[...]
[traceshaders] glCreateShader(GL_VERTEX_SHADER) → shader=73
[traceshaders] glShaderSource(shader=73)
[traceshaders] ================================================================================
#version 410
#extension GL_ARB_explicit_attrib_location : require
#ifdef GL_ARB_separate_shader_objects
#extension GL_ARB_separate_shader_objects : enable
#endif
subroutine void SubroutineType();
struct vec1 {
	float x;
};
struct uvec1 {
	uint x;
};
struct ivec1 {
	int x;
};
vec4 InstrHelper;
out gl_PerVertex {
vec4 gl_Position;
float gl_PointSize;
float gl_ClipDistance[];};
layout(std140) uniform;
uniform cbuffer_0 { // $Globals 
     vec4 Const0[63];
};
uniform cbuffer_1 { // VSOffsetConstants 
     vec4 Const1[5];
};
layout(location = 0)  in  vec4 dcl_Input0;
vec4 Input0;
layout(location = 1)  in  vec4 dcl_Input1;
vec4 Input1;
layout(location = 2)  in  vec4 dcl_Input2;
vec4 Input2;
layout(location = 4)  in  vec4 dcl_Input4;
vec4 Input4;
layout(location = 5)  in  vec4 dcl_Input5;
vec4 Input5;
layout(location = 6)  in  vec4 dcl_Input6;
vec4 Input6;
layout(location = 7)  in  vec4 dcl_Input7;
vec4 Input7;
layout(location = 0)  out  vec4 VtxGeoOutput0;
#define Output0 VtxGeoOutput0
layout(location = 1)  out  vec4 VtxGeoOutput1;
#define Output1 VtxGeoOutput1
layout(location = 2)  out  vec4 VtxGeoOutput2;
#define Output2 VtxGeoOutput2
layout(location = 3)  out  vec4 VtxGeoOutput3;
#define Output3 VtxGeoOutput3
layout(location = 4)  out  vec4 VtxGeoOutput4;
#define Output4 VtxGeoOutput4
#undef Output5
#define Output5 phase0_Output5
vec4 phase0_Output5;
vec4 Temp[4];
ivec4 Temp_int[4];
uvec4 Temp_uint[4];
void main()
{
    Input0 = dcl_Input0;
    Input1 = dcl_Input1;
    Input2 = dcl_Input2;
    Input4 = dcl_Input4;
    Input5 = dcl_Input5;
    Input6 = dcl_Input6;
    Input7 = dcl_Input7;
    Temp[0].xyz = Input1.xyz * vec3(uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u)) + vec3(uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u));
    Temp[1] = Input2 * vec4(uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u)) + vec4(uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u));
    Temp[2].xyz = Temp[0].xyz * Temp[1].yzx;
    Temp[0].xyz = Temp[1].xyz * Temp[0].yzx + (-Temp[2].xyz);
    Temp[0].xyz = Temp[1].www * Temp[0].xyz;
    Temp[2].xyz = Temp[1].yzx * Temp[0].xyz;
    Temp[2].xyz = Temp[0].zxy * Temp[1].zxy + (-Temp[2].xyz);
    Temp[2].xyz = Temp[1].www * Temp[2].xyz;
    Temp[2].yw = Temp[2].yy * Const0[1].xz;
    Temp[2].xy = Const0[0].xz * Temp[2].xx + Temp[2].yw;
    Temp[2].xy = Const0[2].xz * Temp[2].zz + Temp[2].xy;
    Temp[0].zw = Temp[0].zz * Const0[1].xz;
    Temp[0].yz = Const0[0].xz * Temp[0].yy + Temp[0].zw;
    Temp[0].xy = Const0[2].xz * Temp[0].xx + Temp[0].yz;
    Temp[2].z = Temp[0].x;
    Temp[3].xy = Temp[1].yy * Const0[1].xz;
    Temp[1].xy = Const0[0].xz * Temp[1].xx + Temp[3].xy;
    Temp[0].xw = Const0[2].xz * Temp[1].zz + Temp[1].xy;
    Output1.w = Temp[1].w * Const0[51].x;
    Temp[2].w = Temp[0].x;
    Temp[0].x = dot(Temp[2].xzw, Temp[2].xzw);
    Temp[0].x = ( ( Temp[0].x < 0.0 ) ? 0.0 : ( ( Temp[0].x > 0.0 ) ? inversesqrt( Temp[0].x ) : ( 3.4028235E+38 * sign( Temp[0].x ) ) ) );
    Output0.xyz = Temp[0].xxx * Temp[2].xzw;
    Temp[0].z = Temp[2].y;
    Temp[0].x = dot(Temp[0].yzw, Temp[0].yzw);
    Temp[0].x = ( ( Temp[0].x < 0.0 ) ? 0.0 : ( ( Temp[0].x > 0.0 ) ? inversesqrt( Temp[0].x ) : ( 3.4028235E+38 * sign( Temp[0].x ) ) ) );
    Output1.xyz = Temp[0].xxx * Temp[0].zyw;
    Output2.xy = Input7.xy * Const0[52].xy + Const0[52].wz;
    Output3.xy = Input4.xy;
    Output3.zw = Input5.yx;
    Output4.xy = Input6.xy;
    Output4.zw = vec2(uintBitsToFloat(0u), uintBitsToFloat(0u));
    Temp[0].x = Const0[18].w * uintBitsToFloat(1067114824u) + Const0[46].w;
    Temp[1].xyz = Input0.xyz * Const0[57].xyz + Const0[56].xyz;
    Temp[0].y = Const0[49].w * Temp[1].x + Temp[0].x;
    Temp[0].y = Temp[0].y * uintBitsToFloat(1065353216u);
    Temp[0].z = uintBitsToFloat((Temp[0].y>=(-Temp[0].y)) ? 0xFFFFFFFFu : 0u);
    Temp[0].y = fract(abs(Temp[0].y));
    Temp[0].y = (floatBitsToInt(Temp[0]).z != 0) ? Temp[0].y : (-Temp[0].y);
    Temp[0].y = Temp[0].y * uintBitsToFloat(1086918619u);
    Temp[0].z = Const0[50].x * Temp[1].x + Temp[0].x;
    Temp[0].z = Temp[0].z * uintBitsToFloat(1056964608u);
    Temp[0].w = uintBitsToFloat((Temp[0].z>=(-Temp[0].z)) ? 0xFFFFFFFFu : 0u);
    Temp[0].z = fract(abs(Temp[0].z));
    Temp[0].z = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].z = Temp[0].z * uintBitsToFloat(1086918619u);
    Temp[0].yz = sin(Temp[0].yz);
    Temp[0].y = Temp[0].z + Temp[0].y;
    Temp[0].y = Temp[0].y * Const0[50].y;
    Temp[0].zw = Temp[0].xx * vec2(uintBitsToFloat(1065353216u), uintBitsToFloat(1061481552u));
    Temp[0].x = Temp[0].x + Const0[48].x;
    Temp[2].xy = uintBitsToFloat(uvec2(greaterThanEqual(Temp[0].zwzz, (-Temp[0].zwzz)).xy) * 0xFFFFFFFFu);
    Temp[0].zw = fract(abs(Temp[0].zw));
    Temp[0].z = (floatBitsToInt(Temp[2]).x != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].w = (floatBitsToInt(Temp[2]).y != 0) ? Temp[0].w : (-Temp[0].w);
    Temp[0].zw = Temp[0].zw * vec2(uintBitsToFloat(1086918619u), uintBitsToFloat(1086918619u));
    Temp[0].zw = sin(Temp[0].zw);
    Temp[0].z = Temp[0].w + Temp[0].z;
    Temp[0].z = Temp[0].z + uintBitsToFloat(1065353216u);
    Temp[0].w = Temp[1].x * Const0[49].z;
    Temp[0].z = Temp[0].z * Temp[0].w;
    Temp[2].z = Temp[0].z * uintBitsToFloat(1045220557u) + Temp[0].y;
    Temp[0].y = Const0[46].z * Temp[1].x + (-Temp[0].x);
    Temp[0].xz = Const0[48].yw * Temp[1].xx + (-Temp[0].xx);
    Temp[0].xyz = Temp[0].xyz * vec3(uintBitsToFloat(1045494470u), uintBitsToFloat(1055439406u), uintBitsToFloat(1065353216u));
    Temp[0].w = uintBitsToFloat((Temp[0].y>=(-Temp[0].y)) ? 0xFFFFFFFFu : 0u);
    Temp[0].y = fract(abs(Temp[0].y));
    Temp[0].y = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].y : (-Temp[0].y);
    Temp[0].y = Temp[0].y * uintBitsToFloat(1086918619u);
    Temp[3].xy = uintBitsToFloat(uvec2(greaterThanEqual(Temp[0].xzxx, (-Temp[0].xzxx)).xy) * 0xFFFFFFFFu);
    Temp[0].xz = fract(abs(Temp[0].xz));
    Temp[0].x = (floatBitsToInt(Temp[3]).x != 0) ? Temp[0].x : (-Temp[0].x);
    Temp[0].z = (floatBitsToInt(Temp[3]).y != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].xz = Temp[0].xz * vec2(uintBitsToFloat(1086918619u), uintBitsToFloat(1086918619u));
    Temp[0].xyz = sin(Temp[0].xyz);
    Temp[0].x = Temp[0].x + Temp[0].y;
    Temp[0].y = Temp[0].x * Const0[49].y;
    Temp[0].x = abs(Temp[0].x) * Const0[48].z;
    Temp[2].y = Const0[49].x * Temp[0].z + Temp[0].y;
    Temp[0].y = Const0[46].y * uintBitsToFloat(3256877056u) + uintBitsToFloat(3240099840u);
    Temp[2].x = Temp[0].x * Temp[0].y;
    Temp[0].xyz = Temp[2].xyz * Input6.xxx;
    Temp[1].w = Input0.w;
    Temp[0].w = uintBitsToFloat(0u);
    Temp[0] = Temp[0] + Temp[1];
    Temp[1] = Temp[0].yyyy * Const0[1];
    Temp[1] = Const0[0] * Temp[0].xxxx + Temp[1];
    Temp[1] = Const0[2] * Temp[0].zzzz + Temp[1];
    Temp[0] = Const0[3] * Temp[0].wwww + Temp[1];
    Temp[1] = Temp[0].yyyy * Const1[1];
    Temp[1] = Const1[0] * Temp[0].xxxx + Temp[1];
    Temp[1] = Const1[2] * Temp[0].zzzz + Temp[1];
    Output5 = Const1[3] * Temp[0].wwww + Temp[1];
    gl_Position = vec4(phase0_Output5);
    gl_Position.y = -gl_Position.y;
    gl_Position.z = gl_Position.z * 2.0 - gl_Position.w;
    return;
}

