gl_common.c

来自「君正早期ucos系统(只有早期的才不没有打包成库),MPLAYER,文件系统,图」· C语言 代码 · 共 1,599 行 · 第 1/4 页

  ppm_skip(f);
  if (fscanf(f, "%u", &h) != 1)
    return 0;
  ppm_skip(f);
  if (fscanf(f, "%u", &m) != 1)
    return 0;
  val = fgetc(f);
  if (!isspace(val))
    return 0;
  if (w > MAXDIM || h > MAXDIM)
    return 0;
  bpp = (m > 255) ? 6 : 3;
  data = malloc(w * h * bpp);
  if (fread(data, w * bpp, h, f) != h)
    return 0;
  if (!fmt) {
    fmt = (m > 255) ? hqtexfmt : 3;
    if (fmt == GL_FLOAT_RGB32_NV && target != GL_TEXTURE_RECTANGLE)
      fmt = GL_RGB16;
  }
  glCreateClearTex(target, fmt, filter, w, h, 0);
  glUploadTex(target, GL_RGB, (m > 255) ? GL_UNSIGNED_SHORT : GL_UNSIGNED_BYTE,
              data, w * bpp, 0, 0, w, h, 0);
  free(data);
  if (width) *width = w;
  if (height) *height = h;
  if (maxval) *maxval = m;
  return 1;
}

/**
 * \brief return the number of bytes per pixel for the given format
 * \param format OpenGL format
 * \param type OpenGL type
 * \return bytes per pixel
 * \ingroup glgeneral
 *
 * Does not handle all possible variants, just those used by MPlayer
 */
int glFmt2bpp(GLenum format, GLenum type) {
  int component_size = 0;
  switch (type) {
    case GL_UNSIGNED_BYTE_3_3_2:
    case GL_UNSIGNED_BYTE_2_3_3_REV:
      return 1;
    case GL_UNSIGNED_SHORT_5_5_5_1:
    case GL_UNSIGNED_SHORT_1_5_5_5_REV:
    case GL_UNSIGNED_SHORT_5_6_5:
    case GL_UNSIGNED_SHORT_5_6_5_REV:
      return 2;
    case GL_UNSIGNED_BYTE:
      component_size = 1;
      break;
    case GL_UNSIGNED_SHORT:
      component_size = 2;
      break;
  }
  switch (format) {
    case GL_LUMINANCE:
    case GL_ALPHA:
      return component_size;
    case GL_RGB:
    case GL_BGR:
      return 3 * component_size;
    case GL_RGBA:
    case GL_BGRA:
      return 4 * component_size;
  }
  return 0; // unknown
}

/**
 * \brief upload a texture, handling things like stride and slices
 * \param target texture target, usually GL_TEXTURE_2D
 * \param format OpenGL format of data
 * \param type OpenGL type of data
 * \param dataptr data to upload
 * \param stride data stride
 * \param x x offset in texture
 * \param y y offset in texture
 * \param w width of the texture part to upload
 * \param h height of the texture part to upload
 * \param slice height of an upload slice, 0 for all at once
 * \ingroup gltexture
 */
void glUploadTex(GLenum target, GLenum format, GLenum type,
                 const void *dataptr, int stride,
                 int x, int y, int w, int h, int slice) {
  const uint8_t *data = dataptr;
  int y_max = y + h;
  if (w <= 0 || h <= 0) return;
  if (slice <= 0)
    slice = h;
  if (stride < 0) {
    data += (h - 1) * stride;
    stride = -stride;
  }
  // this is not always correct, but should work for MPlayer
  glAdjustAlignment(stride);
  glPixelStorei(GL_UNPACK_ROW_LENGTH, stride / glFmt2bpp(format, type));
  for (; y + slice <= y_max; y += slice) {
    glTexSubImage2D(target, 0, x, y, w, slice, format, type, data);
    data += stride * slice;
  }
  if (y < y_max)
    glTexSubImage2D(target, 0, x, y, w, y_max - y, format, type, data);
}

