add verbatim

verbatim/cube_rotate.vs.asm

@@ -0,0 +1,67 @@
+-- CONST[0] = {0.159155, 0.5, 6.283185, -3.141593}
+-- CONST[1] = {theta1, theta2, 0.2, 0.5}
+
+-- each instruction is only allowed to use a single unique `const`
+-- address
+--
+-- instructions may use multiple `temp` addresses, so const[1] is
+-- moved to temp[0]:
+--
+temp[0].xy   = VE_ADD  const[1].xy__ const[1].00__ ;
+
+-- ME_SIN and ME_COS clamp their inputs to [-pi,+pi] prior to the
+-- sin/cos calculation.
+--
+-- This 3-instruction sequence remaps the range
+-- [-inf,+inf] to [-pi,+pi]
+temp[0].xy   = VE_MAD   temp[0].xy__   const[0].xx__  const[0].yy__ ;
+temp[0].xy   = VE_FRC   temp[0].xy__ ;
+temp[0].xy   = VE_MAD   temp[0].xy__   const[0].zz__  const[0].ww__ ;
+
+-- sin and cos
+temp[3].x    = ME_SIN   temp[0].___x ;
+temp[3].y    = ME_COS   temp[0].___x ;
+temp[3].z    = ME_SIN   temp[0].___y ;
+temp[3].w    = ME_COS   temp[0].___y ;
+
+-- temp[3] now contains:
+-- temp[3] = {sin(theta1), cos(theta1), sin(theta2), cos(theta2)}
+
+----------------------------------------------------------------------
+-- first rotation: X-axis rotation:
+----------------------------------------------------------------------
+
+-- y_ = (-z0 * st1)
+-- z_ = ( z0 * ct1)
+temp[1].yz    = VE_MUL   input[0]._-zz_  temp[3]._xy_ ;
+
+-- x1 = (x0 *   1 +      0)
+-- y1 = (y0 * ct1 + nz0st1)
+-- z1 = (y0 * st1 +  z0ct1)
+temp[1].xyz   = VE_MAD   input[0].xyy_   temp[3].1yx_   temp[1].0yz_ ;
+
+----------------------------------------------------------------------
+-- second rotation: Y-axis rotation:
+----------------------------------------------------------------------
+
+-- x_ = (-z1 * st2)
+-- z_ = ( z1 * ct2)
+temp[2].xz    = VE_MUL   temp[1].-z_z_   temp[3].z_w_ ;
+
+-- x2 = (x1 * ct2 + nz1st2)
+-- y2 = (y1 *   1 +      0)
+-- z2 = (x1 * st2 +  z1ct2)
+temp[2].xyz   = VE_MAD   temp[1].xyx_    temp[3].w1z_   temp[2].x0z_ ;
+
+----------------------------------------------------------------------
+-- scale
+----------------------------------------------------------------------
+
+temp[3].xyz   = VE_MAD   temp[2].xyz_    const[1].zzz_  const[1].00w_ ;
+
+----------------------------------------------------------------------
+-- output
+----------------------------------------------------------------------
+
+out[0].xyzw   = VE_MUL   temp[3].xyzz    temp[3].11-z1 ;
+out[1].xyzw   = VE_ADD   input[1].xyzw   input[1].0000 ;

verbatim/cube_rotate.vs.glsl

@@ -0,0 +1,39 @@
+#version 120
+
+attribute vec3 pos;
+attribute vec2 tex;
+
+uniform float theta1;
+uniform float theta2;
+
+vec3 rotate(vec3 v)
+{
+  float ct1 = cos(theta1);
+  float st1 = sin(theta1);
+
+  float ct2 = cos(theta2);
+  float st2 = sin(theta2);
+
+  float x0 = v.x;
+  float y0 = v.y;
+  float z0 = v.z;
+
+  float x1 = x0;
+  float y1 = y0 * ct1 - z0 * st1;
+  float z1 = y0 * st1 + z0 * ct1;
+
+  float x2 = x1 * ct2 - z1 * st2;
+  float y2 = y1;
+  float z2 = x1 * st2 + z1 * ct2;
+
+  return vec3(x2 * 0.2,
+              y2 * 0.2,
+              z2 * 0.2 + 0.5);
+}
+
+void main()
+{
+  vec3 rpos = rotate(pos);
+  gl_Position = vec4(rpos.xy, -rpos.z * rpos.z, rpos.z);
+  gl_TexCoord[0] = vec4(tex, 0, 0);
+}

verbatim/cube_rotate_3_temp.vs.asm

@@ -0,0 +1 @@
+temp[2].xyz   = VE_MAD   temp[1].xyx_    temp[3].w1z_   temp[2].x0z_ ;

verbatim/fragment_shader_equivalent_single_color.glsl

@@ -0,0 +1 @@
+colorbuffer_A = max(vec4(1, 1, 0, 0), vec4(1, 1, 0, 0));