[traceshaders] ================================================================================
[traceshaders] glCompileShader(shader=73)
[traceshaders] glCreateProgram() = 74
[traceshaders] glAttachShader(program=74, shader=73)
[traceshaders] glLinkProgram(program=74)
SHADER KEY
  instance_divisors = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
  as_es = 0
  as_ls = 0
  export_prim_id = 0
VERT
PROPERTY NEXT_SHADER FRAG
DCL IN[0]
DCL IN[1]
DCL IN[2]
DCL IN[3]
DCL IN[4]
DCL IN[5]
DCL IN[6]
DCL OUT[0], POSITION
DCL OUT[1], GENERIC[0]
DCL OUT[2], GENERIC[1]
DCL OUT[3], GENERIC[2]
DCL OUT[4], GENERIC[3]
DCL OUT[5], GENERIC[4]
DCL CONST[1][0..62]
DCL CONST[2][0..4]
DCL TEMP[0..11], LOCAL
IMM[0] FLT32 {    2.0000,    -1.0000,     0.0000, 340282346638528859811704183484516925440.0000}
IMM[1] UINT32 {0, 16, 32, 816}
IMM[2] UINT32 {832, 288, 736, 912}
IMM[3] FLT32 {    1.2100,     6.2832,     0.5000,     1.0000}
IMM[4] UINT32 {896, 784, 4294967295, 800}
IMM[5] INT32 {0, 1, 0, 0}
IMM[6] FLT32 {    1.0000,     0.7692,     0.2000,   -40.0000}
IMM[7] UINT32 {768, 48, 1, 0}
IMM[8] FLT32 {    0.2041,     0.4545,     1.0000,   -10.0000}
  0: MAD TEMP[0].xyz, IN[1].xyzz, IMM[0].xxxx, IMM[0].yyyy
  1: MAD TEMP[1], IN[2], IMM[0].xxxx, IMM[0].yyyy
  2: MUL TEMP[2].xyz, TEMP[0].xyzz, TEMP[1].yzxx
  3: MAD TEMP[0].xyz, TEMP[1].xyzz, TEMP[0].yzxx, -TEMP[2].xyzz
  4: MUL TEMP[0].xyz, TEMP[1].wwww, TEMP[0].xyzz
  5: MUL TEMP[2].xyz, TEMP[1].yzxx, TEMP[0].xyzz
  6: MAD TEMP[2].xyz, TEMP[0].zxyy, TEMP[1].zxyy, -TEMP[2].xyzz
  7: MUL TEMP[2].xyz, TEMP[1].wwww, TEMP[2].xyzz
  8: MUL TEMP[3].xy, TEMP[2].yyyy, CONST[1][1].xzzz
  9: MAD TEMP[2].xy, CONST[1][0].xzzz, TEMP[2].xxxx, TEMP[3].xyyy
 10: MAD TEMP[2].xy, CONST[1][2].xzzz, TEMP[2].zzzz, TEMP[2].xyyy
 11: MUL TEMP[3].xy, TEMP[0].zzzz, CONST[1][1].xzzz
 12: MAD TEMP[3].xy, CONST[1][0].xzzz, TEMP[0].yyyy, TEMP[3].xyyy
 13: MAD TEMP[0].xy, CONST[1][2].xzzz, TEMP[0].xxxx, TEMP[3].xyyy
 14: MOV TEMP[2].z, TEMP[0].xxxx
 15: MUL TEMP[3].xy, TEMP[1].yyyy, CONST[1][1].xzzz
 16: MAD TEMP[1].xy, CONST[1][0].xzzz, TEMP[1].xxxx, TEMP[3].xyyy
 17: MAD TEMP[4].xy, CONST[1][2].xzzz, TEMP[1].zzzz, TEMP[1].xyyy
 18: MOV TEMP[0].w, TEMP[4].xxxy
 19: MUL TEMP[5].x, TEMP[1].wwww, CONST[1][51].xxxx
 20: MOV TEMP[5].w, TEMP[5].xxxx
 21: MOV TEMP[2].w, TEMP[4].xxxx
 22: DP3 TEMP[0].x, TEMP[2].xzww, TEMP[2].xzww
 23: FSLT TEMP[4].x, TEMP[0].xxxx, IMM[0].zzzz
 24: UIF TEMP[4].xxxx :0
 25:   MOV TEMP[4].x, IMM[0].zzzz
 26: ELSE :0
 27:   FSLT TEMP[6].x, IMM[0].zzzz, TEMP[0].xxxx
 28:   UIF TEMP[6].xxxx :0
 29:     RSQ TEMP[6].x, TEMP[0].xxxx
 30:   ELSE :0
 31:     SSG TEMP[7].x, TEMP[0].xxxx
 32:     MUL TEMP[6].x, IMM[0].wwww, TEMP[7].xxxx
 33:   ENDIF
 34:   MOV TEMP[4].x, TEMP[6].xxxx
 35: ENDIF
 36: MUL TEMP[4].xyz, TEMP[4].xxxx, TEMP[2].xzww
 37: MOV TEMP[0].z, TEMP[2].yyyy
 38: DP3 TEMP[0].x, TEMP[0].yzww, TEMP[0].yzww
 39: FSLT TEMP[6].x, TEMP[0].xxxx, IMM[0].zzzz
 40: UIF TEMP[6].xxxx :0
 41:   MOV TEMP[6].x, IMM[0].zzzz
 42: ELSE :0
 43:   FSLT TEMP[7].x, IMM[0].zzzz, TEMP[0].xxxx
 44:   UIF TEMP[7].xxxx :0
 45:     RSQ TEMP[7].x, TEMP[0].xxxx
 46:   ELSE :0
 47:     SSG TEMP[8].x, TEMP[0].xxxx
 48:     MUL TEMP[7].x, IMM[0].wwww, TEMP[8].xxxx
 49:   ENDIF
 50:   MOV TEMP[6].x, TEMP[7].xxxx
 51: ENDIF
 52: MUL TEMP[5].xyz, TEMP[6].xxxx, TEMP[0].zyww
 53: MAD TEMP[6].xy, IN[6].xyyy, CONST[1][52].xyyy, CONST[1][52].wzzz
 54: MOV TEMP[7].xy, IN[3].xyxx
 55: MOV TEMP[7].zw, IN[4].xxyx
 56: MOV TEMP[8].xy, IN[5].xyxx
 57: MOV TEMP[8].zw, IMM[0].zzzz
 58: MAD TEMP[0].x, CONST[1][18].wwww, IMM[3].xxxx, CONST[1][46].wwww
 59: MAD TEMP[1].xyz, IN[0].xyzz, CONST[1][57].xyzz, CONST[1][56].xyzz
 60: MAD TEMP[9].x, CONST[1][49].wwww, TEMP[1].xxxx, TEMP[0].xxxx
 61: FSGE TEMP[10].x, TEMP[9].xxxx, -TEMP[9].xxxx
 62: UIF TEMP[10].xxxx :0
 63:   MOV TEMP[10].x, IMM[4].zzzz
 64: ELSE :0
 65:   MOV TEMP[10].x, IMM[1].xxxx
 66: ENDIF
 67: MOV TEMP[0].z, TEMP[10].xxxx
 68: ABS TEMP[9].x, TEMP[9].xxxx
 69: FRC TEMP[9].x, TEMP[9].xxxx
 70: USNE TEMP[10].x, TEMP[0].zzzz, IMM[5].xxxx
 71: UIF TEMP[10].xxxx :0
 72:   MOV TEMP[10].x, TEMP[9].xxxx
 73: ELSE :0
 74:   MOV TEMP[10].x, -TEMP[9].xxxx
 75: ENDIF
 76: MUL TEMP[9].x, TEMP[10].xxxx, IMM[3].yyyy
 77: MAD TEMP[10].x, CONST[1][50].xxxx, TEMP[1].xxxx, TEMP[0].xxxx
 78: MUL TEMP[10].x, TEMP[10].xxxx, IMM[3].zzzz
 79: FSGE TEMP[11].x, TEMP[10].xxxx, -TEMP[10].xxxx
 80: UIF TEMP[11].xxxx :0
 81:   MOV TEMP[11].x, IMM[4].zzzz
 82: ELSE :0
 83:   MOV TEMP[11].x, IMM[1].xxxx
 84: ENDIF
 85: MOV TEMP[0].w, TEMP[11].xxxx
 86: ABS TEMP[10].x, TEMP[10].xxxx
 87: FRC TEMP[10].x, TEMP[10].xxxx
 88: USNE TEMP[11].x, TEMP[0].wwww, IMM[5].xxxx
 89: UIF TEMP[11].xxxx :0
 90:   MOV TEMP[11].x, TEMP[10].xxxx
 91: ELSE :0
 92:   MOV TEMP[11].x, -TEMP[10].xxxx
 93: ENDIF
 94: MUL TEMP[10].x, TEMP[11].xxxx, IMM[3].yyyy
 95: SIN TEMP[9].x, TEMP[9].xxxx
 96: SIN TEMP[9].y, TEMP[10].xxxx
 97: ADD TEMP[9].x, TEMP[9].yyyy, TEMP[9].xxxx
 98: MUL TEMP[9].x, TEMP[9].xxxx, CONST[1][50].yyyy
 99: MUL TEMP[10].xy, TEMP[0].xxxx, IMM[6].xyyy
100: ADD TEMP[0].x, TEMP[0].xxxx, CONST[1][48].xxxx
101: FSGE TEMP[11].xy, TEMP[10].xyxx, -TEMP[10].xyxx
102: AND TEMP[11].xy, TEMP[11].xyyy, IMM[5].yyyy
103: INEG TEMP[11].xy, TEMP[11].xyyy
104: MOV TEMP[2].xy, TEMP[11].xyxx
105: ABS TEMP[10].xy, TEMP[10].xyyy
106: FRC TEMP[10].xy, TEMP[10].xyyy
107: USNE TEMP[11].x, TEMP[2].xxxx, IMM[5].xxxx
108: UIF TEMP[11].xxxx :0
109:   MOV TEMP[11].x, TEMP[10].xxxx
110: ELSE :0
111:   MOV TEMP[11].x, -TEMP[10].xxxx
112: ENDIF
113: MOV TEMP[0].z, TEMP[11].xxxx
114: USNE TEMP[11].x, TEMP[2].yyyy, IMM[5].xxxx
115: UIF TEMP[11].xxxx :0
116:   MOV TEMP[11].x, TEMP[10].yyyy
117: ELSE :0
118:   MOV TEMP[11].x, -TEMP[10].yyyy
119: ENDIF
120: MOV TEMP[0].w, TEMP[11].xxxx
121: MUL TEMP[10].xy, TEMP[0].zwww, IMM[3].yyyy
122: SIN TEMP[11].x, TEMP[10].xxxx
123: SIN TEMP[11].y, TEMP[10].yyyy
124: ADD TEMP[10].x, TEMP[11].yyyy, TEMP[11].xxxx
125: ADD TEMP[10].x, TEMP[10].xxxx, IMM[3].wwww
126: MUL TEMP[11].x, TEMP[1].xxxx, CONST[1][49].zzzz
127: MUL TEMP[10].x, TEMP[10].xxxx, TEMP[11].xxxx
128: MAD TEMP[9].x, TEMP[10].xxxx, IMM[6].zzzz, TEMP[9].xxxx
129: MOV TEMP[2].z, TEMP[9].xxxx
130: MAD TEMP[9].x, CONST[1][46].zzzz, TEMP[1].xxxx, -TEMP[0].xxxx
131: MOV TEMP[0].y, TEMP[9].xxxx
132: MAD TEMP[9].xy, CONST[1][48].ywww, TEMP[1].xxxx, -TEMP[0].xxxx
133: MOV TEMP[0].xz, TEMP[9].xxyx
134: MUL TEMP[0].xyz, TEMP[0].xyzz, IMM[8].xyzz
135: FSGE TEMP[9].x, TEMP[0].yyyy, -TEMP[0].yyyy
136: UIF TEMP[9].xxxx :0
137:   MOV TEMP[9].x, IMM[4].zzzz
138: ELSE :0
139:   MOV TEMP[9].x, IMM[1].xxxx
140: ENDIF
141: MOV TEMP[0].w, TEMP[9].xxxx
142: ABS TEMP[9].x, TEMP[0].yyyy
143: FRC TEMP[9].x, TEMP[9].xxxx
144: USNE TEMP[10].x, TEMP[0].wwww, IMM[5].xxxx
145: UIF TEMP[10].xxxx :0
146:   MOV TEMP[10].x, TEMP[9].xxxx
147: ELSE :0
148:   MOV TEMP[10].x, -TEMP[9].xxxx
149: ENDIF
150: MUL TEMP[9].x, TEMP[10].xxxx, IMM[3].yyyy
151: FSGE TEMP[10].xy, TEMP[0].xzxx, -TEMP[0].xzxx
152: AND TEMP[10].xy, TEMP[10].xyyy, IMM[5].yyyy
153: INEG TEMP[10].xy, TEMP[10].xyyy
154: MOV TEMP[3].xy, TEMP[10].xyxx
155: ABS TEMP[10].xy, TEMP[0].xzzz
156: FRC TEMP[10].xy, TEMP[10].xyyy
157: USNE TEMP[11].x, TEMP[3].xxxx, IMM[5].xxxx
158: UIF TEMP[11].xxxx :0
159:   MOV TEMP[11].x, TEMP[10].xxxx
160: ELSE :0
161:   MOV TEMP[11].x, -TEMP[10].xxxx
162: ENDIF
163: MOV TEMP[0].x, TEMP[11].xxxx
164: USNE TEMP[3].x, TEMP[3].yyyy, IMM[5].xxxx
165: UIF TEMP[3].xxxx :0
166:   MOV TEMP[3].x, TEMP[10].yyyy
167: ELSE :0
168:   MOV TEMP[3].x, -TEMP[10].yyyy
169: ENDIF
170: MOV TEMP[0].z, TEMP[3].xxxx
171: MUL TEMP[3].xy, TEMP[0].xzzz, IMM[3].yyyy
172: SIN TEMP[10].x, TEMP[3].xxxx
173: SIN TEMP[10].y, TEMP[9].xxxx
174: SIN TEMP[10].z, TEMP[3].yyyy
175: ADD TEMP[0].x, TEMP[10].xxxx, TEMP[10].yyyy
176: MUL TEMP[3].x, TEMP[0].xxxx, CONST[1][49].yyyy
177: ABS TEMP[9].x, TEMP[0].xxxx
178: MUL TEMP[0].x, TEMP[9].xxxx, CONST[1][48].zzzz
179: MAD TEMP[3].x, CONST[1][49].xxxx, TEMP[10].zzzz, TEMP[3].xxxx
180: MOV TEMP[2].y, TEMP[3].xxxx
181: MAD TEMP[3].x, CONST[1][46].yyyy, IMM[6].wwww, IMM[8].wwww
182: MUL TEMP[2].x, TEMP[0].xxxx, TEMP[3].xxxx
183: MUL TEMP[0].xyz, TEMP[2].xyzz, IN[5].xxxx
184: MOV TEMP[1].w, IN[0].wwww
185: MOV TEMP[0].w, IMM[0].zzzz
186: ADD TEMP[0], TEMP[0], TEMP[1]
187: MUL TEMP[1], TEMP[0].yyyy, CONST[1][1]
188: MAD TEMP[1], CONST[1][0], TEMP[0].xxxx, TEMP[1]
189: MAD TEMP[1], CONST[1][2], TEMP[0].zzzz, TEMP[1]
190: MAD TEMP[0], CONST[1][3], TEMP[0].wwww, TEMP[1]
191: MUL TEMP[1], TEMP[0].yyyy, CONST[2][1]
192: MAD TEMP[1], CONST[2][0], TEMP[0].xxxx, TEMP[1]
193: MAD TEMP[1], CONST[2][2], TEMP[0].zzzz, TEMP[1]
194: MAD TEMP[0], CONST[2][3], TEMP[0].wwww, TEMP[1]
195: MOV TEMP[1].xw, TEMP[0].xxxw
196: MOV TEMP[1].y, -TEMP[0].yyyy
197: MAD TEMP[0].x, TEMP[0].zzzz, IMM[0].xxxx, -TEMP[0].wwww
198: MOV TEMP[1].z, TEMP[0].xxxx
199: MOV OUT[1], TEMP[4]
200: MOV OUT[2], TEMP[5]
201: MOV OUT[3], TEMP[6]
202: MOV OUT[4], TEMP[7]
203: MOV OUT[5], TEMP[8]
204: MOV OUT[0], TEMP[1]
205: END
radeonsi: Compiling shader 49
TGSI shader LLVM IR:

