예제 #1
0
static void *
create_frag_shader_weave(struct vl_compositor *c)
{
   struct ureg_program *shader;
   struct ureg_src i_tc[2];
   struct ureg_src csc[3];
   struct ureg_src sampler[3];
   struct ureg_dst t_tc[2];
   struct ureg_dst t_texel[2];
   struct ureg_dst o_fragment;
   unsigned i, j;

   shader = ureg_create(TGSI_PROCESSOR_FRAGMENT);
   if (!shader)
      return false;

   i_tc[0] = ureg_DECL_fs_input(shader, TGSI_SEMANTIC_GENERIC, VS_O_VTOP, TGSI_INTERPOLATE_LINEAR);
   i_tc[1] = ureg_DECL_fs_input(shader, TGSI_SEMANTIC_GENERIC, VS_O_VBOTTOM, TGSI_INTERPOLATE_LINEAR);

   for (i = 0; i < 3; ++i) {
      csc[i] = ureg_DECL_constant(shader, i);
      sampler[i] = ureg_DECL_sampler(shader, i);
   }

   for (i = 0; i < 2; ++i) {
      t_tc[i] = ureg_DECL_temporary(shader);
      t_texel[i] = ureg_DECL_temporary(shader);
   }
   o_fragment = ureg_DECL_output(shader, TGSI_SEMANTIC_COLOR, 0);

   /* calculate the texture offsets
    * t_tc.x = i_tc.x
    * t_tc.y = (round(i_tc.y) + 0.5) / height * 2
    */
   for (i = 0; i < 2; ++i) {
      ureg_MOV(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_X), i_tc[i]);
      ureg_ROUND(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_YZ), i_tc[i]);
      ureg_MOV(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_W),
               ureg_imm1f(shader, i ? 0.75f : 0.25f));
      ureg_ADD(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_YZ),
               ureg_src(t_tc[i]), ureg_imm1f(shader, 0.5f));
      ureg_MUL(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_Y),
               ureg_src(t_tc[i]), ureg_scalar(i_tc[0], TGSI_SWIZZLE_W));
      ureg_MUL(shader, ureg_writemask(t_tc[i], TGSI_WRITEMASK_Z),
               ureg_src(t_tc[i]), ureg_scalar(i_tc[1], TGSI_SWIZZLE_W));
   }

   /* fetch the texels
    * texel[0..1].x = tex(t_tc[0..1][0])
    * texel[0..1].y = tex(t_tc[0..1][1])
    * texel[0..1].z = tex(t_tc[0..1][2])
    */
   for (i = 0; i < 2; ++i)
      for (j = 0; j < 3; ++j) {
         struct ureg_src src = ureg_swizzle(ureg_src(t_tc[i]),
            TGSI_SWIZZLE_X, j ? TGSI_SWIZZLE_Z : TGSI_SWIZZLE_Y, TGSI_SWIZZLE_W, TGSI_SWIZZLE_W);

         ureg_TEX(shader, ureg_writemask(t_texel[i], TGSI_WRITEMASK_X << j),
                  TGSI_TEXTURE_3D, src, sampler[j]);
      }

   /* calculate linear interpolation factor
    * factor = |round(i_tc.y) - i_tc.y| * 2
    */
   ureg_ROUND(shader, ureg_writemask(t_tc[0], TGSI_WRITEMASK_YZ), i_tc[0]);
   ureg_ADD(shader, ureg_writemask(t_tc[0], TGSI_WRITEMASK_YZ),
            ureg_src(t_tc[0]), ureg_negate(i_tc[0]));
   ureg_MUL(shader, ureg_writemask(t_tc[0], TGSI_WRITEMASK_XY),
            ureg_abs(ureg_src(t_tc[0])), ureg_imm1f(shader, 2.0f));
   ureg_LRP(shader, t_texel[0], ureg_swizzle(ureg_src(t_tc[0]),
            TGSI_SWIZZLE_Y, TGSI_SWIZZLE_Z, TGSI_SWIZZLE_Z, TGSI_SWIZZLE_Z),
            ureg_src(t_texel[1]), ureg_src(t_texel[0]));

