_context *ctx,
                  const struct gl_texture_object *tObj,
                  const struct gl_texture_image *img,
                  const GLfloat texcoord[4],
                  GLfloat rgba[])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;    /* without border, power of two */
   const GLint height = img->Height2;  /* without border, power of two */
   GLint i, j;
   (void) ctx;

   i = nearest_texel_location(tObj->Sampler.WrapS, img, width, texcoord[0]);
   j = nearest_texel_location(tObj->Sampler.WrapT, img, height, texcoord[1]);

   /* skip over the border, if any */
   i += img->Border;
   j += img->Border;

   if (i < 0 || i >= (GLint) img->Width || j < 0 || j >= (GLint) img->Height) {
      /* Need this test for GL_CLAMP_TO_BORDER mode */
      get_border_color(tObj, img, rgba);
   }
   else {
      swImg->FetchTexel(swImg, i, j, 0, rgba);
   }
}


/**
 * Return the texture sample for coordinate (s,t) using GL_LINEAR filter.
 * New sampling code contributed by Lynn Quam <quam@ai.sri.com>.
 */
static inline void
sample_2d_linear(struct gl_context *ctx,
                 const struct gl_texture_object *tObj,
                 const struct gl_texture_image *img,
                 const GLfloat texcoord[4],
                 GLfloat rgba[])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;
   const GLint height = img->Height2;
   GLint i0, j0, i1, j1;
   GLbitfield useBorderColor = 0x0;
   GLfloat a, b;
   GLfloat t00[4], t10[4], t01[4], t11[4]; /* sampled texel colors */

   linear_texel_locations(tObj->Sampler.WrapS, img, width, texcoord[0],  &i0, &i1, &a);
   linear_texel_locations(tObj->Sampler.WrapT, img, height, texcoord[1], &j0, &j1, &b);

   if (img->Border) {
      i0 += img->Border;
      i1 += img->Border;
      j0 += img->Border;
      j1 += img->Border;
   }
   else {
      if (i0 < 0 || i0 >= width)   useBorderColor |= I0BIT;
      if (i1 < 0 || i1 >= width)   useBorderColor |= I1BIT;
      if (j0 < 0 || j0 >= height)  useBorderColor |= J0BIT;
      if (j1 < 0 || j1 >= height)  useBorderColor |= J1BIT;
   }

   /* fetch four texel colors */
   if (useBorderColor & (I0BIT | J0BIT)) {
      get_border_color(tObj, img, t00);
   }
   else {
      swImg->FetchTexel(swImg, i0, j0, 0, t00);
   }
   if (useBorderColor & (I1BIT | J0BIT)) {
      get_border_color(tObj, img, t10);
   }
   else {
      swImg->FetchTexel(swImg, i1, j0, 0, t10);
   }
   if (useBorderColor & (I0BIT | J1BIT)) {
      get_border_color(tObj, img, t01);
   }
   else {
      swImg->FetchTexel(swImg, i0, j1, 0, t01);
   }
   if (useBorderColor & (I1BIT | J1BIT)) {
      get_border_color(tObj, img, t11);
   }
   else {
      swImg->FetchTexel(swImg, i1, j1, 0, t11);
   }

   lerp_rgba_2d(rgba, a, b, t00, t10, t01, t11);
}


/**
 * As above, but we know WRAP_S == REPEAT and WRAP_T == REPEAT.
 * We don't have to worry about the texture border.
 */
static inline void
sample_2d_linear_repeat(struct gl_context *ctx,
                        const struct gl_texture_object *tObj,
                        const struct gl_texture_image *img,
                        const GLfloat texcoord[4],
                        GLfloat rgba[])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;
   const GLint height = img->Height2;
   GLint i0, j0, i1, j1;
   GLfloat wi, wj;
   GLfloat t00[4], t10[4], t01[4], t11[4]; /* sampled texel colors */

   (void) ctx;

   ASSERT(tObj->Sampler.WrapS == GL_REPEAT);
   ASSERT(tObj->Sampler.WrapT == GL_REPEAT);
   ASSERT(img->Border == 0);
   ASSERT(swImg->_IsPowerOfTwo);

   linear_repeat_texel_location(width,  texcoord[0], &i0, &i1, &wi);
   linear_repeat_texel_location(height, texcoord[1], &j0, &j1, &wj);

   swImg->FetchTexel(swImg, i0, j0, 0, t00);
   swImg->FetchTexel(swImg, i1, j0, 0, t10);
   swImg->FetchTexel(swImg, i0, j1, 0, t01);
   swImg->FetchTexel(swImg, i1, j1, 0, t11);

   lerp_rgba_2d(rgba, wi, wj, t00, t10, t01, t11);
}


static void
sample_2d_nearest_mipmap_nearest(struct gl_context *ctx,
                                 const struct gl_texture_object *tObj,
                                 GLuint n, const GLfloat texcoord[][4],
                                 const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_2d_nearest(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_2d_linear_mipmap_nearest(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_2d_linear(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_2d_nearest_mipmap_linear(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_2d_nearest(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                           texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_2d_nearest(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_2d_nearest(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


static void
sample_2d_linear_mipmap_linear( struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4] )
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_2d_linear(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                          texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_2d_linear(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_2d_linear(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


static void
sample_2d_linear_mipmap_linear_repeat(struct gl_context *ctx,
                                      const struct gl_texture_object *tObj,
                                      GLuint n, const GLfloat texcoord[][4],
                                      const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   ASSERT(tObj->Sampler.WrapS == GL_REPEAT);
   ASSERT(tObj->Sampler.WrapT == GL_REPEAT);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_2d_linear_repeat(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                                 texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_2d_linear_repeat(ctx, tObj, tObj->Image[0][level  ],
                                 texcoord[i], t0);
         sample_2d_linear_repeat(ctx, tObj, tObj->Image[0][level+1],
                                 texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


/** Sample 2D texture, nearest filtering for both min/magnification */
static void
sample_nearest_2d(struct gl_context *ctx,
                  const struct gl_texture_object *tObj, GLuint n,
                  const GLfloat texcoords[][4],
                  const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   (void) lambda;
   for (i = 0; i < n; i++) {
      sample_2d_nearest(ctx, tObj, image, texcoords[i], rgba[i]);
   }
}


/** Sample 2D texture, linear filtering for both min/magnification */
static void
sample_linear_2d(struct gl_context *ctx,
                 const struct gl_texture_object *tObj, GLuint n,
                 const GLfloat texcoords[][4],
                 const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg = swrast_texture_image_const(image);
   (void) lambda;
   if (tObj->Sampler.WrapS == GL_REPEAT &&
       tObj->Sampler.WrapT == GL_REPEAT &&
       swImg->_IsPowerOfTwo &&
       image->Border == 0) {
      for (i = 0; i < n; i++) {
         sample_2d_linear_repeat(ctx, tObj, image, texcoords[i], rgba[i]);
      }
   }
   else {
      for (i = 0; i < n; i++) {
         sample_2d_linear(ctx, tObj, image, texcoords[i], rgba[i]);
      }
   }
}