; ModuleID = 'tgsi'
source_filename = "tgsi"
target triple = "amdgcn--"

define amdgpu_vs <{ float, float, float }> @main([17 x <16 x i8>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, [32 x <8 x i32>] addrspace(2)* byval, [16 x <8 x i32>] addrspace(2)* byval, [16 x <4 x i32>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, i32 inreg, i32 inreg, i32 inreg, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32) {
main_body:
  %20 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 1, !amdgpu.uniform !0
  %21 = load <16 x i8>, <16 x i8> addrspace(2)* %20, align 16, !tbaa !1
  %22 = call float @llvm.SI.load.const(<16 x i8> %21, i32 0)
  %23 = call float @llvm.SI.load.const(<16 x i8> %21, i32 4)
  %24 = call float @llvm.SI.load.const(<16 x i8> %21, i32 8)
  %25 = call float @llvm.SI.load.const(<16 x i8> %21, i32 12)
  %26 = call float @llvm.SI.load.const(<16 x i8> %21, i32 16)
  %27 = call float @llvm.SI.load.const(<16 x i8> %21, i32 20)
  %28 = call float @llvm.SI.load.const(<16 x i8> %21, i32 24)
  %29 = call float @llvm.SI.load.const(<16 x i8> %21, i32 28)
  %30 = call float @llvm.SI.load.const(<16 x i8> %21, i32 32)
  %31 = call float @llvm.SI.load.const(<16 x i8> %21, i32 36)
  %32 = call float @llvm.SI.load.const(<16 x i8> %21, i32 40)
  %33 = call float @llvm.SI.load.const(<16 x i8> %21, i32 44)
  %34 = call float @llvm.SI.load.const(<16 x i8> %21, i32 48)
  %35 = call float @llvm.SI.load.const(<16 x i8> %21, i32 52)
  %36 = call float @llvm.SI.load.const(<16 x i8> %21, i32 56)
  %37 = call float @llvm.SI.load.const(<16 x i8> %21, i32 60)
  %38 = call float @llvm.SI.load.const(<16 x i8> %21, i32 300)
  %39 = call float @llvm.SI.load.const(<16 x i8> %21, i32 740)
  %40 = call float @llvm.SI.load.const(<16 x i8> %21, i32 744)
  %41 = call float @llvm.SI.load.const(<16 x i8> %21, i32 748)
  %42 = call float @llvm.SI.load.const(<16 x i8> %21, i32 768)
  %43 = call float @llvm.SI.load.const(<16 x i8> %21, i32 772)
  %44 = call float @llvm.SI.load.const(<16 x i8> %21, i32 776)
  %45 = call float @llvm.SI.load.const(<16 x i8> %21, i32 780)
  %46 = call float @llvm.SI.load.const(<16 x i8> %21, i32 784)
  %47 = call float @llvm.SI.load.const(<16 x i8> %21, i32 788)
  %48 = call float @llvm.SI.load.const(<16 x i8> %21, i32 792)
  %49 = call float @llvm.SI.load.const(<16 x i8> %21, i32 796)
  %50 = call float @llvm.SI.load.const(<16 x i8> %21, i32 800)
  %51 = call float @llvm.SI.load.const(<16 x i8> %21, i32 804)
  %52 = call float @llvm.SI.load.const(<16 x i8> %21, i32 816)
  %53 = call float @llvm.SI.load.const(<16 x i8> %21, i32 832)
  %54 = call float @llvm.SI.load.const(<16 x i8> %21, i32 836)
  %55 = call float @llvm.SI.load.const(<16 x i8> %21, i32 840)
  %56 = call float @llvm.SI.load.const(<16 x i8> %21, i32 844)
  %57 = call float @llvm.SI.load.const(<16 x i8> %21, i32 896)
  %58 = call float @llvm.SI.load.const(<16 x i8> %21, i32 900)
  %59 = call float @llvm.SI.load.const(<16 x i8> %21, i32 904)
  %60 = call float @llvm.SI.load.const(<16 x i8> %21, i32 912)
  %61 = call float @llvm.SI.load.const(<16 x i8> %21, i32 916)
  %62 = call float @llvm.SI.load.const(<16 x i8> %21, i32 920)
  %63 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 2, !amdgpu.uniform !0
  %64 = load <16 x i8>, <16 x i8> addrspace(2)* %63, align 16, !tbaa !1
  %65 = call float @llvm.SI.load.const(<16 x i8> %64, i32 0)
  %66 = call float @llvm.SI.load.const(<16 x i8> %64, i32 4)
  %67 = call float @llvm.SI.load.const(<16 x i8> %64, i32 8)
  %68 = call float @llvm.SI.load.const(<16 x i8> %64, i32 12)
  %69 = call float @llvm.SI.load.const(<16 x i8> %64, i32 16)
  %70 = call float @llvm.SI.load.const(<16 x i8> %64, i32 20)
  %71 = call float @llvm.SI.load.const(<16 x i8> %64, i32 24)
  %72 = call float @llvm.SI.load.const(<16 x i8> %64, i32 28)
  %73 = call float @llvm.SI.load.const(<16 x i8> %64, i32 32)
  %74 = call float @llvm.SI.load.const(<16 x i8> %64, i32 36)
  %75 = call float @llvm.SI.load.const(<16 x i8> %64, i32 40)
  %76 = call float @llvm.SI.load.const(<16 x i8> %64, i32 44)
  %77 = call float @llvm.SI.load.const(<16 x i8> %64, i32 48)
  %78 = call float @llvm.SI.load.const(<16 x i8> %64, i32 52)
  %79 = call float @llvm.SI.load.const(<16 x i8> %64, i32 56)
  %80 = call float @llvm.SI.load.const(<16 x i8> %64, i32 60)
  %81 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 0, !amdgpu.uniform !0
  %82 = load <16 x i8>, <16 x i8> addrspace(2)* %81, align 16, !tbaa !1
  %83 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %82, i32 0, i32 %13)
  %84 = extractelement <4 x float> %83, i32 0
  %85 = extractelement <4 x float> %83, i32 1
  %86 = extractelement <4 x float> %83, i32 2
  %87 = extractelement <4 x float> %83, i32 3
  %88 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 1, !amdgpu.uniform !0
  %89 = load <16 x i8>, <16 x i8> addrspace(2)* %88, align 16, !tbaa !1
  %90 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %89, i32 0, i32 %14)
  %91 = extractelement <4 x float> %90, i32 0
  %92 = extractelement <4 x float> %90, i32 1
  %93 = extractelement <4 x float> %90, i32 2
  %94 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 2, !amdgpu.uniform !0
  %95 = load <16 x i8>, <16 x i8> addrspace(2)* %94, align 16, !tbaa !1
  %96 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %95, i32 0, i32 %15)
  %97 = extractelement <4 x float> %96, i32 0
  %98 = extractelement <4 x float> %96, i32 1
  %99 = extractelement <4 x float> %96, i32 2
  %100 = extractelement <4 x float> %96, i32 3
  %101 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 3, !amdgpu.uniform !0
  %102 = load <16 x i8>, <16 x i8> addrspace(2)* %101, align 16, !tbaa !1
  %103 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %102, i32 0, i32 %16)
  %104 = extractelement <4 x float> %103, i32 0
  %105 = extractelement <4 x float> %103, i32 1
  %106 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 4, !amdgpu.uniform !0
  %107 = load <16 x i8>, <16 x i8> addrspace(2)* %106, align 16, !tbaa !1
  %108 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %107, i32 0, i32 %17)
  %109 = extractelement <4 x float> %108, i32 0
  %110 = extractelement <4 x float> %108, i32 1
  %111 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 5, !amdgpu.uniform !0
  %112 = load <16 x i8>, <16 x i8> addrspace(2)* %111, align 16, !tbaa !1
  %113 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %112, i32 0, i32 %18)
  %114 = extractelement <4 x float> %113, i32 0
  %115 = extractelement <4 x float> %113, i32 1
  %116 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 6, !amdgpu.uniform !0
  %117 = load <16 x i8>, <16 x i8> addrspace(2)* %116, align 16, !tbaa !1
  %118 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %117, i32 0, i32 %19)
  %119 = extractelement <4 x float> %118, i32 0
  %120 = extractelement <4 x float> %118, i32 1
  %121 = fmul float %91, 2.000000e+00
  %122 = fadd float %121, -1.000000e+00
  %123 = fmul float %92, 2.000000e+00
  %124 = fadd float %123, -1.000000e+00
  %125 = fmul float %93, 2.000000e+00
  %126 = fadd float %125, -1.000000e+00
  %127 = fmul float %97, 2.000000e+00
  %128 = fadd float %127, -1.000000e+00
  %129 = fmul float %98, 2.000000e+00
  %130 = fadd float %129, -1.000000e+00
  %131 = fmul float %99, 2.000000e+00
  %132 = fadd float %131, -1.000000e+00
  %133 = fmul float %100, 2.000000e+00
  %134 = fadd float %133, -1.000000e+00
  %135 = fmul float %122, %130
  %136 = fmul float %124, %132
  %137 = fmul float %126, %128
  %138 = fmul float %128, %124
  %139 = fsub float %138, %135
  %140 = fmul float %130, %126
  %141 = fsub float %140, %136
  %142 = fmul float %132, %122
  %143 = fsub float %142, %137
  %144 = fmul float %134, %139
  %145 = fmul float %134, %141
  %146 = fmul float %134, %143
  %147 = fmul float %130, %144
  %148 = fmul float %132, %145
  %149 = fmul float %128, %146
  %150 = fmul float %146, %132
  %151 = fsub float %150, %147
  %152 = fmul float %144, %128
  %153 = fsub float %152, %148
  %154 = fmul float %145, %130
  %155 = fsub float %154, %149
  %156 = fmul float %134, %151
  %157 = fmul float %134, %153
  %158 = fmul float %134, %155
  %159 = fmul float %157, %26
  %160 = fmul float %157, %28
  %161 = fmul float %22, %156
  %162 = fadd float %161, %159
  %163 = fmul float %24, %156
  %164 = fadd float %163, %160
  %165 = fmul float %30, %158
  %166 = fadd float %165, %162
  %167 = fmul float %32, %158
  %168 = fadd float %167, %164
  %169 = fmul float %146, %26
  %170 = fmul float %146, %28
  %171 = fmul float %22, %145
  %172 = fadd float %171, %169
  %173 = fmul float %24, %145
  %174 = fadd float %173, %170
  %175 = fmul float %30, %144
  %176 = fadd float %175, %172
  %177 = fmul float %32, %144
  %178 = fadd float %177, %174
  %179 = fmul float %130, %26
  %180 = fmul float %130, %28
  %181 = fmul float %22, %128
  %182 = fadd float %181, %179
  %183 = fmul float %24, %128
  %184 = fadd float %183, %180
  %185 = fmul float %30, %132
  %186 = fadd float %185, %182
  %187 = fmul float %32, %132
  %188 = fadd float %187, %184
  %189 = fmul float %134, %52
  %190 = fmul float %166, %166
  %191 = fmul float %176, %176
  %192 = fadd float %191, %190
  %193 = fmul float %186, %186
  %194 = fadd float %192, %193
  br i1 false, label %ENDIF, label %ELSE