   /* and finally do colour space transformation
    * fragment = csc * texel
    */
   ureg_MOV(shader, ureg_writemask(t_texel[0], TGSI_WRITEMASK_W), ureg_imm1f(shader, 1.0f));
   for (i = 0; i < 3; ++i)
      ureg_DP4(shader, ureg_writemask(o_fragment, TGSI_WRITEMASK_X << i), csc[i], ureg_src(t_texel[0]));

   ureg_MOV(shader, ureg_writemask(o_fragment, TGSI_WRITEMASK_W), ureg_imm1f(shader, 1.0f));

   for (i = 0; i < 2; ++i) {
      ureg_release_temporary(shader, t_texel[i]);
      ureg_release_temporary(shader, t_tc[i]);
   }

   ureg_END(shader);

   return ureg_create_shader_and_destroy(shader, c->pipe);
}
예제 #2
0
static void *
create_deint_frag_shader(struct vl_deint_filter *filter, unsigned field,
                         struct vertex2f *sizes, bool spatial_filter)
{
   struct ureg_program *shader;
   struct ureg_src i_vtex;
   struct ureg_src sampler_cur;
   struct ureg_src sampler_prevprev;
   struct ureg_src sampler_prev;
   struct ureg_src sampler_next;
   struct ureg_dst o_fragment;
   struct ureg_dst t_tex;
   struct ureg_dst t_comp_top, t_comp_bot;
   struct ureg_dst t_diff;
   struct ureg_dst t_a, t_b;
   struct ureg_dst t_weave, t_linear;

   shader = ureg_create(PIPE_SHADER_FRAGMENT);
   if (!shader) {
      return NULL;
   }

   t_tex = ureg_DECL_temporary(shader);
   t_comp_top = ureg_DECL_temporary(shader);
   t_comp_bot = ureg_DECL_temporary(shader);
   t_diff = ureg_DECL_temporary(shader);
   t_a = ureg_DECL_temporary(shader);
   t_b = ureg_DECL_temporary(shader);
   t_weave = ureg_DECL_temporary(shader);
   t_linear = ureg_DECL_temporary(shader);

   i_vtex = ureg_DECL_fs_input(shader, TGSI_SEMANTIC_GENERIC, VS_O_VTEX, TGSI_INTERPOLATE_LINEAR);
   sampler_prevprev = ureg_DECL_sampler(shader, 0);
   sampler_prev = ureg_DECL_sampler(shader, 1);
   sampler_cur = ureg_DECL_sampler(shader, 2);
   sampler_next = ureg_DECL_sampler(shader, 3);
   o_fragment = ureg_DECL_output(shader, TGSI_SEMANTIC_COLOR, 0);

   // we don't care about ZW interpolation (allows better optimization)
   ureg_MOV(shader, t_tex, i_vtex);
   ureg_MOV(shader, ureg_writemask(t_tex, TGSI_WRITEMASK_ZW),
            ureg_imm1f(shader, 0));

   // sample between texels for cheap lowpass
   ureg_ADD(shader, t_comp_top, ureg_src(t_tex),
            ureg_imm4f(shader, sizes->x * 0.5f, sizes->y * -0.5f, 0, 0));
   ureg_ADD(shader, t_comp_bot, ureg_src(t_tex),
            ureg_imm4f(shader, sizes->x * -0.5f, sizes->y * 0.5f, 1.0f, 0));