verbatim/fragment_shader_equivalent_vertex_color.glsl

@@ -0,0 +1,2 @@
+src0.rgb = temp[0].rgb
+colorbuffer_A = max(vec4(src0.rgb, 1), vec4(src0.rgb, 1))

verbatim/mesa_cube_rotate.vs.asm

@@ -0,0 +1,27 @@
+temp[0].xyzw = VE_ADD   const[2].xyzw  const[2].0000  const[2].0000 ;
+temp[1].xyzw = VE_ADD   const[2].xyzw  const[2].0000  const[2].0000 ;
+temp[0].x    = VE_MAD   const[0].x___  temp[1].x___   temp[0].y___  ;
+temp[0].x    = VE_FRC   temp[0].x___   temp[0].0000   temp[0].0000  ;
+temp[0].x    = VE_MAD   temp[0].x___   const[2].z___  const[2].w___ ;
+temp[0].y    = ME_COS   temp[0].xxxx   temp[0].0000   temp[0].0000  ;
+temp[0].x    = ME_SIN   temp[0].xxxx   temp[0].0000   temp[0].0000  ;
+temp[1].xyzw = VE_ADD   const[2].xyzw  const[2].0000  const[2].0000 ;
+temp[2].xyzw = VE_ADD   const[2].xyzw  const[2].0000  const[2].0000 ;
+temp[0].z    = VE_MAD   const[1].__x_  temp[2].__x_   temp[1].__y_  ;
+temp[0].z    = VE_FRC   temp[0].__z_   temp[0].0000   temp[0].0000  ;
+temp[0].z    = VE_MAD   temp[0].__z_   const[2].__z_  const[2].__w_ ;
+temp[0].w    = ME_COS   temp[0].zzzz   temp[0].0000   temp[0].0000  ;
+temp[0].z    = ME_SIN   temp[0].zzzz   temp[0].0000   temp[0].0000  ;
+temp[1].xy   = VE_MUL   input[0].yz__  temp[0].xx__   temp[0].0000  ;
+temp[0].x    = VE_MAD   input[0].z___  temp[0].y___   temp[1].x___  ;
+temp[1].z    = VE_MUL   temp[0].__x_   temp[0].__z_   temp[0].0000  ;
+temp[1].z    = VE_MAD   input[0].__x_  temp[0].__w_   temp[1].__-z_ ;
+temp[0].y    = VE_MAD   input[0]._y__  temp[0]._y__   temp[1]._-y__ ;
+temp[0].x    = VE_MUL   temp[0].x___   temp[0].w___   temp[0].0000  ;
+temp[0].x    = VE_MAD   input[0].x___  temp[0].z___   temp[0].x___  ;
+temp[2].xyzw = VE_ADD   const[2].xyzw  const[2].0000  const[2].0000 ;
+temp[0].x    = VE_MAD   temp[0].x___   const[3].x___  temp[2].y___  ;
+temp[0].z    = VE_MUL   temp[0].__x_   temp[0].__x_   temp[0].0000  ;
+out[0].xz    = VE_MAD   temp[1].z_1_   const[3].x_0_  temp[0].0_-z_ ;
+out[1].xy    = VE_ADD   input[1].xy__  input[1].0000  input[1].0000 ;
+out[0].yw    = VE_MAD   temp[0]._y_1   const[3]._x_0  temp[0]._0_x  ;

verbatim/mesa_glclear.fs.asm

@@ -0,0 +1,4 @@
+OUT TEX_SEM_WAIT
+src0.a = const[0] :
+  out[0].a    = MAX src0.a src0.a ,
+  out[0].rgb  = MAX src0.rgb src0.rgb ;

verbatim/mesa_glclear.vs.asm

@@ -0,0 +1,2 @@
+out[0].xyzw  = VE_ADD  input[0].xyzw  input[0].0000 ;
+out[1].xyzw  = VE_ADD  input[1].xyzw  input[1].0000 ;

verbatim/shadertoy_circle.fs.asm

@@ -0,0 +1,21 @@
+-- d = length(uv)
+src0.rgb = temp[0] :
+  temp[0].r    = DP3 src0.rg0 src0.rg0 ;
+src0.rgb = temp[0] :
+  temp[0].a    = RSQ |src0.r| ;
+src0.a   = temp[0] :
+  temp[0].a    = RCP src0.a ;
+
+-- d = abs(d - 0.5) * 1 + -0.1
+src0.a = float(48), -- 0.5
+src1.a = temp[0],   -- d
+src2.a = float(29), -- 0.1015625
+srcp.a = sub :      -- (src1.a - src0.a)
+  temp[0].a    = MAD |srcp.a| src0.1 -src2.a ;
+
+-- d = (d >= 0.0) ? 1.0 : 0.0
+-- out.rgba = vec4(d, 0, 0, 1)
+OUT TEX_SEM_WAIT
+src0.a = temp[0] :
+  out[0].a    = MAX src0.1 src0.1 ,
+  out[0].rgb  = CMP src0.100 src0.000 src0.a00 ;