ELSE:                                             ; preds = %main_body
  %195 = fcmp ogt float %194, 0.000000e+00
  br i1 %195, label %IF49, label %ELSE50

ENDIF:                                            ; preds = %IF49, %ELSE50, %main_body
  %temp16.0 = phi float [ 0.000000e+00, %main_body ], [ %205, %IF49 ], [ %209, %ELSE50 ]
  %196 = fmul float %temp16.0, %166
  %197 = fmul float %temp16.0, %176
  %198 = fmul float %temp16.0, %186
  %199 = fmul float %178, %178
  %200 = fmul float %168, %168
  %201 = fadd float %200, %199
  %202 = fmul float %188, %188
  %203 = fadd float %201, %202
  br i1 false, label %ENDIF51, label %ELSE53

IF49:                                             ; preds = %ELSE
  %204 = call float @llvm.sqrt.f32(float %194)
  %205 = fdiv float 1.000000e+00, %204
  br label %ENDIF

ELSE50:                                           ; preds = %ELSE
  %206 = fcmp ogt float %194, 0.000000e+00
  %207 = select i1 %206, float 1.000000e+00, float %194
  %208 = fcmp oge float %207, 0.000000e+00
  %.op = fmul float %207, 0x47EFFFFFE0000000
  %209 = select i1 %208, float %.op, float 0xC7EFFFFFE0000000
  br label %ENDIF

ELSE53:                                           ; preds = %ENDIF
  %210 = fcmp ogt float %203, 0.000000e+00
  br i1 %210, label %IF55, label %ELSE56

ENDIF51:                                          ; preds = %IF55, %ELSE56, %ENDIF
  %temp24.1 = phi float [ 0.000000e+00, %ENDIF ], [ %383, %IF55 ], [ %387, %ELSE56 ]
  %211 = fmul float %temp24.1, %168
  %212 = fmul float %temp24.1, %178
  %213 = fmul float %temp24.1, %188
  %214 = fmul float %119, %53
  %215 = fadd float %214, %56
  %216 = fmul float %120, %54
  %217 = fadd float %216, %55
  %218 = fmul float %38, 0x3FF35C2900000000
  %219 = fadd float %218, %41
  %220 = fmul float %84, %60
  %221 = fadd float %220, %57
  %222 = fmul float %85, %61
  %223 = fadd float %222, %58
  %224 = fmul float %86, %62
  %225 = fadd float %224, %59
  %226 = fmul float %49, %221
  %227 = fadd float %226, %219
  %228 = fsub float -0.000000e+00, %227
  %229 = fcmp oge float %227, %228
  %230 = call float @llvm.fabs.f32(float %227)
  %231 = call float @llvm.floor.f32(float %230)
  %232 = fsub float %230, %231
  %233 = fsub float -0.000000e+00, %232
  %temp40.1 = select i1 %229, float %232, float %233
  %234 = fmul float %temp40.1, 0x401921FB60000000
  %235 = fmul float %50, %221
  %236 = fadd float %235, %219
  %237 = fmul float %236, 5.000000e-01
  %238 = fsub float -0.000000e+00, %237
  %239 = fcmp oge float %237, %238
  %240 = call float @llvm.fabs.f32(float %237)
  %241 = call float @llvm.floor.f32(float %240)
  %242 = fsub float %240, %241
  %243 = fsub float -0.000000e+00, %242
  %temp44.1 = select i1 %239, float %242, float %243
  %244 = fmul float %temp44.1, 0x401921FB60000000
  %245 = call float @llvm.sin.f32(float %234)
  %246 = call float @llvm.sin.f32(float %244)
  %247 = fadd float %246, %245
  %248 = fmul float %247, %51
  %249 = fmul float %219, 0x3FE89D8A00000000
  %250 = fadd float %219, %42
  %251 = fsub float -0.000000e+00, %219
  %252 = fcmp oge float %219, %251
  %253 = fsub float -0.000000e+00, %249
  %254 = fcmp oge float %249, %253
  %255 = call float @llvm.fabs.f32(float %219)
  %256 = call float @llvm.fabs.f32(float %249)
  %257 = call float @llvm.floor.f32(float %255)
  %258 = fsub float %255, %257
  %259 = call float @llvm.floor.f32(float %256)
  %260 = fsub float %256, %259
  %261 = fsub float -0.000000e+00, %258
  %temp44.2 = select i1 %252, float %258, float %261
  %262 = fsub float -0.000000e+00, %260
  %temp44.3 = select i1 %254, float %260, float %262
  %263 = fmul float %temp44.2, 0x401921FB60000000
  %264 = fmul float %temp44.3, 0x401921FB60000000
  %265 = call float @llvm.sin.f32(float %263)
  %266 = call float @llvm.sin.f32(float %264)
  %267 = fadd float %266, %265
  %268 = fadd float %267, 1.000000e+00
  %269 = fmul float %221, %48
  %270 = fmul float %268, %269
  %271 = fmul float %270, 0x3FC99999A0000000
  %272 = fadd float %271, %248
  %273 = fmul float %40, %221
  %274 = fsub float %273, %250
  %275 = fmul float %43, %221
  %276 = fsub float %275, %250
  %277 = fmul float %45, %221
  %278 = fsub float %277, %250
  %279 = fmul float %276, 0x3FCA1F58C0000000
  %280 = fmul float %274, 0x3FDD1745C0000000
  %281 = fsub float -0.000000e+00, %280
  %282 = fcmp oge float %280, %281
  %283 = call float @llvm.fabs.f32(float %280)
  %284 = call float @llvm.floor.f32(float %283)
  %285 = fsub float %283, %284
  %286 = fsub float -0.000000e+00, %285
  %temp40.2 = select i1 %282, float %285, float %286
  %287 = fmul float %temp40.2, 0x401921FB60000000
  %288 = fsub float -0.000000e+00, %279
  %289 = fcmp oge float %279, %288
  %290 = fsub float -0.000000e+00, %278
  %291 = fcmp oge float %278, %290
  %292 = call float @llvm.fabs.f32(float %279)
  %293 = call float @llvm.fabs.f32(float %278)
  %294 = call float @llvm.floor.f32(float %292)
  %295 = fsub float %292, %294
  %296 = call float @llvm.floor.f32(float %293)
  %297 = fsub float %293, %296
  %298 = fsub float -0.000000e+00, %295
  %temp44.4 = select i1 %289, float %295, float %298
  %299 = fsub float -0.000000e+00, %297
  %temp12.0 = select i1 %291, float %297, float %299
  %300 = fmul float %temp44.4, 0x401921FB60000000
  %301 = fmul float %temp12.0, 0x401921FB60000000
  %302 = call float @llvm.sin.f32(float %300)
  %303 = call float @llvm.sin.f32(float %287)
  %304 = call float @llvm.sin.f32(float %301)
  %305 = fadd float %302, %303
  %306 = fmul float %305, %47
  %307 = call float @llvm.fabs.f32(float %305)
  %308 = fmul float %307, %44
  %309 = fmul float %46, %304
  %310 = fadd float %309, %306
  %311 = fmul float %39, -4.000000e+01
  %312 = fadd float %311, -1.000000e+01
  %313 = fmul float %308, %312
  %314 = fmul float %313, %114
  %315 = fmul float %310, %114
  %316 = fmul float %272, %114
  %317 = fadd float %314, %221
  %318 = fadd float %315, %223
  %319 = fadd float %316, %225
  %320 = fadd float %87, 0.000000e+00
  %321 = fmul float %318, %26
  %322 = fmul float %318, %27
  %323 = fmul float %318, %28
  %324 = fmul float %318, %29
  %325 = fmul float %22, %317
  %326 = fadd float %325, %321
  %327 = fmul float %23, %317
  %328 = fadd float %327, %322
  %329 = fmul float %24, %317
  %330 = fadd float %329, %323
  %331 = fmul float %25, %317
  %332 = fadd float %331, %324
  %333 = fmul float %30, %319
  %334 = fadd float %333, %326
  %335 = fmul float %31, %319
  %336 = fadd float %335, %328
  %337 = fmul float %32, %319
  %338 = fadd float %337, %330
  %339 = fmul float %33, %319
  %340 = fadd float %339, %332
  %341 = fmul float %34, %320
  %342 = fadd float %341, %334
  %343 = fmul float %35, %320
  %344 = fadd float %343, %336
  %345 = fmul float %36, %320
  %346 = fadd float %345, %338
  %347 = fmul float %37, %320
  %348 = fadd float %347, %340
  %349 = fmul float %344, %69
  %350 = fmul float %344, %70
  %351 = fmul float %344, %71
  %352 = fmul float %344, %72
  %353 = fmul float %65, %342
  %354 = fadd float %353, %349
  %355 = fmul float %66, %342
  %356 = fadd float %355, %350
  %357 = fmul float %67, %342
  %358 = fadd float %357, %351
  %359 = fmul float %68, %342
  %360 = fadd float %359, %352
  %361 = fmul float %73, %346
  %362 = fadd float %361, %354
  %363 = fmul float %74, %346
  %364 = fadd float %363, %356
  %365 = fmul float %75, %346
  %366 = fadd float %365, %358
  %367 = fmul float %76, %346
  %368 = fadd float %367, %360
  %369 = fmul float %77, %348
  %370 = fadd float %369, %362
  %371 = fmul float %78, %348
  %372 = fadd float %371, %364
  %373 = fmul float %79, %348
  %374 = fadd float %373, %366
  %375 = fmul float %80, %348
  %376 = fadd float %375, %368
  %377 = fsub float -0.000000e+00, %372
  %378 = fmul float %374, 2.000000e+00
  %379 = fsub float %378, %376
  %380 = bitcast i32 %11 to float
  %381 = insertvalue <{ float, float, float }> undef, float %380, 2
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 32, i32 0, float %196, float %197, float %198, float undef)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 33, i32 0, float %211, float %212, float %213, float %189)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 34, i32 0, float %215, float %217, float undef, float undef)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 35, i32 0, float %104, float %105, float %110, float %109)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 36, i32 0, float %114, float %115, float 0.000000e+00, float 0.000000e+00)
  call void @llvm.SI.export(i32 15, i32 0, i32 1, i32 12, i32 0, float %370, float %377, float %379, float %376)
  ret <{ float, float, float }> %381