   if (field == 0) {
      /* interpolating top field -> current field is a bottom field */
      // cur vs prev2
      ureg_TEX(shader, t_a, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_cur);
      ureg_TEX(shader, t_b, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_prevprev);
      ureg_ADD(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_X), ureg_src(t_a), ureg_negate(ureg_src(t_b)));
      // prev vs next
      ureg_TEX(shader, t_a, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_top), sampler_prev);
      ureg_TEX(shader, t_b, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_top), sampler_next);
      ureg_ADD(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_Y), ureg_src(t_a), ureg_negate(ureg_src(t_b)));
   } else {
      /* interpolating bottom field -> current field is a top field */
      // cur vs prev2
      ureg_TEX(shader, t_a, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_top), sampler_cur);
      ureg_TEX(shader, t_b, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_top), sampler_prevprev);
      ureg_ADD(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_X), ureg_src(t_a), ureg_negate(ureg_src(t_b)));
      // prev vs next
      ureg_TEX(shader, t_a, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_prev);
      ureg_TEX(shader, t_b, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_next);
      ureg_ADD(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_Y), ureg_src(t_a), ureg_negate(ureg_src(t_b)));
   }

   // absolute maximum of differences
   ureg_MAX(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_X), ureg_abs(ureg_src(t_diff)),
            ureg_scalar(ureg_abs(ureg_src(t_diff)), TGSI_SWIZZLE_Y));

   if (field == 0) {
      /* weave with prev top field */
      ureg_TEX(shader, t_weave, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_tex), sampler_prev);
      /* get linear interpolation from current bottom field */
      ureg_ADD(shader, t_comp_top, ureg_src(t_tex), ureg_imm4f(shader, 0, sizes->y * -1.0f, 1.0f, 0));
      ureg_TEX(shader, t_linear, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_top), sampler_cur);
   } else {
      /* weave with prev bottom field */
      ureg_ADD(shader, t_comp_bot, ureg_src(t_tex), ureg_imm4f(shader, 0, 0, 1.0f, 0));
      ureg_TEX(shader, t_weave, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_prev);
      /* get linear interpolation from current top field */
      ureg_ADD(shader, t_comp_bot, ureg_src(t_tex), ureg_imm4f(shader, 0, sizes->y * 1.0f, 0, 0));
      ureg_TEX(shader, t_linear, TGSI_TEXTURE_2D_ARRAY, ureg_src(t_comp_bot), sampler_cur);
   }

   // mix between weave and linear
   // fully weave if diff < 6 (0.02353), fully interpolate if diff > 14 (0.05490)
   ureg_ADD(shader, ureg_writemask(t_diff, TGSI_WRITEMASK_X), ureg_src(t_diff),
            ureg_imm4f(shader, -0.02353f, 0, 0, 0));
   ureg_MUL(shader, ureg_saturate(ureg_writemask(t_diff, TGSI_WRITEMASK_X)),
            ureg_src(t_diff), ureg_imm4f(shader, 31.8750f, 0, 0, 0));
   ureg_LRP(shader, ureg_writemask(t_tex, TGSI_WRITEMASK_X), ureg_src(t_diff),
            ureg_src(t_linear), ureg_src(t_weave));
   ureg_MOV(shader, o_fragment, ureg_scalar(ureg_src(t_tex), TGSI_SWIZZLE_X));

   ureg_release_temporary(shader, t_tex);
   ureg_release_temporary(shader, t_comp_top);
   ureg_release_temporary(shader, t_comp_bot);
   ureg_release_temporary(shader, t_diff);
   ureg_release_temporary(shader, t_a);
   ureg_release_temporary(shader, t_b);
   ureg_release_temporary(shader, t_weave);
   ureg_release_temporary(shader, t_linear);
   ureg_END(shader);

   return ureg_create_shader_and_destroy(shader, filter->pipe);
}
예제 #3
0
void *
util_make_fs_msaa_resolve_bilinear(struct pipe_context *pipe,
                                   enum tgsi_texture_type tgsi_tex,
                                   unsigned nr_samples,
                                   enum tgsi_return_type stype)
{
   struct ureg_program *ureg;
   struct ureg_src sampler, coord;
   struct ureg_dst out, tmp, top, bottom;
   struct ureg_dst tmp_coord[4], tmp_sum[4];
   unsigned i, c;

   ureg = ureg_create(PIPE_SHADER_FRAGMENT);
   if (!ureg)
      return NULL;