/**
 * Optimized 2-D texture sampling:
 *    S and T wrap mode == GL_REPEAT
 *    GL_NEAREST min/mag filter
 *    No border, 
 *    RowStride == Width,
 *    Format = GL_RGB
 */
static void
opt_sample_rgb_2d(struct gl_context *ctx,
                  const struct gl_texture_object *tObj,
                  GLuint n, const GLfloat texcoords[][4],
                  const GLfloat lambda[], GLfloat rgba[][4])
{
   const struct gl_texture_image *img = tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLfloat width = (GLfloat) img->Width;
   const GLfloat height = (GLfloat) img->Height;
   const GLint colMask = img->Width - 1;
   const GLint rowMask = img->Height - 1;
   const GLint shift = img->WidthLog2;
   GLuint k;
   (void) ctx;
   (void) lambda;
   ASSERT(tObj->Sampler.WrapS==GL_REPEAT);
   ASSERT(tObj->Sampler.WrapT==GL_REPEAT);
   ASSERT(img->Border==0);
   ASSERT(img->TexFormat == MESA_FORMAT_RGB888);
   ASSERT(swImg->_IsPowerOfTwo);
   (void) swImg;

   for (k=0; k<n; k++) {
      GLint i = IFLOOR(texcoords[k][0] * width) & colMask;
      GLint j = IFLOOR(texcoords[k][1] * height) & rowMask;
      GLint pos = (j << shift) | i;
      GLubyte *texel = swImg->Map + 3 * pos;
      rgba[k][RCOMP] = UBYTE_TO_FLOAT(texel[2]);
      rgba[k][GCOMP] = UBYTE_TO_FLOAT(texel[1]);
      rgba[k][BCOMP] = UBYTE_TO_FLOAT(texel[0]);
      rgba[k][ACOMP] = 1.0F;
   }
}


/**
 * Optimized 2-D texture sampling:
 *    S and T wrap mode == GL_REPEAT
 *    GL_NEAREST min/mag filter
 *    No border
 *    RowStride == Width,
 *    Format = GL_RGBA
 */
static void
opt_sample_rgba_2d(struct gl_context *ctx,
                   const struct gl_texture_object *tObj,
                   GLuint n, const GLfloat texcoords[][4],
                   const GLfloat lambda[], GLfloat rgba[][4])
{
   const struct gl_texture_image *img = tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLfloat width = (GLfloat) img->Width;
   const GLfloat height = (GLfloat) img->Height;
   const GLint colMask = img->Width - 1;
   const GLint rowMask = img->Height - 1;
   const GLint shift = img->WidthLog2;
   GLuint i;
   (void) ctx;
   (void) lambda;
   ASSERT(tObj->Sampler.WrapS==GL_REPEAT);
   ASSERT(tObj->Sampler.WrapT==GL_REPEAT);
   ASSERT(img->Border==0);
   ASSERT(img->TexFormat == MESA_FORMAT_RGBA8888);
   ASSERT(swImg->_IsPowerOfTwo);
   (void) swImg;

   for (i = 0; i < n; i++) {
      const GLint col = IFLOOR(texcoords[i][0] * width) & colMask;
      const GLint row = IFLOOR(texcoords[i][1] * height) & rowMask;
      const GLint pos = (row << shift) | col;
      const GLuint texel = *((GLuint *) swImg->Map + pos);
      rgba[i][RCOMP] = UBYTE_TO_FLOAT( (texel >> 24)        );
      rgba[i][GCOMP] = UBYTE_TO_FLOAT( (texel >> 16) & 0xff );
      rgba[i][BCOMP] = UBYTE_TO_FLOAT( (texel >>  8) & 0xff );
      rgba[i][ACOMP] = UBYTE_TO_FLOAT( (texel      ) & 0xff );
   }
}


/** Sample 2D texture, using lambda to choose between min/magnification */
static void
sample_lambda_2d(struct gl_context *ctx,
                 const struct gl_texture_object *tObj,
                 GLuint n, const GLfloat texcoords[][4],
                 const GLfloat lambda[], GLfloat rgba[][4])
{
   const struct gl_texture_image *tImg = tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg = swrast_texture_image_const(tImg);
   GLuint minStart, minEnd;  /* texels with minification */
   GLuint magStart, magEnd;  /* texels with magnification */

   const GLboolean repeatNoBorderPOT = (tObj->Sampler.WrapS == GL_REPEAT)
      && (tObj->Sampler.WrapT == GL_REPEAT)
      && (tImg->Border == 0 && (tImg->Width == swImg->RowStride))
      && swImg->_IsPowerOfTwo;

   ASSERT(lambda != NULL);
   compute_min_mag_ranges(tObj, n, lambda,
                          &minStart, &minEnd, &magStart, &magEnd);

   if (minStart < minEnd) {
      /* do the minified texels */
      const GLuint m = minEnd - minStart;
      switch (tObj->Sampler.MinFilter) {
      case GL_NEAREST:
         if (repeatNoBorderPOT) {
            switch (tImg->TexFormat) {
            case MESA_FORMAT_RGB888:
               opt_sample_rgb_2d(ctx, tObj, m, texcoords + minStart,
                                 NULL, rgba + minStart);
               break;
            case MESA_FORMAT_RGBA8888:
               opt_sample_rgba_2d(ctx, tObj, m, texcoords + minStart,
                                  NULL, rgba + minStart);
               break;
            default:
               sample_nearest_2d(ctx, tObj, m, texcoords + minStart,
                                 NULL, rgba + minStart );
            }
         }
         else {
            sample_nearest_2d(ctx, tObj, m, texcoords + minStart,
                              NULL, rgba + minStart);
         }
         break;
      case GL_LINEAR:
         sample_linear_2d(ctx, tObj, m, texcoords + minStart,
                          NULL, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_NEAREST:
         sample_2d_nearest_mipmap_nearest(ctx, tObj, m,
                                          texcoords + minStart,
                                          lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_NEAREST:
         sample_2d_linear_mipmap_nearest(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_LINEAR:
         sample_2d_nearest_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_LINEAR:
         if (repeatNoBorderPOT)
            sample_2d_linear_mipmap_linear_repeat(ctx, tObj, m,
                  texcoords + minStart, lambda + minStart, rgba + minStart);
         else
            sample_2d_linear_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                        lambda + minStart, rgba + minStart);
         break;
      default:
         _mesa_problem(ctx, "Bad min filter in sample_2d_texture");
         return;
      }
   }

   if (magStart < magEnd) {
      /* do the magnified texels */
      const GLuint m = magEnd - magStart;