IF55:                                             ; preds = %ELSE53
  %382 = call float @llvm.sqrt.f32(float %203)
  %383 = fdiv float 1.000000e+00, %382
  br label %ENDIF51

ELSE56:                                           ; preds = %ELSE53
  %384 = fcmp ogt float %203, 0.000000e+00
  %385 = select i1 %384, float 1.000000e+00, float %203
  %386 = fcmp oge float %385, 0.000000e+00
  %.op89 = fmul float %385, 0x47EFFFFFE0000000
  %387 = select i1 %386, float %.op89, float 0xC7EFFFFFE0000000
  br label %ENDIF51
}

; Function Attrs: nounwind readnone
declare float @llvm.SI.load.const(<16 x i8>, i32) #0

; Function Attrs: nounwind readnone
declare <4 x float> @llvm.SI.vs.load.input(<16 x i8>, i32, i32) #0

; Function Attrs: nounwind readnone
declare float @llvm.sqrt.f32(float) #0

; Function Attrs: nounwind readnone
declare float @llvm.fabs.f32(float) #0

; Function Attrs: nounwind readnone
declare float @llvm.floor.f32(float) #0

; Function Attrs: nounwind readnone
declare float @llvm.sin.f32(float) #0

declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)

attributes #0 = { nounwind readnone }

!0 = !{}
!1 = !{!"const", null, i32 1}

[traceshaders] glDetachShader(program=74, shader=73)
[traceshaders] glDeleteShader(shader=73)
[...]
[traceshaders] glCreateShader(GL_FRAGMENT_SHADER) → shader=89
[traceshaders] glShaderSource(shader=89)
[traceshaders] ================================================================================
#version 410
#extension GL_ARB_explicit_attrib_location : require
#ifdef GL_ARB_separate_shader_objects
#extension GL_ARB_separate_shader_objects : enable
#endif
subroutine void SubroutineType();
struct vec1 {
	float x;
};
struct uvec1 {
	uint x;
};
struct ivec1 {
	int x;
};
vec4 InstrHelper;
layout(std140) uniform;
uniform cbuffer_0 { // $Globals 
     vec4 Const0[63];
};
uniform sampler2D tex0_X_smp0;
uniform sampler2D tex0;
layout(location = 0)  in  vec4 VtxGeoOutput0;
vec4 Input0;
layout(location = 1)  in  vec4 VtxGeoOutput1;
vec4 Input1;
layout(location = 2)  in  vec4 VtxGeoOutput2;
vec4 Input2;
layout(location = 0) out  vec4 PixOutput0;
#define Output0 PixOutput0
layout(location = 1) out  vec4 PixOutput1;
#define Output1 PixOutput1
vec4 Temp[2];
ivec4 Temp_int[2];
uvec4 Temp_uint[2];
void main()
{
    Input0 = VtxGeoOutput0;
    Input1 = VtxGeoOutput1;
    Input2 = VtxGeoOutput2;
    Temp[0].x = Input0.x * Input1.y;
    Temp[0].x = Input1.x * Input0.y + (-Temp[0].x);
    Temp[0].y = Temp[0].x * Input1.w;
    Temp[0].x = Input0.z;
    Temp[0].z = Input1.z;
    Temp[0].w = dot(Temp[0].xyz, Temp[0].xyz);
    Temp[0].w = ( ( Temp[0].w < 0.0 ) ? 0.0 : ( ( Temp[0].w > 0.0 ) ? inversesqrt( Temp[0].w ) : ( 3.4028235E+38 * sign( Temp[0].w ) ) ) );
    Temp[0].xyz = Temp[0].www * Temp[0].xyz;
    Output0.xyz = Temp[0].xyz * vec3(uintBitsToFloat(1056964608u), uintBitsToFloat(1056964608u), uintBitsToFloat(1056964608u)) + vec3(uintBitsToFloat(1056964608u), uintBitsToFloat(1056964608u), uintBitsToFloat(1056964608u));
    Output0.xyz = clamp(Output0.xyz, 0.0, 1.0);
    Temp[0].x = Const0[45].x + uintBitsToFloat(3225419776u);
    Temp[0].x = Temp[0].x * uintBitsToFloat(989855744u);
    Temp[0].x = clamp(Temp[0].x, 0.0, 1.0);
    Output0.w = ( ( ( Temp[0].x ) < 0.0 ) ? 0.0 : sqrt( Temp[0].x ) );
    Temp[0] = texture(tex0_X_smp0, Input2.xy);
    Temp[1] = Temp[0].yyyy * Const0[59];
    Temp[1] = Const0[58] * Temp[0].xxxx + Temp[1];
    Temp[1] = Const0[60] * Temp[0].zzzz + Temp[1];
    Temp[0] = Const0[61] * Temp[0].wwww + Temp[1];
    Output1 = Temp[0] + Const0[62];
    return;
}

[traceshaders] ================================================================================
[traceshaders] glCompileShader(shader=89)
[traceshaders] glCreateProgram() = 90
[traceshaders] glAttachShader(program=90, shader=89)
[traceshaders] glLinkProgram(program=90)
SHADER KEY
  prolog.color_two_side = 0
  prolog.poly_stipple = 0
  prolog.force_persample_interp = 0
  epilog.spi_shader_col_format = 0x0
  epilog.color_is_int8 = 0x0
  epilog.last_cbuf = 0
  epilog.alpha_func = 0
  epilog.alpha_to_one = 0
  epilog.poly_line_smoothing = 0
  epilog.clamp_color = 0
FRAG
DCL IN[0], GENERIC[0], PERSPECTIVE
DCL IN[1], GENERIC[1], PERSPECTIVE
DCL IN[2], GENERIC[2], PERSPECTIVE
DCL OUT[0], COLOR
DCL OUT[1], COLOR[1]
DCL SAMP[0]
DCL SVIEW[0], 2D, FLOAT
DCL CONST[1][0..62]
DCL TEMP[0..3], LOCAL
IMM[0] FLT32 {    0.0000, 340282346638528859811704183484516925440.0000,     0.5000,    -3.0000}
IMM[1] UINT32 {0, 720, 944, 928}
IMM[2] FLT32 {    0.0020,     0.0000,     0.0000,     0.0000}
IMM[3] UINT32 {960, 976, 992, 0}
  0: MUL TEMP[0].x, IN[0].xxxx, IN[1].yyyy
  1: MAD TEMP[0].x, IN[1].xxxx, IN[0].yyyy, -TEMP[0].xxxx
  2: MUL TEMP[1].x, TEMP[0].xxxx, IN[1].wwww
  3: MOV TEMP[0].y, TEMP[1].xxxx
  4: MOV TEMP[0].x, IN[0].zzzz
  5: MOV TEMP[0].z, IN[1].zzzz
  6: DP3 TEMP[1].x, TEMP[0].xyzz, TEMP[0].xyzz
  7: FSLT TEMP[2].x, TEMP[1].xxxx, IMM[0].xxxx
  8: UIF TEMP[2].xxxx :0
  9:   MOV TEMP[2].x, IMM[0].xxxx
 10: ELSE :0
 11:   FSLT TEMP[3].x, IMM[0].xxxx, TEMP[1].xxxx
 12:   UIF TEMP[3].xxxx :0
 13:     RSQ TEMP[3].x, TEMP[1].xxxx
 14:   ELSE :0
 15:     SSG TEMP[1].x, TEMP[1].xxxx
 16:     MUL TEMP[3].x, IMM[0].yyyy, TEMP[1].xxxx
 17:   ENDIF
 18:   MOV TEMP[2].x, TEMP[3].xxxx
 19: ENDIF
 20: MUL TEMP[0].xyz, TEMP[2].xxxx, TEMP[0].xyzz
 21: MAD TEMP[1].xyz, TEMP[0].xyzz, IMM[0].zzzz, IMM[0].zzzz
 22: MOV_SAT TEMP[1].xyz, TEMP[1].xyzz
 23: ADD TEMP[0].x, CONST[1][45].xxxx, IMM[0].wwww
 24: MUL TEMP[0].x, TEMP[0].xxxx, IMM[2].xxxx
 25: MOV_SAT TEMP[0].x, TEMP[0].xxxx
 26: FSLT TEMP[2].x, TEMP[0].xxxx, IMM[0].xxxx
 27: UIF TEMP[2].xxxx :0
 28:   MOV TEMP[2].x, IMM[0].xxxx
 29: ELSE :0
 30:   SQRT TEMP[2].x, TEMP[0].xxxx
 31: ENDIF
 32: MOV TEMP[1].w, TEMP[2].xxxx
 33: MOV TEMP[2].xy, IN[2].xyyy
 34: TEX TEMP[2], TEMP[2], SAMP[0], 2D
 35: MUL TEMP[3], TEMP[2].yyyy, CONST[1][59]
 36: MAD TEMP[3], CONST[1][58], TEMP[2].xxxx, TEMP[3]
 37: MAD TEMP[3], CONST[1][60], TEMP[2].zzzz, TEMP[3]
 38: MAD TEMP[0], CONST[1][61], TEMP[2].wwww, TEMP[3]
 39: ADD TEMP[0], TEMP[0], CONST[1][62]
 40: MOV OUT[0], TEMP[1]
 41: MOV OUT[1], TEMP[0]
 42: END
radeonsi: Compiling shader 57
TGSI shader LLVM IR:

; ModuleID = 'tgsi'
source_filename = "tgsi"
target triple = "amdgcn--"

define amdgpu_ps <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> @main([17 x <16 x i8>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, [32 x <8 x i32>] addrspace(2)* byval, [16 x <8 x i32>] addrspace(2)* byval, [16 x <4 x i32>] addrspace(2)* byval, float inreg, i32 inreg, <2 x i32>, <2 x i32>, <2 x i32>, <3 x i32>, <2 x i32>, <2 x i32>, <2 x i32>, float, float, float, float, float, i32, i32, float, i32) #0 {
main_body:
  %23 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 1, !amdgpu.uniform !0
  %24 = load <16 x i8>, <16 x i8> addrspace(2)* %23, align 16, !tbaa !1
  %25 = call float @llvm.SI.load.const(<16 x i8> %24, i32 720)
  %26 = call float @llvm.SI.load.const(<16 x i8> %24, i32 928)
  %27 = call float @llvm.SI.load.const(<16 x i8> %24, i32 932)
  %28 = call float @llvm.SI.load.const(<16 x i8> %24, i32 936)
  %29 = call float @llvm.SI.load.const(<16 x i8> %24, i32 940)
  %30 = call float @llvm.SI.load.const(<16 x i8> %24, i32 944)
  %31 = call float @llvm.SI.load.const(<16 x i8> %24, i32 948)
  %32 = call float @llvm.SI.load.const(<16 x i8> %24, i32 952)
  %33 = call float @llvm.SI.load.const(<16 x i8> %24, i32 956)
  %34 = call float @llvm.SI.load.const(<16 x i8> %24, i32 960)
  %35 = call float @llvm.SI.load.const(<16 x i8> %24, i32 964)
  %36 = call float @llvm.SI.load.const(<16 x i8> %24, i32 968)
  %37 = call float @llvm.SI.load.const(<16 x i8> %24, i32 972)
  %38 = call float @llvm.SI.load.const(<16 x i8> %24, i32 976)
  %39 = call float @llvm.SI.load.const(<16 x i8> %24, i32 980)
  %40 = call float @llvm.SI.load.const(<16 x i8> %24, i32 984)
  %41 = call float @llvm.SI.load.const(<16 x i8> %24, i32 988)
  %42 = call float @llvm.SI.load.const(<16 x i8> %24, i32 992)
  %43 = call float @llvm.SI.load.const(<16 x i8> %24, i32 996)
  %44 = call float @llvm.SI.load.const(<16 x i8> %24, i32 1000)
  %45 = call float @llvm.SI.load.const(<16 x i8> %24, i32 1004)
  %46 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 0, !amdgpu.uniform !0
  %47 = load <8 x i32>, <8 x i32> addrspace(2)* %46, align 32, !tbaa !1
  %48 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %49 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %48, i64 0, i64 3, !amdgpu.uniform !0
  %50 = load <4 x i32>, <4 x i32> addrspace(2)* %49, align 16, !tbaa !1
  %51 = call float @llvm.SI.fs.interp(i32 0, i32 0, i32 %6, <2 x i32> %8)
  %52 = call float @llvm.SI.fs.interp(i32 1, i32 0, i32 %6, <2 x i32> %8)
  %53 = call float @llvm.SI.fs.interp(i32 2, i32 0, i32 %6, <2 x i32> %8)
  %54 = call float @llvm.SI.fs.interp(i32 0, i32 1, i32 %6, <2 x i32> %8)
  %55 = call float @llvm.SI.fs.interp(i32 1, i32 1, i32 %6, <2 x i32> %8)
  %56 = call float @llvm.SI.fs.interp(i32 2, i32 1, i32 %6, <2 x i32> %8)
  %57 = call float @llvm.SI.fs.interp(i32 3, i32 1, i32 %6, <2 x i32> %8)
  %58 = call float @llvm.SI.fs.interp(i32 0, i32 2, i32 %6, <2 x i32> %8)
  %59 = call float @llvm.SI.fs.interp(i32 1, i32 2, i32 %6, <2 x i32> %8)
  %60 = fmul float %51, %55
  %61 = fmul float %54, %52
  %62 = fsub float %61, %60
  %63 = fmul float %62, %57
  %64 = fmul float %53, %53
  %65 = fmul float %63, %63
  %66 = fadd float %65, %64
  %67 = fmul float %56, %56
  %68 = fadd float %66, %67
  br i1 false, label %ENDIF, label %ELSE

ELSE:                                             ; preds = %main_body
  %69 = fcmp ogt float %68, 0.000000e+00
  br i1 %69, label %IF17, label %ELSE18

ENDIF:                                            ; preds = %IF17, %ELSE18, %main_body
  %temp8.0 = phi float [ 0.000000e+00, %main_body ], [ %140, %IF17 ], [ %144, %ELSE18 ]
  %70 = fmul float %temp8.0, %53
  %71 = fmul float %temp8.0, %63
  %72 = fmul float %temp8.0, %56
  %73 = fmul float %70, 5.000000e-01
  %74 = fadd float %73, 5.000000e-01
  %75 = fmul float %71, 5.000000e-01
  %76 = fadd float %75, 5.000000e-01
  %77 = fmul float %72, 5.000000e-01
  %78 = fadd float %77, 5.000000e-01
  %79 = call float @llvm.AMDGPU.clamp.(float %74, float 0.000000e+00, float 1.000000e+00)
  %80 = call float @llvm.AMDGPU.clamp.(float %76, float 0.000000e+00, float 1.000000e+00)
  %81 = call float @llvm.AMDGPU.clamp.(float %78, float 0.000000e+00, float 1.000000e+00)
  %82 = fadd float %25, -3.000000e+00
  %83 = fmul float %82, 1.953125e-03
  %84 = call float @llvm.AMDGPU.clamp.(float %83, float 0.000000e+00, float 1.000000e+00)
  %85 = fcmp olt float %84, 0.000000e+00
  %86 = call float @llvm.sqrt.f32(float %84)
  %temp8.1 = select i1 %85, float 0.000000e+00, float %86
  %87 = bitcast float %58 to i32
  %88 = bitcast float %59 to i32
  %89 = insertelement <2 x i32> undef, i32 %87, i32 0
  %90 = insertelement <2 x i32> %89, i32 %88, i32 1
  %91 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %90, <8 x i32> %47, <4 x i32> %50, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %92 = extractelement <4 x float> %91, i32 0
  %93 = extractelement <4 x float> %91, i32 1
  %94 = extractelement <4 x float> %91, i32 2
  %95 = extractelement <4 x float> %91, i32 3
  %96 = fmul float %93, %30
  %97 = fmul float %93, %31
  %98 = fmul float %93, %32
  %99 = fmul float %93, %33
  %100 = fmul float %26, %92
  %101 = fadd float %100, %96
  %102 = fmul float %27, %92
  %103 = fadd float %102, %97
  %104 = fmul float %28, %92
  %105 = fadd float %104, %98
  %106 = fmul float %29, %92
  %107 = fadd float %106, %99
  %108 = fmul float %34, %94
  %109 = fadd float %108, %101
  %110 = fmul float %35, %94
  %111 = fadd float %110, %103
  %112 = fmul float %36, %94
  %113 = fadd float %112, %105
  %114 = fmul float %37, %94
  %115 = fadd float %114, %107
  %116 = fmul float %38, %95
  %117 = fadd float %116, %109
  %118 = fmul float %39, %95
  %119 = fadd float %118, %111
  %120 = fmul float %40, %95
  %121 = fadd float %120, %113
  %122 = fmul float %41, %95
  %123 = fadd float %122, %115
  %124 = fadd float %117, %42
  %125 = fadd float %119, %43
  %126 = fadd float %121, %44
  %127 = fadd float %123, %45
  %128 = bitcast float %5 to i32
  %129 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> undef, i32 %128, 10
  %130 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %129, float %79, 11
  %131 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %130, float %80, 12
  %132 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %131, float %81, 13
  %133 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %132, float %temp8.1, 14
  %134 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %133, float %124, 15
  %135 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %134, float %125, 16
  %136 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %135, float %126, 17
  %137 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %136, float %127, 18
  %138 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %137, float %21, 24
  ret <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %138

IF17:                                             ; preds = %ELSE
  %139 = call float @llvm.sqrt.f32(float %68)
  %140 = fdiv float 1.000000e+00, %139
  br label %ENDIF

ELSE18:                                           ; preds = %ELSE
  %141 = fcmp ogt float %68, 0.000000e+00
  %142 = select i1 %141, float 1.000000e+00, float %68
  %143 = fcmp oge float %142, 0.000000e+00
  %.op = fmul float %142, 0x47EFFFFFE0000000
  %144 = select i1 %143, float %.op, float 0xC7EFFFFFE0000000
  br label %ENDIF
}

; Function Attrs: nounwind readnone
declare float @llvm.SI.load.const(<16 x i8>, i32) #1

; Function Attrs: nounwind readnone
declare float @llvm.SI.fs.interp(i32, i32, i32, <2 x i32>) #1

; Function Attrs: nounwind readnone
declare float @llvm.sqrt.f32(float) #1

; Function Attrs: readnone
declare float @llvm.AMDGPU.clamp.(float, float, float) #2

; Function Attrs: nounwind readnone
declare <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32>, <8 x i32>, <4 x i32>, i32, i32, i32, i32, i32, i32, i32, i32) #1

attributes #0 = { "InitialPSInputAddr"="36983" }
attributes #1 = { nounwind readnone }
attributes #2 = { readnone }