static void fillUVcoeff(GLfloat *ucoef, GLfloat *vcoef,
                        float uvcos, float uvsin) {
  int i;
  ucoef[0] = 0 * uvcos + 1.403 * uvsin;
  vcoef[0] = 0 * uvsin + 1.403 * uvcos;
  ucoef[1] = -0.344 * uvcos + -0.714 * uvsin;
  vcoef[1] = -0.344 * uvsin + -0.714 * uvcos;
  ucoef[2] = 1.770 * uvcos + 0 * uvsin;
  vcoef[2] = 1.770 * uvsin + 0 * uvcos;
  ucoef[3] = 0;
  vcoef[3] = 0;
  // Coefficients (probably) must be in [0, 1] range, whereas they originally
  // are in [-2, 2] range, so here comes the trick:
  // First put them in the [-0.5, 0.5] range, then add 0.5.
  // This can be undone with the HALF_BIAS and SCALE_BY_FOUR arguments
  // for CombinerInput and CombinerOutput (or the respective ATI variants)
  for (i = 0; i < 4; i++) {
    ucoef[i] = ucoef[i] * 0.25 + 0.5;
    vcoef[i] = vcoef[i] * 0.25 + 0.5;
  }
}

/**
 * \brief Setup register combiners for YUV to RGB conversion.
 * \param uvcos used for saturation and hue adjustment
 * \param uvsin used for saturation and hue adjustment
 */
static void glSetupYUVCombiners(float uvcos, float uvsin) {
  GLfloat ucoef[4];
  GLfloat vcoef[4];
  GLint i;
  if (!CombinerInput || !CombinerOutput ||
      !CombinerParameterfv || !CombinerParameteri) {
    mp_msg(MSGT_VO, MSGL_FATAL, "[gl] Combiner functions missing!\n");
    return;
  }
  glGetIntegerv(GL_MAX_GENERAL_COMBINERS_NV, &i);
  if (i < 2)
    mp_msg(MSGT_VO, MSGL_ERR,
           "[gl] 2 general combiners needed for YUV combiner support (found %i)\n", i);
  glGetIntegerv (GL_MAX_TEXTURE_UNITS, &i);
  if (i < 3)
    mp_msg(MSGT_VO, MSGL_ERR,
           "[gl] 3 texture units needed for YUV combiner support (found %i)\n", i);
  fillUVcoeff(ucoef, vcoef, uvcos, uvsin);
  CombinerParameterfv(GL_CONSTANT_COLOR0_NV, ucoef);
  CombinerParameterfv(GL_CONSTANT_COLOR1_NV, vcoef);

  // UV first, like this green component cannot overflow
  CombinerInput(GL_COMBINER0_NV, GL_RGB, GL_VARIABLE_A_NV,
                GL_TEXTURE1, GL_HALF_BIAS_NORMAL_NV, GL_RGB);
  CombinerInput(GL_COMBINER0_NV, GL_RGB, GL_VARIABLE_B_NV,
                GL_CONSTANT_COLOR0_NV, GL_HALF_BIAS_NORMAL_NV, GL_RGB);
  CombinerInput(GL_COMBINER0_NV, GL_RGB, GL_VARIABLE_C_NV,
                GL_TEXTURE2, GL_HALF_BIAS_NORMAL_NV, GL_RGB);
  CombinerInput(GL_COMBINER0_NV, GL_RGB, GL_VARIABLE_D_NV,
                GL_CONSTANT_COLOR1_NV, GL_HALF_BIAS_NORMAL_NV, GL_RGB);
  CombinerOutput(GL_COMBINER0_NV, GL_RGB, GL_DISCARD_NV, GL_DISCARD_NV,
                 GL_SPARE0_NV, GL_SCALE_BY_FOUR_NV, GL_NONE, GL_FALSE,
                 GL_FALSE, GL_FALSE);

  // stage 2
  CombinerInput(GL_COMBINER1_NV, GL_RGB, GL_VARIABLE_A_NV, GL_SPARE0_NV,
                GL_SIGNED_IDENTITY_NV, GL_RGB);
  CombinerInput(GL_COMBINER1_NV, GL_RGB, GL_VARIABLE_B_NV, GL_ZERO,
                 GL_UNSIGNED_INVERT_NV, GL_RGB);
  CombinerInput(GL_COMBINER1_NV, GL_RGB, GL_VARIABLE_C_NV,
                GL_TEXTURE0, GL_SIGNED_IDENTITY_NV, GL_RGB);
  CombinerInput(GL_COMBINER1_NV, GL_RGB, GL_VARIABLE_D_NV, GL_ZERO,
                GL_UNSIGNED_INVERT_NV, GL_RGB);
  CombinerOutput(GL_COMBINER1_NV, GL_RGB, GL_DISCARD_NV, GL_DISCARD_NV,
                 GL_SPARE0_NV, GL_NONE, GL_NONE, GL_FALSE,
                 GL_FALSE, GL_FALSE);