      switch (tObj->Sampler.MagFilter) {
      case GL_NEAREST:
         if (repeatNoBorderPOT) {
            switch (tImg->TexFormat) {
            case MESA_FORMAT_RGB888:
               opt_sample_rgb_2d(ctx, tObj, m, texcoords + magStart,
                                 NULL, rgba + magStart);
               break;
            case MESA_FORMAT_RGBA8888:
               opt_sample_rgba_2d(ctx, tObj, m, texcoords + magStart,
                                  NULL, rgba + magStart);
               break;
            default:
               sample_nearest_2d(ctx, tObj, m, texcoords + magStart,
                                 NULL, rgba + magStart );
            }
         }
         else {
            sample_nearest_2d(ctx, tObj, m, texcoords + magStart,
                              NULL, rgba + magStart);
         }
         break;
      case GL_LINEAR:
         sample_linear_2d(ctx, tObj, m, texcoords + magStart,
                          NULL, rgba + magStart);
         break;
      default:
         _mesa_problem(ctx, "Bad mag filter in sample_lambda_2d");
         break;
      }
   }
}


/* For anisotropic filtering */
#define WEIGHT_LUT_SIZE 1024

static GLfloat *weightLut = NULL;

/**
 * Creates the look-up table used to speed-up EWA sampling
 */
static void
create_filter_table(void)
{
   GLuint i;
   if (!weightLut) {
      weightLut = (GLfloat *) malloc(WEIGHT_LUT_SIZE * sizeof(GLfloat));

      for (i = 0; i < WEIGHT_LUT_SIZE; ++i) {
         GLfloat alpha = 2;
         GLfloat r2 = (GLfloat) i / (GLfloat) (WEIGHT_LUT_SIZE - 1);
         GLfloat weight = (GLfloat) exp(-alpha * r2);
         weightLut[i] = weight;
      }
   }
}


/**
 * Elliptical weighted average (EWA) filter for producing high quality
 * anisotropic filtered results.
 * Based on the Higher Quality Elliptical Weighted Avarage Filter
 * published by Paul S. Heckbert in his Master's Thesis
 * "Fundamentals of Texture Mapping and Image Warping" (1989)
 */
static void
sample_2d_ewa(struct gl_context *ctx,
              const struct gl_texture_object *tObj,
              const GLfloat texcoord[4],
              const GLfloat dudx, const GLfloat dvdx,
              const GLfloat dudy, const GLfloat dvdy, const GLint lod,
              GLfloat rgba[])
{
   GLint level = lod > 0 ? lod : 0;
   GLfloat scaling = 1.0 / (1 << level);
   const struct gl_texture_image *img = tObj->Image[0][level];
   const struct gl_texture_image *mostDetailedImage =
      tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg =
      swrast_texture_image_const(mostDetailedImage);
   GLfloat tex_u=-0.5 + texcoord[0] * swImg->WidthScale * scaling;
   GLfloat tex_v=-0.5 + texcoord[1] * swImg->HeightScale * scaling;

   GLfloat ux = dudx * scaling;
   GLfloat vx = dvdx * scaling;
   GLfloat uy = dudy * scaling;
   GLfloat vy = dvdy * scaling;

   /* compute ellipse coefficients to bound the region: 
    * A*x*x + B*x*y + C*y*y = F.
    */
   GLfloat A = vx*vx+vy*vy+1;
   GLfloat B = -2*(ux*vx+uy*vy);
   GLfloat C = ux*ux+uy*uy+1;
   GLfloat F = A*C-B*B/4.0;

   /* check if it is an ellipse */
   /* ASSERT(F > 0.0); */

   /* Compute the ellipse's (u,v) bounding box in texture space */
   GLfloat d = -B*B+4.0*C*A;
   GLfloat box_u = 2.0 / d * sqrt(d*C*F); /* box_u -> half of bbox with   */
   GLfloat box_v = 2.0 / d * sqrt(A*d*F); /* box_v -> half of bbox height */

   GLint u0 = floor(tex_u - box_u);
   GLint u1 = ceil (tex_u + box_u);
   GLint v0 = floor(tex_v - box_v);
   GLint v1 = ceil (tex_v + box_v);

   GLfloat num[4] = {0.0F, 0.0F, 0.0F, 0.0F};
   GLfloat newCoord[2];
   GLfloat den = 0.0F;
   GLfloat ddq;
   GLfloat U = u0 - tex_u;
   GLint v;

   /* Scale ellipse formula to directly index the Filter Lookup Table.
    * i.e. scale so that F = WEIGHT_LUT_SIZE-1
    */
   double formScale = (double) (WEIGHT_LUT_SIZE - 1) / F;
   A *= formScale;
   B *= formScale;
   C *= formScale;
   /* F *= formScale; */ /* no need to scale F as we don't use it below here */

   /* Heckbert MS thesis, p. 59; scan over the bounding box of the ellipse
    * and incrementally update the value of Ax^2+Bxy*Cy^2; when this
    * value, q, is less than F, we're inside the ellipse
    */
   ddq = 2 * A;
   for (v = v0; v <= v1; ++v) {
      GLfloat V = v - tex_v;
      GLfloat dq = A * (2 * U + 1) + B * V;
      GLfloat q = (C * V + B * U) * V + A * U * U;

      GLint u;
      for (u = u0; u <= u1; ++u) {
         /* Note that the ellipse has been pre-scaled so F = WEIGHT_LUT_SIZE - 1 */
         if (q < WEIGHT_LUT_SIZE) {
            /* as a LUT is used, q must never be negative;
             * should not happen, though
             */
            const GLint qClamped = q >= 0.0F ? q : 0;
            GLfloat weight = weightLut[qClamped];

            newCoord[0] = u / ((GLfloat) img->Width2);
            newCoord[1] = v / ((GLfloat) img->Height2);

            sample_2d_nearest(ctx, tObj, img, newCoord, rgba);
            num[0] += weight * rgba[0];
            num[1] += weight * rgba[1];
            num[2] += weight * rgba[2];
            num[3] += weight * rgba[3];

            den += weight;
         }
         q += dq;
         dq += ddq;
      }
   }

   if (den <= 0.0F) {
      /* Reaching this place would mean
       * that no pixels intersected the ellipse.
       * This should never happen because
       * the filter we use always
       * intersects at least one pixel.
       */

      /*rgba[0]=0;
      rgba[1]=0;
      rgba[2]=0;
      rgba[3]=0;*/
      /* not enough pixels in resampling, resort to direct interpolation */
      sample_2d_linear(ctx, tObj, img, texcoord, rgba);
      return;
   }

   rgba[0] = num[0] / den;
   rgba[1] = num[1] / den;
   rgba[2] = num[2] / den;
   rgba[3] = num[3] / den;
}