verbatim/shadertoy_palette.fs.asm

@@ -0,0 +1,51 @@
+-- CONST[0] = { time, 0, 0, 0 }
+-- CONST[1] = { PI_2, I_PI_2, 0, 0 },
+
+-- d = length(uv)
+src0.rgb = temp[0] :
+  temp[0].r    = DP3 src0.rg0 src0.rg0 ;
+src0.rgb = temp[0] :
+  temp[0].a    = RSQ |src0.r| ;
+src0.a   = temp[0] :
+  temp[0].a    = RCP src0.a ;
+
+-- i = vec3(0.25, 0.40625, 0.5625);
+-- td = time + d
+src0.a = float(40)   , -- 0.25
+src1.a = float(45)   , -- 0.40625
+src2.rgb = float(49) , -- 0.5625
+src0.rgb = const[0]  ,
+src2.a = temp[0] :
+  temp[0].rgb = MAD src0.a10 src1.1a0 src2.00r ,
+  temp[1].a   = MAD src0.1 src0.r src2.a ;
+
+-- t = i + vec3(td)
+src0.a   = temp[1] ,
+src0.rgb = temp[0] :
+  temp[0].rgb = MAD src0.111 src0.rgb src0.aaa ;
+
+-- j = fract(t)
+src0.rgb = temp[0] :
+  temp[0].rgb = FRC src0.rgb ;
+
+-- k = cos(j * 2pi)
+src0.rgb = temp[0] :
+              COS src0.r ,
+  temp[0].r = SOP ;
+src0.rgb = temp[0] :
+              COS src0.g ,
+  temp[0].g = SOP ;
+src0.rgb = temp[0] :
+              COS src0.b ,
+  temp[0].b = SOP ;
+
+-- l = k * vec3(0.5, 0.5, 0.5) + vec3(0.5, 0.5, 0.5)
+src0.rgb = temp[0] ,
+src1.rgb = float(48) : -- 0.5
+  temp[0].rgb = MAD src0.rgb src1.rrr src1.rrr ;
+
+OUT TEX_SEM_WAIT
+src0.a   = temp[0] ,
+src0.rgb = temp[0] :
+  out[0].a     = MAX src0.1 src0.1 ,
+  out[0].rgb   = MAX src0.rgb src0.rgb ;

verbatim/texture_dual.fs.asm

@@ -0,0 +1,17 @@
+TEX
+temp[1].rgba = LD tex[0] temp[0].rgaa ;
+
+TEX TEX_SEM_ACQUIRE
+temp[2].rgba = LD tex[1] temp[0].rgaa ;
+
+-- (temp[0].rrrr * (temp[2].rgba - temp[1].rgba)) + temp[1].rgba
+OUT TEX_SEM_WAIT
+src0.a = temp[1],
+src1.a = temp[2],
+srcp.a = sub,
+src0.rgb = temp[1],
+src1.rgb = temp[2],
+src2.rgb = temp[0],
+srcp.rgb = sub :
+  out[0].a    = MAD src2.r srcp.a src0.a ,
+  out[0].rgb  = MAD src2.rrr srcp.rgb src0.rgb ;

verbatim/texture_dual.fs.glsl

@@ -0,0 +1,3 @@
+vec4 c1 = texture2D(texture[0], gl_TexCoord[0].xy);
+vec4 c2 = texture2D(texture[1], gl_TexCoord[0].xy);
+gl_FragColor = mix(c1, c2, gl_TexCoord[0].x);

verbatim/vap_prog_stream_vertices.c

@@ -0,0 +1,6 @@
+const float vertices[] = {
+  // position          // color
+   0.5f, -0.5f, 0.0f,  1.0f, 0.0f, 0.0f,  // bottom right
+  -0.5f, -0.5f, 0.0f,  0.0f, 1.0f, 0.0f,  // bottom left
+   0.0f,  0.5f, 0.0f,  0.0f, 0.0f, 1.0f   // top
+};

verbatim/vap_prog_stream_vertices2.c

@@ -0,0 +1,10 @@
+const float vertices[] = {
+  // position
+   0.5f, -0.5f, 0.0f,  // bottom right
+  -0.5f, -0.5f, 0.0f,  // bottom left
+   0.0f,  0.5f, 0.0f,  // top
+  // color
+   1.0f,  0.0f, 0.0f,  // bottom right
+   0.0f,  1.0f, 0.0f,  // bottom left
+   0.0f,  0.0f, 1.0f   // top
+};

verbatim/vertex_shader_equivalent_single_color.glsl

@@ -0,0 +1 @@
+out[0] = input[0].xyzw + vec4(0, 0, 0, 0);

verbatim/vertex_shader_equivalent_vertex_color.glsl

@@ -0,0 +1,2 @@
+out[1] = input[1].xyz + vec3(0, 0, 0);
+out[0] = vec4(input[0].xyz, 1) + vec4(0, 0, 0, 0);