  // leave final combiner stage in default mode
  CombinerParameteri(GL_NUM_GENERAL_COMBINERS_NV, 2);
}

/**
 * \brief Setup ATI version of register combiners for YUV to RGB conversion.
 * \param uvcos used for saturation and hue adjustment
 * \param uvsin used for saturation and hue adjustment
 *
 * ATI called this fragment shader, but the name is confusing in the
 * light of a very different OpenGL 2.0 extension with the same name
 */
static void glSetupYUVCombinersATI(float uvcos, float uvsin) {
  GLfloat ucoef[4];
  GLfloat vcoef[4];
  GLint i;
  if (!BeginFragmentShader || !EndFragmentShader ||
      !SetFragmentShaderConstant || !SampleMap ||
      !ColorFragmentOp2 || !ColorFragmentOp3) {
    mp_msg(MSGT_VO, MSGL_FATAL, "[gl] Combiner (ATI) functions missing!\n");
    return;
  }
  glGetIntegerv(GL_NUM_FRAGMENT_REGISTERS_ATI, &i);
  if (i < 3)
    mp_msg(MSGT_VO, MSGL_ERR,
           "[gl] 3 registers needed for YUV combiner (ATI) support (found %i)\n", i);
  glGetIntegerv (GL_MAX_TEXTURE_UNITS, &i);
  if (i < 3)
    mp_msg(MSGT_VO, MSGL_ERR,
           "[gl] 3 texture units needed for YUV combiner (ATI) support (found %i)\n", i);
  fillUVcoeff(ucoef, vcoef, uvcos, uvsin);
  BeginFragmentShader();
  SetFragmentShaderConstant(GL_CON_0_ATI, ucoef);
  SetFragmentShaderConstant(GL_CON_1_ATI, vcoef);
  SampleMap(GL_REG_0_ATI, GL_TEXTURE0, GL_SWIZZLE_STR_ATI);
  SampleMap(GL_REG_1_ATI, GL_TEXTURE1, GL_SWIZZLE_STR_ATI);
  SampleMap(GL_REG_2_ATI, GL_TEXTURE2, GL_SWIZZLE_STR_ATI);
  // UV first, like this green component cannot overflow
  ColorFragmentOp2(GL_MUL_ATI, GL_REG_1_ATI, GL_NONE, GL_NONE,
                   GL_REG_1_ATI, GL_NONE, GL_BIAS_BIT_ATI,
                   GL_CON_0_ATI, GL_NONE, GL_BIAS_BIT_ATI);
  ColorFragmentOp3(GL_MAD_ATI, GL_REG_2_ATI, GL_NONE, GL_4X_BIT_ATI,
                   GL_REG_2_ATI, GL_NONE, GL_BIAS_BIT_ATI,
                   GL_CON_1_ATI, GL_NONE, GL_BIAS_BIT_ATI,
                   GL_REG_1_ATI, GL_NONE, GL_NONE);
  ColorFragmentOp2(GL_ADD_ATI, GL_REG_0_ATI, GL_NONE, GL_NONE,
                   GL_REG_0_ATI, GL_NONE, GL_NONE,
                   GL_REG_2_ATI, GL_NONE, GL_NONE);
  EndFragmentShader();
}

/**
 * \brief helper function for gen_spline_lookup_tex
 * \param x subpixel-position ((0,1) range) to calculate weights for
 * \param dst where to store transformed weights, must provide space for 4 GLfloats
 *
 * calculates the weights and stores them after appropriate transformation
 * for the scaler fragment program.
 */
static void store_weights(float x, GLfloat *dst) {
  float w0 = (((-1 * x + 3) * x - 3) * x + 1) / 6;
  float w1 = ((( 3 * x - 6) * x + 0) * x + 4) / 6;
  float w2 = (((-3 * x + 3) * x + 3) * x + 1) / 6;
  float w3 = ((( 1 * x + 0) * x + 0) * x + 0) / 6;
  *dst++ = 1 + x - w1 / (w0 + w1);
  *dst++ = 1 - x + w3 / (w2 + w3);
  *dst++ = w0 + w1;
  *dst++ = 0;
}