/**
 * Anisotropic filtering using footprint assembly as outlined in the
 * EXT_texture_filter_anisotropic spec:
 * http://www.opengl.org/registry/specs/EXT/texture_filter_anisotropic.txt
 * Faster than EWA but has less quality (more aliasing effects)
 */
static void
sample_2d_footprint(struct gl_context *ctx,
                 const struct gl_texture_object *tObj,
                 const GLfloat texcoord[4],
                 const GLfloat dudx, const GLfloat dvdx,
                 const GLfloat dudy, const GLfloat dvdy, const GLint lod,
                 GLfloat rgba[])
{
   GLint level = lod > 0 ? lod : 0;
   GLfloat scaling = 1.0F / (1 << level);
   const struct gl_texture_image *img = tObj->Image[0][level];

   GLfloat ux = dudx * scaling;
   GLfloat vx = dvdx * scaling;
   GLfloat uy = dudy * scaling;
   GLfloat vy = dvdy * scaling;

   GLfloat Px2 = ux * ux + vx * vx; /* squared length of dx */
   GLfloat Py2 = uy * uy + vy * vy; /* squared length of dy */

   GLint numSamples;
   GLfloat ds;
   GLfloat dt;

   GLfloat num[4] = {0.0F, 0.0F, 0.0F, 0.0F};
   GLfloat newCoord[2];
   GLint s;

   /*  Calculate the per anisotropic sample offsets in s,t space. */
   if (Px2 > Py2) {
      numSamples = ceil(SQRTF(Px2));
      ds = ux / ((GLfloat) img->Width2);
      dt = vx / ((GLfloat) img->Height2);
   }
   else {
      numSamples = ceil(SQRTF(Py2));
      ds = uy / ((GLfloat) img->Width2);
      dt = vy / ((GLfloat) img->Height2);
   }

   for (s = 0; s<numSamples; s++) {
      newCoord[0] = texcoord[0] + ds * ((GLfloat)(s+1) / (numSamples+1) -0.5);
      newCoord[1] = texcoord[1] + dt * ((GLfloat)(s+1) / (numSamples+1) -0.5);

      sample_2d_linear(ctx, tObj, img, newCoord, rgba);
      num[0] += rgba[0];
      num[1] += rgba[1];
      num[2] += rgba[2];
      num[3] += rgba[3];
   }

   rgba[0] = num[0] / numSamples;
   rgba[1] = num[1] / numSamples;
   rgba[2] = num[2] / numSamples;
   rgba[3] = num[3] / numSamples;
}


/**
 * Returns the index of the specified texture object in the
 * gl_context texture unit array.
 */
static inline GLuint
texture_unit_index(const struct gl_context *ctx,
                   const struct gl_texture_object *tObj)
{
   const GLuint maxUnit
      = (ctx->Texture._EnabledCoordUnits > 1) ? ctx->Const.MaxTextureUnits : 1;
   GLuint u;

   /* XXX CoordUnits vs. ImageUnits */
   for (u = 0; u < maxUnit; u++) {
      if (ctx->Texture.Unit[u]._Current == tObj)
         break; /* found */
   }
   if (u >= maxUnit)
      u = 0; /* not found, use 1st one; should never happen */
   
   return u;
}


/**
 * Sample 2D texture using an anisotropic filter.
 * NOTE: the const GLfloat lambda_iso[] parameter does *NOT* contain
 * the lambda float array but a "hidden" SWspan struct which is required
 * by this function but is not available in the texture_sample_func signature.
 * See _swrast_texture_span( struct gl_context *ctx, SWspan *span ) on how
 * this function is called.
 */
static void
sample_lambda_2d_aniso(struct gl_context *ctx,
                       const struct gl_texture_object *tObj,
                       GLuint n, const GLfloat texcoords[][4],
                       const GLfloat lambda_iso[], GLfloat rgba[][4])
{
   const struct gl_texture_image *tImg = tObj->Image[0][tObj->BaseLevel];
   const struct swrast_texture_image *swImg = swrast_texture_image_const(tImg);
   const GLfloat maxEccentricity =
      tObj->Sampler.MaxAnisotropy * tObj->Sampler.MaxAnisotropy;
   
   /* re-calculate the lambda values so that they are usable with anisotropic
    * filtering
    */
   SWspan *span = (SWspan *)lambda_iso; /* access the "hidden" SWspan struct */

   /* based on interpolate_texcoords(struct gl_context *ctx, SWspan *span)
    * in swrast/s_span.c
    */
   
   /* find the texture unit index by looking up the current texture object
    * from the context list of available texture objects.
    */
   const GLuint u = texture_unit_index(ctx, tObj);
   const GLuint attr = FRAG_ATTRIB_TEX0 + u;
   GLfloat texW, texH;

   const GLfloat dsdx = span->attrStepX[attr][0];
   const GLfloat dsdy = span->attrStepY[attr][0];
   const GLfloat dtdx = span->attrStepX[attr][1];
   const GLfloat dtdy = span->attrStepY[attr][1];
   const GLfloat dqdx = span->attrStepX[attr][3];
   const GLfloat dqdy = span->attrStepY[attr][3];
   GLfloat s = span->attrStart[attr][0] + span->leftClip * dsdx;
   GLfloat t = span->attrStart[attr][1] + span->leftClip * dtdx;
   GLfloat q = span->attrStart[attr][3] + span->leftClip * dqdx;

   GLuint i;
   
   /* on first access create the lookup table containing the filter weights. */
   if (!weightLut) {
      create_filter_table();
   }

   texW = swImg->WidthScale;
   texH = swImg->HeightScale;

   for (i = 0; i < n; i++) {
      const GLfloat invQ = (q == 0.0F) ? 1.0F : (1.0F / q);
      
      GLfloat dudx = texW * ((s + dsdx) / (q + dqdx) - s * invQ);
      GLfloat dvdx = texH * ((t + dtdx) / (q + dqdx) - t * invQ);
      GLfloat dudy = texW * ((s + dsdy) / (q + dqdy) - s * invQ);
      GLfloat dvdy = texH * ((t + dtdy) / (q + dqdy) - t * invQ);
      
      /* note: instead of working with Px and Py, we will use the 
       * squared length instead, to avoid sqrt.
       */
      GLfloat Px2 = dudx * dudx + dvdx * dvdx;
      GLfloat Py2 = dudy * dudy + dvdy * dvdy;

      GLfloat Pmax2;
      GLfloat Pmin2;
      GLfloat e;
      GLfloat lod;

      s += dsdx;
      t += dtdx;
      q += dqdx;
      
      if (Px2 < Py2) {
         Pmax2 = Py2;
         Pmin2 = Px2;
      }
      else {
         Pmax2 = Px2;
         Pmin2 = Py2;
      }
      