   /* Declarations. */
   sampler = ureg_DECL_sampler(ureg, 0);
   ureg_DECL_sampler_view(ureg, 0, tgsi_tex, stype, stype, stype, stype);
   coord = ureg_DECL_fs_input(ureg, TGSI_SEMANTIC_GENERIC, 0,
                              TGSI_INTERPOLATE_LINEAR);
   out = ureg_DECL_output(ureg, TGSI_SEMANTIC_COLOR, 0);
   for (c = 0; c < 4; c++)
      tmp_sum[c] = ureg_DECL_temporary(ureg);
   for (c = 0; c < 4; c++)
      tmp_coord[c] = ureg_DECL_temporary(ureg);
   tmp = ureg_DECL_temporary(ureg);
   top = ureg_DECL_temporary(ureg);
   bottom = ureg_DECL_temporary(ureg);

   /* Instructions. */
   for (c = 0; c < 4; c++)
      ureg_MOV(ureg, tmp_sum[c], ureg_imm1f(ureg, 0));

   /* Get 4 texture coordinates for the bilinear filter. */
   ureg_F2U(ureg, tmp_coord[0], coord); /* top-left */
   ureg_UADD(ureg, tmp_coord[1], ureg_src(tmp_coord[0]),
             ureg_imm4u(ureg, 1, 0, 0, 0)); /* top-right */
   ureg_UADD(ureg, tmp_coord[2], ureg_src(tmp_coord[0]),
             ureg_imm4u(ureg, 0, 1, 0, 0)); /* bottom-left */
   ureg_UADD(ureg, tmp_coord[3], ureg_src(tmp_coord[0]),
             ureg_imm4u(ureg, 1, 1, 0, 0)); /* bottom-right */

   for (i = 0; i < nr_samples; i++) {
      for (c = 0; c < 4; c++) {
         /* Read one sample. */
         ureg_MOV(ureg, ureg_writemask(tmp_coord[c], TGSI_WRITEMASK_W),
                  ureg_imm1u(ureg, i));
         ureg_TXF(ureg, tmp, tgsi_tex, ureg_src(tmp_coord[c]), sampler);

         if (stype == TGSI_RETURN_TYPE_UINT)
            ureg_U2F(ureg, tmp, ureg_src(tmp));
         else if (stype == TGSI_RETURN_TYPE_SINT)
            ureg_I2F(ureg, tmp, ureg_src(tmp));

         /* Add it to the sum.*/
         ureg_ADD(ureg, tmp_sum[c], ureg_src(tmp_sum[c]), ureg_src(tmp));
      }
   }

   /* Calculate the average. */
   for (c = 0; c < 4; c++)
      ureg_MUL(ureg, tmp_sum[c], ureg_src(tmp_sum[c]),
               ureg_imm1f(ureg, 1.0 / nr_samples));

   /* Take the 4 average values and apply a standard bilinear filter. */
   ureg_FRC(ureg, tmp, coord);

   ureg_LRP(ureg, top,
            ureg_scalar(ureg_src(tmp), 0),
            ureg_src(tmp_sum[1]),
            ureg_src(tmp_sum[0]));

   ureg_LRP(ureg, bottom,
            ureg_scalar(ureg_src(tmp), 0),
            ureg_src(tmp_sum[3]),
            ureg_src(tmp_sum[2]));

   ureg_LRP(ureg, tmp,
            ureg_scalar(ureg_src(tmp), 1),
            ureg_src(bottom),
            ureg_src(top));

   /* Convert to the texture format and return. */
   if (stype == TGSI_RETURN_TYPE_UINT)
      ureg_F2U(ureg, out, ureg_src(tmp));
   else if (stype == TGSI_RETURN_TYPE_SINT)
      ureg_F2I(ureg, out, ureg_src(tmp));
   else
      ureg_MOV(ureg, out, ureg_src(tmp));

   ureg_END(ureg);

   return ureg_create_shader_and_destroy(ureg, pipe);
}