//! to avoid artefacts this should be rather large
#define LOOKUP_BSPLINE_RES (2 * 1024)
/**
 * \brief creates the 1D lookup texture needed for fast higher-order filtering
 * \param unit texture unit to attach texture to
 */
static void gen_spline_lookup_tex(GLenum unit) {
  GLfloat tex[4 * LOOKUP_BSPLINE_RES];
  GLfloat *tp = tex;
  int i;
  for (i = 0; i < LOOKUP_BSPLINE_RES; i++) {
    float x = (float)(i + 0.5) / LOOKUP_BSPLINE_RES;
    store_weights(x, tp);
    tp += 4;
  }
  store_weights(0, tex);
  store_weights(1, &tex[4 * (LOOKUP_BSPLINE_RES - 1)]);
  ActiveTexture(unit);
  glTexImage1D(GL_TEXTURE_1D, 0, GL_RGBA16, LOOKUP_BSPLINE_RES, 0, GL_RGBA, GL_FLOAT, tex);
  glTexParameterf(GL_TEXTURE_1D, GL_TEXTURE_PRIORITY, 1.0);
  glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
  glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
  glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_WRAP_S, GL_REPEAT);
  ActiveTexture(GL_TEXTURE0);
}

static const char *bilin_filt_template =
  "TEX yuv.%c, fragment.texcoord[%c], texture[%c], %s;";

#define BICUB_FILT_MAIN(textype) \
  /* first y-interpolation */ \
  "ADD coord, fragment.texcoord[%c].xyxy, cdelta.xyxw;" \
  "ADD coord2, fragment.texcoord[%c].xyxy, cdelta.zyzw;" \
  "TEX a.r, coord.xyxy, texture[%c], "textype";" \
  "TEX a.g, coord.zwzw, texture[%c], "textype";" \
  /* second y-interpolation */ \
  "TEX b.r, coord2.xyxy, texture[%c], "textype";" \
  "TEX b.g, coord2.zwzw, texture[%c], "textype";" \
  "LRP a.b, parmy.b, a.rrrr, a.gggg;" \
  "LRP a.a, parmy.b, b.rrrr, b.gggg;" \
  /* x-interpolation */ \
  "LRP yuv.%c, parmx.b, a.bbbb, a.aaaa;"

static const char *bicub_filt_template_2D =
  "MAD coord.xy, fragment.texcoord[%c], {%f, %f}, {0.5, 0.5};"
  "TEX parmx, coord.x, texture[%c], 1D;"
  "MUL cdelta.xz, parmx.rrgg, {-%f, 0, %f, 0};"
  "TEX parmy, coord.y, texture[%c], 1D;"
  "MUL cdelta.yw, parmy.rrgg, {0, -%f, 0, %f};"
  BICUB_FILT_MAIN("2D");

static const char *bicub_filt_template_RECT =
  "ADD coord, fragment.texcoord[%c], {0.5, 0.5};"
  "TEX parmx, coord.x, texture[%c], 1D;"
  "MUL cdelta.xz, parmx.rrgg, {-1, 0, 1, 0};"
  "TEX parmy, coord.y, texture[%c], 1D;"
  "MUL cdelta.yw, parmy.rrgg, {0, -1, 0, 1};"
  BICUB_FILT_MAIN("RECT");

#define CALCWEIGHTS(t, s) \
  "MAD "t", {-0.5, 0.1666, 0.3333, -0.3333}, "s", {1, 0, -0.5, 0.5};" \
  "MAD "t", "t", "s", {0, 0, -0.5, 0.5};" \
  "MAD "t", "t", "s", {-0.6666, 0, 0.8333, 0.1666};" \
  "RCP a.x, "t".z;" \
  "RCP a.y, "t".w;" \
  "MAD "t".xy, "t".xyxy, a.xyxy, {1, 1, 0, 0};" \
  "ADD "t".x, "t".xxxx, "s";" \
  "SUB "t".y, "t".yyyy, "s";"

static const char *bicub_notex_filt_template_2D =
  "MAD coord.xy, fragment.texcoord[%c], {%f, %f}, {0.5, 0.5};"
  "FRC coord.xy, coord.xyxy;"
  CALCWEIGHTS("parmx", "coord.xxxx")
  "MUL cdelta.xz, parmx.rrgg, {-%f, 0, %f, 0};"
  CALCWEIGHTS("parmy", "coord.yyyy")
  "MUL cdelta.yw, parmy.rrgg, {0, -%f, 0, %f};"
  BICUB_FILT_MAIN("2D");