      /* if the eccentricity of the ellipse is too big, scale up the shorter
       * of the two vectors to limit the maximum amount of work per pixel
       */
      e = Pmax2 / Pmin2;
      if (e > maxEccentricity) {
         /* GLfloat s=e / maxEccentricity;
            minor[0] *= s;
            minor[1] *= s;
            Pmin2 *= s; */
         Pmin2 = Pmax2 / maxEccentricity;
      }
      
      /* note: we need to have Pmin=sqrt(Pmin2) here, but we can avoid
       * this since 0.5*log(x) = log(sqrt(x))
       */
      lod = 0.5 * LOG2(Pmin2);
      
      /* If the ellipse covers the whole image, we can
       * simply return the average of the whole image.
       */
      if (lod >= tObj->_MaxLevel) {
         sample_2d_linear(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                          texcoords[i], rgba[i]);
      }
      else {
         /* don't bother interpolating between multiple LODs; it doesn't
          * seem to be worth the extra running time.
          */
         sample_2d_ewa(ctx, tObj, texcoords[i],
                       dudx, dvdx, dudy, dvdy, floor(lod), rgba[i]);

         /* unused: */
         (void) sample_2d_footprint;
         /*
         sample_2d_footprint(ctx, tObj, texcoords[i],
                             dudx, dvdx, dudy, dvdy, floor(lod), rgba[i]);
         */
      }
   }
}



/**********************************************************************/
/*                    3-D Texture Sampling Functions                  */
/**********************************************************************/

/**
 * Return the texture sample for coordinate (s,t,r) using GL_NEAREST filter.
 */
static inline void
sample_3d_nearest(struct gl_context *ctx,
                  const struct gl_texture_object *tObj,
                  const struct gl_texture_image *img,
                  const GLfloat texcoord[4],
                  GLfloat rgba[4])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;     /* without border, power of two */
   const GLint height = img->Height2;   /* without border, power of two */
   const GLint depth = img->Depth2;     /* without border, power of two */
   GLint i, j, k;
   (void) ctx;

   i = nearest_texel_location(tObj->Sampler.WrapS, img, width, texcoord[0]);
   j = nearest_texel_location(tObj->Sampler.WrapT, img, height, texcoord[1]);
   k = nearest_texel_location(tObj->Sampler.WrapR, img, depth, texcoord[2]);

   if (i < 0 || i >= (GLint) img->Width ||
       j < 0 || j >= (GLint) img->Height ||
       k < 0 || k >= (GLint) img->Depth) {
      /* Need this test for GL_CLAMP_TO_BORDER mode */
      get_border_color(tObj, img, rgba);
   }
   else {
      swImg->FetchTexel(swImg, i, j, k, rgba);
   }
}


/**
 * Return the texture sample for coordinate (s,t,r) using GL_LINEAR filter.
 */
static void
sample_3d_linear(struct gl_context *ctx,
                 const struct gl_texture_object *tObj,
                 const struct gl_texture_image *img,
                 const GLfloat texcoord[4],
                 GLfloat rgba[4])
{
   const struct swrast_texture_image *swImg = swrast_texture_image_const(img);
   const GLint width = img->Width2;
   const GLint height = img->Height2;
   const GLint depth = img->Depth2;
   GLint i0, j0, k0, i1, j1, k1;
   GLbitfield useBorderColor = 0x0;
   GLfloat a, b, c;
   GLfloat t000[4], t010[4], t001[4], t011[4];
   GLfloat t100[4], t110[4], t101[4], t111[4];

   linear_texel_locations(tObj->Sampler.WrapS, img, width, texcoord[0],  &i0, &i1, &a);
   linear_texel_locations(tObj->Sampler.WrapT, img, height, texcoord[1], &j0, &j1, &b);
   linear_texel_locations(tObj->Sampler.WrapR, img, depth, texcoord[2],  &k0, &k1, &c);

   if (img->Border) {
      i0 += img->Border;
      i1 += img->Border;
      j0 += img->Border;
      j1 += img->Border;
      k0 += img->Border;
      k1 += img->Border;
   }
   else {
      /* check if sampling texture border color */
      if (i0 < 0 || i0 >= width)   useBorderColor |= I0BIT;
      if (i1 < 0 || i1 >= width)   useBorderColor |= I1BIT;
      if (j0 < 0 || j0 >= height)  useBorderColor |= J0BIT;
      if (j1 < 0 || j1 >= height)  useBorderColor |= J1BIT;
      if (k0 < 0 || k0 >= depth)   useBorderColor |= K0BIT;
      if (k1 < 0 || k1 >= depth)   useBorderColor |= K1BIT;
   }

   /* Fetch texels */
   if (useBorderColor & (I0BIT | J0BIT | K0BIT)) {
      get_border_color(tObj, img, t000);
   }
   else {
      swImg->FetchTexel(swImg, i0, j0, k0, t000);
   }
   if (useBorderColor & (I1BIT | J0BIT | K0BIT)) {
      get_border_color(tObj, img, t100);
   }
   else {
      swImg->FetchTexel(swImg, i1, j0, k0, t100);
   }
   if (useBorderColor & (I0BIT | J1BIT | K0BIT)) {
      get_border_color(tObj, img, t010);
   }
   else {
      swImg->FetchTexel(swImg, i0, j1, k0, t010);
   }
   if (useBorderColor & (I1BIT | J1BIT | K0BIT)) {
      get_border_color(tObj, img, t110);
   }
   else {
      swImg->FetchTexel(swImg, i1, j1, k0, t110);
   }

   if (useBorderColor & (I0BIT | J0BIT | K1BIT)) {
      get_border_color(tObj, img, t001);
   }
   else {
      swImg->FetchTexel(swImg, i0, j0, k1, t001);
   }
   if (useBorderColor & (I1BIT | J0BIT | K1BIT)) {
      get_border_color(tObj, img, t101);
   }
   else {
      swImg->FetchTexel(swImg, i1, j0, k1, t101);
   }
   if (useBorderColor & (I0BIT | J1BIT | K1BIT)) {
      get_border_color(tObj, img, t011);
   }
   else {
      swImg->FetchTexel(swImg, i0, j1, k1, t011);
   }
   if (useBorderColor & (I1BIT | J1BIT | K1BIT)) {
      get_border_color(tObj, img, t111);
   }
   else {
      swImg->FetchTexel(swImg, i1, j1, k1, t111);
   }

   /* trilinear interpolation of samples */
   lerp_rgba_3d(rgba, a, b, c, t000, t100, t010, t110, t001, t101, t011, t111);
}


static void
sample_3d_nearest_mipmap_nearest(struct gl_context *ctx,
                                 const struct gl_texture_object *tObj,
                                 GLuint n, const GLfloat texcoord[][4],
                                 const GLfloat lambda[], GLfloat rgba[][4] )
{
   GLuint i;
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_3d_nearest(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_3d_linear_mipmap_nearest(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      sample_3d_linear(ctx, tObj, tObj->Image[0][level], texcoord[i], rgba[i]);
   }
}


static void
sample_3d_nearest_mipmap_linear(struct gl_context *ctx,
                                const struct gl_texture_object *tObj,
                                GLuint n, const GLfloat texcoord[][4],
                                const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_3d_nearest(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                           texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_3d_nearest(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_3d_nearest(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


static void
sample_3d_linear_mipmap_linear(struct gl_context *ctx,
                               const struct gl_texture_object *tObj,
                               GLuint n, const GLfloat texcoord[][4],
                               const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      if (level >= tObj->_MaxLevel) {
         sample_3d_linear(ctx, tObj, tObj->Image[0][tObj->_MaxLevel],
                          texcoord[i], rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_3d_linear(ctx, tObj, tObj->Image[0][level  ], texcoord[i], t0);
         sample_3d_linear(ctx, tObj, tObj->Image[0][level+1], texcoord[i], t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


/** Sample 3D texture, nearest filtering for both min/magnification */
static void
sample_nearest_3d(struct gl_context *ctx,
                  const struct gl_texture_object *tObj, GLuint n,
                  const GLfloat texcoords[][4], const GLfloat lambda[],
                  GLfloat rgba[][4])
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   (void) lambda;
   for (i = 0; i < n; i++) {
      sample_3d_nearest(ctx, tObj, image, texcoords[i], rgba[i]);
   }
}


/** Sample 3D texture, linear filtering for both min/magnification */
static void
sample_linear_3d(struct gl_context *ctx,
                 const struct gl_texture_object *tObj, GLuint n,
                 const GLfloat texcoords[][4],
                 const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   struct gl_texture_image *image = tObj->Image[0][tObj->BaseLevel];
   (void) lambda;
   for (i = 0; i < n; i++) {
      sample_3d_linear(ctx, tObj, image, texcoords[i], rgba[i]);
   }
}


/** Sample 3D texture, using lambda to choose between min/magnification */
static void
sample_lambda_3d(struct gl_context *ctx,
                 const struct gl_texture_object *tObj, GLuint n,
                 const GLfloat texcoords[][4], const GLfloat lambda[],
                 GLfloat rgba[][4])
{
   GLuint minStart, minEnd;  /* texels with minification */
   GLuint magStart, magEnd;  /* texels with magnification */
   GLuint i;

   ASSERT(lambda != NULL);
   compute_min_mag_ranges(tObj, n, lambda,
                          &minStart, &minEnd, &magStart, &magEnd);

   if (minStart < minEnd) {
      /* do the minified texels */
      GLuint m = minEnd - minStart;
      switch (tObj->Sampler.MinFilter) {
      case GL_NEAREST:
         for (i = minStart; i < minEnd; i++)
            sample_3d_nearest(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                              texcoords[i], rgba[i]);
         break;
      case GL_LINEAR:
         for (i = minStart; i < minEnd; i++)
            sample_3d_linear(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                             texcoords[i], rgba[i]);
         break;
      case GL_NEAREST_MIPMAP_NEAREST:
         sample_3d_nearest_mipmap_nearest(ctx, tObj, m, texcoords + minStart,
                                          lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_NEAREST:
         sample_3d_linear_mipmap_nearest(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_LINEAR:
         sample_3d_nearest_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                         lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_LINEAR:
         sample_3d_linear_mipmap_linear(ctx, tObj, m, texcoords + minStart,
                                        lambda + minStart, rgba + minStart);
         break;
      default:
         _mesa_problem(ctx, "Bad min filter in sample_3d_texture");
         return;
      }
   }

   if (magStart < magEnd) {
      /* do the magnified texels */
      switch (tObj->Sampler.MagFilter) {
      case GL_NEAREST:
         for (i = magStart; i < magEnd; i++)
            sample_3d_nearest(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                              texcoords[i], rgba[i]);
         break;
      case GL_LINEAR:
         for (i = magStart; i < magEnd; i++)
            sample_3d_linear(ctx, tObj, tObj->Image[0][tObj->BaseLevel],
                             texcoords[i], rgba[i]);
         break;
      default:
         _mesa_problem(ctx, "Bad mag filter in sample_3d_texture");
         return;
      }
   }
}


/**********************************************************************/
/*                Texture Cube Map Sampling Functions                 */
/**********************************************************************/

/**
 * Choose one of six sides of a texture cube map given the texture
 * coord (rx,ry,rz).  Return pointer to corresponding array of texture
 * images.
 */
static const struct gl_texture_image **
choose_cube_face(const struct gl_texture_object *texObj,
                 const GLfloat texcoord[4], GLfloat newCoord[4])
{
   /*
      major axis
      direction     target                             sc     tc    ma
      ----------    -------------------------------    ---    ---   ---
       +rx          TEXTURE_CUBE_MAP_POSITIVE_X_EXT    -rz    -ry   rx
       -rx          TEXTURE_CUBE_MAP_NEGATIVE_X_EXT    +rz    -ry   rx
       +ry          TEXTURE_CUBE_MAP_POSITIVE_Y_EXT    +rx    +rz   ry
       -ry          TEXTURE_CUBE_MAP_NEGATIVE_Y_EXT    +rx    -rz   ry
       +rz          TEXTURE_CUBE_MAP_POSITIVE_Z_EXT    +rx    -ry   rz
       -rz          TEXTURE_CUBE_MAP_NEGATIVE_Z_EXT    -rx    -ry   rz
   */
   const GLfloat rx = texcoord[0];
   const GLfloat ry = texcoord[1];
   const GLfloat rz = texcoord[2];
   const GLfloat arx = FABSF(rx), ary = FABSF(ry), arz = FABSF(rz);
   GLuint face;
   GLfloat sc, tc, ma;

   if (arx >= ary && arx >= arz) {
      if (rx >= 0.0F) {
         face = FACE_POS_X;
         sc = -rz;
         tc = -ry;
         ma = arx;
      }
      else {
         face = FACE_NEG_X;
         sc = rz;
         tc = -ry;
         ma = arx;
      }
   }
   else if (ary >= arx && ary >= arz) {
      if (ry >= 0.0F) {
         face = FACE_POS_Y;
         sc = rx;
         tc = rz;
         ma = ary;
      }
      else {
         face = FACE_NEG_Y;
         sc = rx;
         tc = -rz;
         ma = ary;
      }
   }
   else {
      if (rz > 0.0F) {
         face = FACE_POS_Z;
         sc = rx;
         tc = -ry;
         ma = arz;
      }
      else {
         face = FACE_NEG_Z;
         sc = -rx;
         tc = -ry;
         ma = arz;
      }
   }