static const char *bicub_notex_filt_template_RECT =
  "ADD coord, fragment.texcoord[%c], {0.5, 0.5};"
  "FRC coord.xy, coord.xyxy;"
  CALCWEIGHTS("parmx", "coord.xxxx")
  "MUL cdelta.xz, parmx.rrgg, {-1, 0, 1, 0};"
  CALCWEIGHTS("parmy", "coord.yyyy")
  "MUL cdelta.yw, parmy.rrgg, {0, -1, 0, 1};"
  BICUB_FILT_MAIN("RECT");

#define BICUB_X_FILT_MAIN(textype) \
  "ADD coord, fragment.texcoord[%c].xyxy, cdelta.xyxw;" \
  "ADD coord2, fragment.texcoord[%c].xyxy, cdelta.zyzw;" \
  "TEX a.r, coord, texture[%c], "textype";" \
  "TEX b.r, coord2, texture[%c], "textype";" \
  /* x-interpolation */ \
  "LRP yuv.%c, parmx.b, a.rrrr, b.rrrr;"

static const char *bicub_x_filt_template_2D =
  "MAD coord.x, fragment.texcoord[%c], {%f, %f}, {0.5, 0.5};"
  "TEX parmx, coord, texture[%c], 1D;"
  "MUL cdelta.xz, parmx.rrgg, {-%f, 0, %f, 0};"
  BICUB_X_FILT_MAIN("2D");

static const char *bicub_x_filt_template_RECT =
  "ADD coord.x, fragment.texcoord[%c], {0.5, 0.5};"
  "TEX parmx, coord, texture[%c], 1D;"
  "MUL cdelta.xz, parmx.rrgg, {-1, 0, 1, 0};"
  BICUB_X_FILT_MAIN("RECT");

static const char *yuv_prog_template =
  "PARAM ycoef = {%.4f, %.4f, %.4f};"
  "PARAM ucoef = {%.4f, %.4f, %.4f};"
  "PARAM vcoef = {%.4f, %.4f, %.4f};"
  "PARAM offsets = {%.4f, %.4f, %.4f};"
  "TEMP res;"
  "MAD res.rgb, yuv.rrrr, ycoef, offsets;"
  "MAD res.rgb, yuv.gggg, ucoef, res;"
  "MAD result.color.rgb, yuv.bbbb, vcoef, res;"
  "END";

static const char *yuv_pow_prog_template =
  "PARAM ycoef = {%.4f, %.4f, %.4f};"
  "PARAM ucoef = {%.4f, %.4f, %.4f};"
  "PARAM vcoef = {%.4f, %.4f, %.4f};"
  "PARAM offsets = {%.4f, %.4f, %.4f};"
  "PARAM gamma = {%.4f, %.4f, %.4f};"
  "TEMP res;"
  "MAD res.rgb, yuv.rrrr, ycoef, offsets;"
  "MAD res.rgb, yuv.gggg, ucoef, res;"
  "MAD_SAT res.rgb, yuv.bbbb, vcoef, res;"
  "POW result.color.r, res.r, gamma.r;"
  "POW result.color.g, res.g, gamma.g;"
  "POW result.color.b, res.b, gamma.b;"
  "END";

static const char *yuv_lookup_prog_template =
  "PARAM ycoef = {%.4f, %.4f, %.4f, 0};"
  "PARAM ucoef = {%.4f, %.4f, %.4f, 0};"
  "PARAM vcoef = {%.4f, %.4f, %.4f, 0};"
  "PARAM offsets = {%.4f, %.4f, %.4f, 0.125};"
  "TEMP res;"
  "MAD res, yuv.rrrr, ycoef, offsets;"
  "MAD res.rgb, yuv.gggg, ucoef, res;"
  "MAD res.rgb, yuv.bbbb, vcoef, res;"
  "TEX result.color.r, res.raaa, texture[%c], 2D;"
  "ADD res.a, res.a, 0.25;"
  "TEX result.color.g, res.gaaa, texture[%c], 2D;"
  "ADD res.a, res.a, 0.25;"
  "TEX result.color.b, res.baaa, texture[%c], 2D;"
  "END";