   { 
      const float ima = 1.0F / ma;
      newCoord[0] = ( sc * ima + 1.0F ) * 0.5F;
      newCoord[1] = ( tc * ima + 1.0F ) * 0.5F;
   }

   return (const struct gl_texture_image **) texObj->Image[face];
}


static void
sample_nearest_cube(struct gl_context *ctx,
                    const struct gl_texture_object *tObj, GLuint n,
                    const GLfloat texcoords[][4], const GLfloat lambda[],
                    GLfloat rgba[][4])
{
   GLuint i;
   (void) lambda;
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      images = choose_cube_face(tObj, texcoords[i], newCoord);
      sample_2d_nearest(ctx, tObj, images[tObj->BaseLevel],
                        newCoord, rgba[i]);
   }
}


static void
sample_linear_cube(struct gl_context *ctx,
                   const struct gl_texture_object *tObj, GLuint n,
                   const GLfloat texcoords[][4],
                   const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   (void) lambda;
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      images = choose_cube_face(tObj, texcoords[i], newCoord);
      sample_2d_linear(ctx, tObj, images[tObj->BaseLevel],
                       newCoord, rgba[i]);
   }
}


static void
sample_cube_nearest_mipmap_nearest(struct gl_context *ctx,
                                   const struct gl_texture_object *tObj,
                                   GLuint n, const GLfloat texcoord[][4],
                                   const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      GLint level;
      images = choose_cube_face(tObj, texcoord[i], newCoord);

      /* XXX we actually need to recompute lambda here based on the newCoords.
       * But we would need the texcoords of adjacent fragments to compute that
       * properly, and we don't have those here.
       * For now, do an approximation:  subtracting 1 from the chosen mipmap
       * level seems to work in some test cases.
       * The same adjustment is done in the next few functions.
      */
      level = nearest_mipmap_level(tObj, lambda[i]);
      level = MAX2(level - 1, 0);

      sample_2d_nearest(ctx, tObj, images[level], newCoord, rgba[i]);
   }
}


static void
sample_cube_linear_mipmap_nearest(struct gl_context *ctx,
                                  const struct gl_texture_object *tObj,
                                  GLuint n, const GLfloat texcoord[][4],
                                  const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      GLint level = nearest_mipmap_level(tObj, lambda[i]);
      level = MAX2(level - 1, 0); /* see comment above */
      images = choose_cube_face(tObj, texcoord[i], newCoord);
      sample_2d_linear(ctx, tObj, images[level], newCoord, rgba[i]);
   }
}


static void
sample_cube_nearest_mipmap_linear(struct gl_context *ctx,
                                  const struct gl_texture_object *tObj,
                                  GLuint n, const GLfloat texcoord[][4],
                                  const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      level = MAX2(level - 1, 0); /* see comment above */
      images = choose_cube_face(tObj, texcoord[i], newCoord);
      if (level >= tObj->_MaxLevel) {
         sample_2d_nearest(ctx, tObj, images[tObj->_MaxLevel],
                           newCoord, rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];  /* texels */
         const GLfloat f = FRAC(lambda[i]);
         sample_2d_nearest(ctx, tObj, images[level  ], newCoord, t0);
         sample_2d_nearest(ctx, tObj, images[level+1], newCoord, t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


static void
sample_cube_linear_mipmap_linear(struct gl_context *ctx,
                                 const struct gl_texture_object *tObj,
                                 GLuint n, const GLfloat texcoord[][4],
                                 const GLfloat lambda[], GLfloat rgba[][4])
{
   GLuint i;
   ASSERT(lambda != NULL);
   for (i = 0; i < n; i++) {
      const struct gl_texture_image **images;
      GLfloat newCoord[4];
      GLint level = linear_mipmap_level(tObj, lambda[i]);
      level = MAX2(level - 1, 0); /* see comment above */
      images = choose_cube_face(tObj, texcoord[i], newCoord);
      if (level >= tObj->_MaxLevel) {
         sample_2d_linear(ctx, tObj, images[tObj->_MaxLevel],
                          newCoord, rgba[i]);
      }
      else {
         GLfloat t0[4], t1[4];
         const GLfloat f = FRAC(lambda[i]);
         sample_2d_linear(ctx, tObj, images[level  ], newCoord, t0);
         sample_2d_linear(ctx, tObj, images[level+1], newCoord, t1);
         lerp_rgba(rgba[i], f, t0, t1);
      }
   }
}


/** Sample cube texture, using lambda to choose between min/magnification */
static void
sample_lambda_cube(struct gl_context *ctx,
                   const struct gl_texture_object *tObj, GLuint n,
                   const GLfloat texcoords[][4], const GLfloat lambda[],
                   GLfloat rgba[][4])
{
   GLuint minStart, minEnd;  /* texels with minification */
   GLuint magStart, magEnd;  /* texels with magnification */

   ASSERT(lambda != NULL);
   compute_min_mag_ranges(tObj, n, lambda,
                          &minStart, &minEnd, &magStart, &magEnd);

   if (minStart < minEnd) {
      /* do the minified texels */
      const GLuint m = minEnd - minStart;
      switch (tObj->Sampler.MinFilter) {
      case GL_NEAREST:
         sample_nearest_cube(ctx, tObj, m, texcoords + minStart,
                             lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR:
         sample_linear_cube(ctx, tObj, m, texcoords + minStart,
                            lambda + minStart, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_NEAREST:
         sample_cube_nearest_mipmap_nearest(ctx, tObj, m,
                                            texcoords + minStart,
                                           lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_NEAREST:
         sample_cube_linear_mipmap_nearest(ctx, tObj, m,
                                           texcoords + minStart,
                                           lambda + minStart, rgba + minStart);
         break;
      case GL_NEAREST_MIPMAP_LINEAR:
         sample_cube_nearest_mipmap_linear(ctx, tObj, m,
                                           texcoords + minStart,
                                           lambda + minStart, rgba + minStart);
         break;
      case GL_LINEAR_MIPMAP_LINEAR:
         sample_cube_linear_mipmap_linear(ctx, tObj, m,
                                          texcoords + minStart,
                                          lambda + minStart, rgba + minStart);
         break;
      default:
         _mesa_problem(ctx, "Bad min filter in sample_lambda_cube");
         break;
      }
   }

   if (magStart < magEnd) {
      /* do the magnified texels */
      const GLuint m = magEnd - magStart;
      switch (tObj->Sampler.MagFilter) {
      case GL_NEAREST:
         sample_nearest_cube(ctx, tObj, m, texcoords + magStart,
                             lambda + magStart, rgba + magStart);
         break;
      case GL_LINEAR:
         sample_linear_cube(ctx, tObj, m, texcoords + magStart,
                            lambda + magStart, rgba + magStart);
         break;
      default:
         _mesa_problem(ctx, "Bad mag filter in sample_lambda_cube");
         break;
      }
   }
}

/**
 * Compare texcoord against depth sample.  Return 1.0 or the ambient value.
 */
static inline GLfloat
shadow_compare(GLenum function, GLfloat coord, GLfloat depthSample,
               GLfloat ambient)
{
   switch (function) {
   case GL_LEQUAL:
      return (coord <= depthSample) ? 1.0F : ambient;
   case GL_GEQUAL:
      return (coord >= depthSample) ? 1.0F : ambient;
   case GL_LESS:
      return (coord < depthSample) ? 1.0F : ambient;
   case GL_GREATER:
      return (coord > depthSample) ? 1.0F : ambient;
   case GL_EQUAL:
      return (coord == depthSample) ? 1.0F : ambient;
   case GL_NOTEQUAL:
      return (coord != depthSample) ? 1.0F : ambient;
   case GL_ALWAYS:
      return 1.0F;
   case GL_NEVER:
      return ambient;
   case GL_NONE:
      return depthSample;
   default:
      _mesa_problem(NULL, "Bad compare func in shadow_compare");
      return ambient;
   }
}


/**
 * Compare texcoord against four depth samples.
 */
static inline GLfloat
shadow_compare4(GLenum function, GLfloat coord,
                GLfloat depth00, GLfloat depth01,
                GLfloat depth10, GLfloat depth11,
                GLfloat ambient, GLfloat wi, GLfloat wj)
{
   const GLfloat d = (1.0F - (GLfloat) ambient) * 0.25F;
   GLfloat luminance = 1.0F;

   switch (function) {
   case GL_LEQUAL:
      if (coord > depth00)  luminance -= d;
      if (coord > depth01)  luminance -= d;
      if (coord > depth10)  luminance -= d;
      if (coord > depth11)  luminance -= d;
      return luminance;
   case GL_GEQUAL:
      if (coord < depth00)  luminance -= d;
      if (coord < depth01)  luminance -= d;
      if (coord < depth10)  luminance -= d;
      if (coord < depth11)  luminance -= d;
      return luminance;
   case GL_LESS:
      if (coord >= depth00)  luminance -= d;
      if (coord >= depth01)  luminance -= d;
      if (coord >= depth10)  luminance -= d;
      if (coord >= depth11)  luminance -= d;
      return luminance;
   case GL_GREATER:
      if (coord <= depth00)  luminance -= d;
      if (coord <= depth01)  luminance -= d;
      if (coord <= depth10)  luminance -= d;
      if (coord <= depth11)  luminance -= d;
      return luminance;
   case GL_EQUAL:
      if (coord != depth00)  luminance -= d;
      if (coord != depth01)  luminance -= d;
      if (coord != depth10)  luminance -= d;
      if (coord != depth11)  luminance -= d;
      return luminance;
   case GL_NOTEQUAL:
      if (coord == depth00)  luminance -= d;
      if (coord == depth01)  luminance -= d;
      if (coord == depth10)  luminance -= d;
      if (coord == depth11)  luminance -= d;
      return luminance;
   case GL_ALWAYS:
      return 1.0F;
   case GL_NEVER:
      return ambient;
   case GL_NONE:
      /* ordinary bilinear filtering */
      return lerp_2d(wi, wj, depth00, depth10, depth01, depth11);
   default:
      _mesa_problem(NULL, "Bad compare func in sample_compare4");
      return ambient;
   }
}

/**
 * We use this function when a texture object is in an "incomplete" state.
 * When a fragment program attempts to sample an incomplete texture we
 * return black (see issue 23 in GL_ARB_fragment_program spec).
 * Note: fragment programs don't observe the texture enable/disable flags.
 */
static void
null_sample_func( struct gl_context *ctx,
                  const struct gl_texture_object *tObj, GLuint n,
                  const GLfloat texcoords[][4], const GLfloat lambda[],
                  GLfloat rgba[][4])
{
   GLuint i;
   (void) ctx;
   (void) tObj;
   (void) texcoords;
   (void) lambda;
   for (i = 0; i < n; i++) {
      rgba[i][RCOMP] = 0;
      rgba[i][GCOMP] = 0;
      rgba[i][BCOMP] = 0;
      rgba[i][ACOMP] = 1.0;
   }
}


/**
 * Choose the texture sampling function for the given texture object.
 */
texture_sample_func
_swrast_choose_texture_sample_func( struct gl_context *ctx,
                                    const struct gl_texture_object *t )
{
   if (!t || !t->_Complete) {
      return &null_sample_func;
   }
   else {
      const GLboolean needLambda =
         (GLboolean) (t->Sampler.MinFilter != t->Sampler.MagFilter);

      switch (t->Target) {
      case GL_TEXTURE_1D:
         if (needLambda) {
            return &sample_lambda_1d;
         }
         else if (t->Sampler.MinFilter == GL_LINEAR) {
            return &sample_linear_1d;
         }
         else {
            ASSERT(t->Sampler.MinFilter == GL_NEAREST);
            return &sample_nearest_1d;
         }
      case GL_TEXTURE_2D:
         if (needLambda) {
            /* Anisotropic filtering extension. Activated only if mipmaps are used */
            if (t->Sampler.MaxAnisotropy > 1.0 &&
                t->Sampler.MinFilter == GL_LINEAR_MIPMAP_LINEAR) {
               return &sample_lambda_2d_aniso;
            }
            return &sample_lambda_2d;
         }
         else if (t->Sampler.MinFilter == GL_LINEAR) {
            return &sample_linear_2d;
         }
         else {
            /* check for a few optimized cases */
            const struct gl_texture_image *img = t->Image[0][t->BaseLevel];
            const struct swrast_texture_image *swImg =
               swrast_texture_image_const(img);
            texture_sample_func func;

            ASSERT(t->Sampler.MinFilter == GL_NEAREST);
            func = &sample_nearest_2d;
            if (t->Sampler.WrapS == GL_REPEAT &&
                t->Sampler.WrapT == GL_REPEAT &&
                swImg->_IsPowerOfTwo &&
                img->Border == 0) {
               if (img->TexFormat == MESA_FORMAT_RGB888)
                  func = &opt_sample_rgb_2d;
               else if (img->TexFormat == MESA_FORMAT_RGBA8888)
                  func = &opt_sample_rgba_2d;
            }

            return func;
         }
      case GL_TEXTURE_3D:
         if (needLambda) {
            return &sample_lambda_3d;
         }
         else if (t->Sampler.MinFilter == GL_LINEAR) {
            return &sample_linear_3d;
         }
         else {
            ASSERT(t->Sampler.MinFilter == GL_NEAREST);
            return &sample_nearest_3d;
         }
      case GL_TEXTURE_CUBE_MAP:
         if (needLambda) {
            return &sample_lambda_cube;
         }
         else if (t->Sampler.MinFilter == GL_LINEAR) {
            return &sample_linear_cube;
         }
         else {
            ASSERT(t->Sampler.MinFilter == GL_NEAREST);
            return &sample_nearest_cube;
         }
      default:
         _mesa_problem(ctx,
                       "invalid target in _swrast_choose_texture_sample_func");
         return &null_sample_func;
      }
   }
}