!0 = !{}
!1 = !{!"const", null, i32 1}

[traceshaders] glDetachShader(program=90, shader=89)
[traceshaders] glDeleteShader(shader=89)
[...]
[traceshaders] glCreateShader(GL_FRAGMENT_SHADER) → shader=381
[traceshaders] glShaderSource(shader=381)
[traceshaders] ================================================================================
#version 410
#extension GL_ARB_explicit_attrib_location : require
#ifdef GL_ARB_separate_shader_objects
#extension GL_ARB_separate_shader_objects : enable
#endif
subroutine void SubroutineType();
struct vec1 {
	float x;
};
struct uvec1 {
	uint x;
};
struct ivec1 {
	int x;
};
vec4 InstrHelper;
layout(std140) uniform;
uniform cbuffer_0 { // $Globals 
     vec4 Const0[46];
};
uniform cbuffer_2 { // PSOffsetConstants 
     vec4 Const2[3];
};
uniform sampler2D tex0_X_smp0;
uniform sampler2D tex0;
uniform sampler2D tex1_X_smp1;
uniform sampler2D tex1;
uniform sampler2D tex2_X_smp2;
uniform sampler2D tex2;
uniform sampler2D tex3_X_smp3;
uniform sampler2D tex3;
uniform sampler2D tex4_X_smp4;
uniform sampler2D tex4;
layout(location = 0)  in  vec4 VtxGeoOutput0;
vec4 Input0;
layout(location = 1)  in  vec4 VtxGeoOutput1;
vec4 Input1;
layout(location = 3)  in  vec4 VtxGeoOutput3;
vec4 Input3;
layout(location = 4)  in  vec4 VtxGeoOutput4;
vec4 Input4;
layout(location = 0) out  vec4 PixOutput0;
#define Output0 PixOutput0
vec4 Temp[5];
ivec4 Temp_int[5];
uvec4 Temp_uint[5];
void main()
{
    Input0 = VtxGeoOutput0;
    Input1 = VtxGeoOutput1;
    Input3 = VtxGeoOutput3;
    Input4 = VtxGeoOutput4;
    Temp[0].x = dot(Input4.xyz, Input4.xyz);
    Temp[0].x = ( ( Temp[0].x < 0.0 ) ? 0.0 : ( ( Temp[0].x > 0.0 ) ? inversesqrt( Temp[0].x ) : ( 3.4028235E+38 * sign( Temp[0].x ) ) ) );
    Temp[0].xyz = Temp[0].xxx * Input4.xyz;
    Temp[0].xyz = Temp[0].zzz * vec3(uintBitsToFloat(0u), uintBitsToFloat(0u), uintBitsToFloat(1073741824u)) + (-Temp[0].xyz);
    Temp[1].x = dot(Temp[0].yz, vec2(uintBitsToFloat(1062274540u), uintBitsToFloat(1058262330u)));
    Temp[1].x = clamp(Temp[1].x, 0.0, 1.0);
    Temp[1].y = dot(Temp[0].zxy, vec3(uintBitsToFloat(1058262330u), uintBitsToFloat(3207922931u), uintBitsToFloat(3201369580u)));
    Temp[1].y = clamp(Temp[1].y, 0.0, 1.0);
    Temp[1].z = dot(Temp[0].xyz, vec3(uintBitsToFloat(1060439283u), uintBitsToFloat(3201369580u), uintBitsToFloat(1058262330u)));
    Temp[1].z = clamp(Temp[1].z, 0.0, 1.0);
    Temp[0].xyz = max(Temp[1].xyz, vec3(uintBitsToFloat(897988541u), uintBitsToFloat(897988541u), uintBitsToFloat(897988541u)));
    InstrHelper.x = ( ( Temp[0].x ) <= 0.0 ) ? -3.4028235E+38 : log2( Temp[0].x );
    InstrHelper.y = ( ( Temp[0].y ) <= 0.0 ) ? -3.4028235E+38 : log2( Temp[0].y );
    InstrHelper.z = ( ( Temp[0].z ) <= 0.0 ) ? -3.4028235E+38 : log2( Temp[0].z );
    Temp[0].xyz = vec3(InstrHelper.xyz);
    Temp[0].w = Const0[29].x + uintBitsToFloat(1065353216u);
    Temp[0].xyz = Temp[0].xyz * Temp[0].www;
    Temp[0].xyz = exp2(Temp[0].xyz);
    Temp[1].xyz = texture(tex1_X_smp1, Input0.xy).xyz;
    Temp[1].xyz = Temp[1].xyz * Const0[44].xyz;
    Temp[0].x = dot(Temp[1].xyz, Temp[0].xyz);
    Temp[0].y = dot(Temp[1].xyz, vec3(uintBitsToFloat(1051372203u), uintBitsToFloat(1051372203u), uintBitsToFloat(1051372203u)));
    InstrHelper.z = ( Input3.x ) / ( ( Input3.w == 0.0 ) ? 3.0E-37 : ( Input3.w ) );
    InstrHelper.w = ( Input3.y ) / ( ( Input3.w == 0.0 ) ? 3.0E-37 : ( Input3.w ) );
    Temp[0].zw = vec2(InstrHelper.zw);
    Temp[0].zw = Temp[0].zw * Const2[0].xy + Const2[0].wz;
    Temp[1].xyz = texture(tex4_X_smp4, Temp[0].zw).xyz;
    Temp[2].xyz = texture(tex3_X_smp3, Temp[0].zw).xyz;
    Temp[3].xyz = texture(tex0_X_smp0, Input0.xy).xyz;
    Temp[3] = Temp[3].xyzz * Const0[45].xyzz;
    Temp[1] = Temp[3] * Temp[0].xxxx + Temp[1].xyzz;
    Temp[0] = Temp[3].xyww * Temp[0].yyyy + Temp[2].xyzz;
    Temp[2].xy = Input1.xy * Const0[28].xy;
    Temp[2].xyz = texture(tex2_X_smp2, Temp[2].xy).xyz;
    Temp[3] = Temp[2].xyzz * Temp[2].xyzz;
    Temp[4] = Temp[2].xyzz * Temp[3];
    Temp[2].w = dot(Temp[4].xyw, vec3(uintBitsToFloat(1065353216u), uintBitsToFloat(1065353216u), uintBitsToFloat(1065353216u)));
    Temp[3] = (-Temp[3].xyww) * Temp[2].xyzz + Temp[2].wwww;
    Temp[2] = Temp[2].xyzz * Const0[22].xyzz;
    Temp[3] = Const0[28].zzzz * Temp[3] + Temp[4];
    Temp[3] = Temp[3] * Const0[24].xyzz;
    Temp[3] = Temp[3] * Const0[28].wwww;
    Temp[1] = Temp[1] * Temp[3];
    Temp[0] = Temp[2] * Temp[0] + Temp[1];
    Output0 = Temp[0] + Const0[21].xyzz;
    return;
}

[traceshaders] ================================================================================
[traceshaders] glCompileShader(shader=381)
[traceshaders] glCreateProgram() = 382
[traceshaders] glAttachShader(program=382, shader=381)
[traceshaders] glLinkProgram(program=382)
SHADER KEY
  prolog.color_two_side = 0
  prolog.poly_stipple = 0
  prolog.force_persample_interp = 0
  epilog.spi_shader_col_format = 0x0
  epilog.color_is_int8 = 0x0
  epilog.last_cbuf = 0
  epilog.alpha_func = 0
  epilog.alpha_to_one = 0
  epilog.poly_line_smoothing = 0
  epilog.clamp_color = 0
FRAG
DCL IN[0], GENERIC[0], PERSPECTIVE
DCL IN[1], GENERIC[1], PERSPECTIVE
DCL IN[2], GENERIC[3], PERSPECTIVE
DCL IN[3], GENERIC[4], PERSPECTIVE
DCL OUT[0], COLOR
DCL SAMP[0]
DCL SAMP[1]
DCL SAMP[2]
DCL SAMP[3]
DCL SAMP[4]
DCL SVIEW[0], 2D, FLOAT
DCL SVIEW[1], 2D, FLOAT
DCL SVIEW[2], 2D, FLOAT
DCL SVIEW[3], 2D, FLOAT
DCL SVIEW[4], 2D, FLOAT
DCL CONST[1][0..2]
DCL CONST[2][0..45]
DCL TEMP[0..6], LOCAL
IMM[0] FLT32 {    0.0000, 340282346638528859811704183484516925440.0000,     2.0000,     0.0000}
IMM[1] FLT32 {    0.8165,     0.5774,    -0.7071,    -0.4082}
IMM[2] FLT32 {    0.7071,    -0.4082,     0.5774, -340282346638528859811704183484516925440.0000}
IMM[3] UINT32 {1, 464, 704, 0}
IMM[4] FLT32 {    1.0000,     0.3333,     0.0000,     0.0000}
IMM[5] UINT32 {720, 448, 352, 384}
IMM[6] UINT32 {336, 0, 0, 0}
  0: DP3 TEMP[0].x, IN[3].xyzz, IN[3].xyzz
  1: FSLT TEMP[1].x, TEMP[0].xxxx, IMM[0].xxxx
  2: UIF TEMP[1].xxxx :0
  3:   MOV TEMP[1].x, IMM[0].xxxx
  4: ELSE :0
  5:   FSLT TEMP[2].x, IMM[0].xxxx, TEMP[0].xxxx
  6:   UIF TEMP[2].xxxx :0
  7:     RSQ TEMP[2].x, TEMP[0].xxxx
  8:   ELSE :0
  9:     SSG TEMP[3].x, TEMP[0].xxxx
 10:     MUL TEMP[2].x, IMM[0].yyyy, TEMP[3].xxxx
 11:   ENDIF
 12:   MOV TEMP[1].x, TEMP[2].xxxx
 13: ENDIF
 14: MUL TEMP[0].xyz, TEMP[1].xxxx, IN[3].xyzz
 15: MAD TEMP[0].xyz, TEMP[0].zzzz, IMM[0].xxzz, -TEMP[0].xyzz
 16: DP2 TEMP[1].x, TEMP[0].yzzz, IMM[1].xyyy
 17: MOV_SAT TEMP[1].x, TEMP[1].xxxx
 18: DP3 TEMP[2].x, TEMP[0].zxyy, IMM[1].yzww
 19: MOV_SAT TEMP[2].x, TEMP[2].xxxx
 20: MOV TEMP[1].y, TEMP[2].xxxx
 21: DP3 TEMP[2].x, TEMP[0].xyzz, IMM[2].xyzz
 22: MOV_SAT TEMP[2].x, TEMP[2].xxxx
 23: MOV TEMP[1].z, TEMP[2].xxxx
 24: MAX TEMP[0].xyz, TEMP[1].xyzz, IMM[0].wwww
 25: FSGE TEMP[2].x, IMM[0].xxxx, TEMP[0].xxxx
 26: UIF TEMP[2].xxxx :0
 27:   MOV TEMP[2].x, IMM[2].wwww
 28: ELSE :0
 29:   LG2 TEMP[2].x, TEMP[0].xxxx
 30: ENDIF
 31: MOV TEMP[2].x, TEMP[2].xxxx
 32: FSGE TEMP[3].x, IMM[0].xxxx, TEMP[0].yyyy
 33: UIF TEMP[3].xxxx :0
 34:   MOV TEMP[3].x, IMM[2].wwww
 35: ELSE :0
 36:   LG2 TEMP[3].x, TEMP[0].yyyy
 37: ENDIF
 38: MOV TEMP[2].y, TEMP[3].xxxx
 39: FSGE TEMP[3].x, IMM[0].xxxx, TEMP[0].zzzz
 40: UIF TEMP[3].xxxx :0
 41:   MOV TEMP[3].x, IMM[2].wwww
 42: ELSE :0
 43:   LG2 TEMP[3].x, TEMP[0].zzzz
 44: ENDIF
 45: MOV TEMP[2].z, TEMP[3].xxxx
 46: ADD TEMP[3].x, CONST[2][29].xxxx, IMM[4].xxxx
 47: MUL TEMP[0].xyz, TEMP[2].xyzz, TEMP[3].xxxx
 48: EX2 TEMP[3].x, TEMP[0].xxxx
 49: EX2 TEMP[3].y, TEMP[0].yyyy
 50: EX2 TEMP[3].z, TEMP[0].zzzz
 51: MOV TEMP[4].xy, IN[0].xyyy
 52: TEX TEMP[4].xyz, TEMP[4], SAMP[1], 2D
 53: MUL TEMP[1].xyz, TEMP[4].xyzz, CONST[2][44].xyzz
 54: DP3 TEMP[0].x, TEMP[1].xyzz, TEMP[3].xyzz
 55: DP3 TEMP[3].x, TEMP[1].xyzz, IMM[4].yyyy
 56: FSEQ TEMP[4].x, IN[2].wwww, IMM[0].xxxx
 57: UIF TEMP[4].xxxx :0
 58:   MOV TEMP[4].x, IMM[4].zzzz
 59: ELSE :0
 60:   MOV TEMP[4].x, IN[2].wwww
 61: ENDIF
 62: RCP TEMP[4].x, TEMP[4].xxxx
 63: MUL TEMP[4].x, IN[2].xxxx, TEMP[4].xxxx
 64: MOV TEMP[2].z, TEMP[4].xxxx
 65: FSEQ TEMP[4].x, IN[2].wwww, IMM[0].xxxx
 66: UIF TEMP[4].xxxx :0
 67:   MOV TEMP[4].x, IMM[4].zzzz
 68: ELSE :0
 69:   MOV TEMP[4].x, IN[2].wwww
 70: ENDIF
 71: RCP TEMP[4].x, TEMP[4].xxxx
 72: MUL TEMP[4].x, IN[2].yyyy, TEMP[4].xxxx
 73: MOV TEMP[2].w, TEMP[4].xxxx
 74: MAD TEMP[2].xy, TEMP[2].zwww, CONST[1][0].xyyy, CONST[1][0].wzzz
 75: MOV TEMP[4].xy, TEMP[2].xyyy
 76: TEX TEMP[4].xyz, TEMP[4], SAMP[4], 2D
 77: MOV TEMP[2].xy, TEMP[2].xyyy
 78: TEX TEMP[2].xyz, TEMP[2], SAMP[3], 2D
 79: MOV TEMP[5].xy, IN[0].xyyy
 80: TEX TEMP[5].xyz, TEMP[5], SAMP[0], 2D
 81: MUL TEMP[5], TEMP[5].xyzz, CONST[2][45].xyzz
 82: MAD TEMP[1], TEMP[5], TEMP[0].xxxx, TEMP[4].xyzz
 83: MAD TEMP[0], TEMP[5].xyww, TEMP[3].xxxx, TEMP[2].xyzz
 84: MUL TEMP[2].xy, IN[1].xyyy, CONST[2][28].xyyy
 85: MOV TEMP[3].xy, TEMP[2].xyyy
 86: TEX TEMP[3].xyz, TEMP[3], SAMP[2], 2D
 87: MUL TEMP[5], TEMP[3].xyzz, TEMP[3].xyzz
 88: MUL TEMP[4], TEMP[3].xyzz, TEMP[5]
 89: DP3 TEMP[6].x, TEMP[4].xyww, IMM[4].xxxx
 90: MAD TEMP[5], -TEMP[5].xyww, TEMP[3].xyzz, TEMP[6].xxxx
 91: MUL TEMP[2], TEMP[3].xyzz, CONST[2][22].xyzz
 92: MAD TEMP[5], CONST[2][28].zzzz, TEMP[5], TEMP[4]
 93: MUL TEMP[5], TEMP[5], CONST[2][24].xyzz
 94: MUL TEMP[5], TEMP[5], CONST[2][28].wwww
 95: MUL TEMP[1], TEMP[1], TEMP[5]
 96: MAD TEMP[0], TEMP[2], TEMP[0], TEMP[1]
 97: ADD TEMP[0], TEMP[0], CONST[2][21].xyzz
 98: MOV OUT[0], TEMP[0]
 99: END
radeonsi: Compiling shader 208
TGSI shader LLVM IR:

; ModuleID = 'tgsi'
source_filename = "tgsi"
target triple = "amdgcn--"

define amdgpu_ps <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> @main([17 x <16 x i8>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, [32 x <8 x i32>] addrspace(2)* byval, [16 x <8 x i32>] addrspace(2)* byval, [16 x <4 x i32>] addrspace(2)* byval, float inreg, i32 inreg, <2 x i32>, <2 x i32>, <2 x i32>, <3 x i32>, <2 x i32>, <2 x i32>, <2 x i32>, float, float, float, float, float, i32, i32, float, i32) #0 {
main_body:
  %23 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 1, !amdgpu.uniform !0
  %24 = load <16 x i8>, <16 x i8> addrspace(2)* %23, align 16, !tbaa !1
  %25 = call float @llvm.SI.load.const(<16 x i8> %24, i32 0)
  %26 = call float @llvm.SI.load.const(<16 x i8> %24, i32 4)
  %27 = call float @llvm.SI.load.const(<16 x i8> %24, i32 8)
  %28 = call float @llvm.SI.load.const(<16 x i8> %24, i32 12)
  %29 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 2, !amdgpu.uniform !0
  %30 = load <16 x i8>, <16 x i8> addrspace(2)* %29, align 16, !tbaa !1
  %31 = call float @llvm.SI.load.const(<16 x i8> %30, i32 336)
  %32 = call float @llvm.SI.load.const(<16 x i8> %30, i32 340)
  %33 = call float @llvm.SI.load.const(<16 x i8> %30, i32 344)
  %34 = call float @llvm.SI.load.const(<16 x i8> %30, i32 352)
  %35 = call float @llvm.SI.load.const(<16 x i8> %30, i32 356)
  %36 = call float @llvm.SI.load.const(<16 x i8> %30, i32 360)
  %37 = call float @llvm.SI.load.const(<16 x i8> %30, i32 384)
  %38 = call float @llvm.SI.load.const(<16 x i8> %30, i32 388)
  %39 = call float @llvm.SI.load.const(<16 x i8> %30, i32 392)
  %40 = call float @llvm.SI.load.const(<16 x i8> %30, i32 448)
  %41 = call float @llvm.SI.load.const(<16 x i8> %30, i32 452)
  %42 = call float @llvm.SI.load.const(<16 x i8> %30, i32 456)
  %43 = call float @llvm.SI.load.const(<16 x i8> %30, i32 460)
  %44 = call float @llvm.SI.load.const(<16 x i8> %30, i32 464)
  %45 = call float @llvm.SI.load.const(<16 x i8> %30, i32 704)
  %46 = call float @llvm.SI.load.const(<16 x i8> %30, i32 708)
  %47 = call float @llvm.SI.load.const(<16 x i8> %30, i32 712)
  %48 = call float @llvm.SI.load.const(<16 x i8> %30, i32 720)
  %49 = call float @llvm.SI.load.const(<16 x i8> %30, i32 724)
  %50 = call float @llvm.SI.load.const(<16 x i8> %30, i32 728)
  %51 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 0, !amdgpu.uniform !0
  %52 = load <8 x i32>, <8 x i32> addrspace(2)* %51, align 32, !tbaa !1
  %53 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %54 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %53, i64 0, i64 3, !amdgpu.uniform !0
  %55 = load <4 x i32>, <4 x i32> addrspace(2)* %54, align 16, !tbaa !1
  %56 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 2, !amdgpu.uniform !0
  %57 = load <8 x i32>, <8 x i32> addrspace(2)* %56, align 32, !tbaa !1
  %58 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %59 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %58, i64 0, i64 7, !amdgpu.uniform !0
  %60 = load <4 x i32>, <4 x i32> addrspace(2)* %59, align 16, !tbaa !1
  %61 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 4, !amdgpu.uniform !0
  %62 = load <8 x i32>, <8 x i32> addrspace(2)* %61, align 32, !tbaa !1
  %63 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %64 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %63, i64 0, i64 11, !amdgpu.uniform !0
  %65 = load <4 x i32>, <4 x i32> addrspace(2)* %64, align 16, !tbaa !1
  %66 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 6, !amdgpu.uniform !0
  %67 = load <8 x i32>, <8 x i32> addrspace(2)* %66, align 32, !tbaa !1
  %68 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %69 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %68, i64 0, i64 15, !amdgpu.uniform !0
  %70 = load <4 x i32>, <4 x i32> addrspace(2)* %69, align 16, !tbaa !1
  %71 = getelementptr [32 x <8 x i32>], [32 x <8 x i32>] addrspace(2)* %2, i64 0, i64 8, !amdgpu.uniform !0
  %72 = load <8 x i32>, <8 x i32> addrspace(2)* %71, align 32, !tbaa !1
  %73 = bitcast [32 x <8 x i32>] addrspace(2)* %2 to [0 x <4 x i32>] addrspace(2)*
  %74 = getelementptr [0 x <4 x i32>], [0 x <4 x i32>] addrspace(2)* %73, i64 0, i64 19, !amdgpu.uniform !0
  %75 = load <4 x i32>, <4 x i32> addrspace(2)* %74, align 16, !tbaa !1
  %76 = call float @llvm.SI.fs.interp(i32 0, i32 0, i32 %6, <2 x i32> %8)
  %77 = call float @llvm.SI.fs.interp(i32 1, i32 0, i32 %6, <2 x i32> %8)
  %78 = call float @llvm.SI.fs.interp(i32 0, i32 1, i32 %6, <2 x i32> %8)
  %79 = call float @llvm.SI.fs.interp(i32 1, i32 1, i32 %6, <2 x i32> %8)
  %80 = call float @llvm.SI.fs.interp(i32 0, i32 2, i32 %6, <2 x i32> %8)
  %81 = call float @llvm.SI.fs.interp(i32 1, i32 2, i32 %6, <2 x i32> %8)
  %82 = call float @llvm.SI.fs.interp(i32 3, i32 2, i32 %6, <2 x i32> %8)
  %83 = call float @llvm.SI.fs.interp(i32 0, i32 3, i32 %6, <2 x i32> %8)
  %84 = call float @llvm.SI.fs.interp(i32 1, i32 3, i32 %6, <2 x i32> %8)
  %85 = call float @llvm.SI.fs.interp(i32 2, i32 3, i32 %6, <2 x i32> %8)
  %86 = fmul float %83, %83
  %87 = fmul float %84, %84
  %88 = fadd float %87, %86
  %89 = fmul float %85, %85
  %90 = fadd float %88, %89
  br i1 false, label %ENDIF, label %ELSE

ELSE:                                             ; preds = %main_body
  %91 = fcmp ogt float %90, 0.000000e+00
  br i1 %91, label %IF29, label %ELSE30

ENDIF:                                            ; preds = %IF29, %ELSE30, %main_body
  %temp4.0 = phi float [ 0.000000e+00, %main_body ], [ %122, %IF29 ], [ %126, %ELSE30 ]
  %92 = fmul float %temp4.0, %83
  %93 = fmul float %temp4.0, %84
  %94 = fmul float %temp4.0, %85
  %95 = fmul float %94, 0.000000e+00
  %96 = fsub float %95, %92
  %97 = fmul float %94, 0.000000e+00
  %98 = fsub float %97, %93
  %99 = fmul float %94, 2.000000e+00
  %100 = fsub float %99, %94
  %101 = fmul float %98, 0x3FEA20BD80000000
  %102 = fmul float %100, 0x3FE279A740000000
  %103 = fadd float %101, %102
  %104 = call float @llvm.AMDGPU.clamp.(float %103, float 0.000000e+00, float 1.000000e+00)
  %105 = fmul float %100, 0x3FE279A740000000
  %106 = fmul float %96, 0xBFE6A09E60000000
  %107 = fadd float %106, %105
  %108 = fmul float %98, 0xBFDA20BD80000000
  %109 = fadd float %107, %108
  %110 = call float @llvm.AMDGPU.clamp.(float %109, float 0.000000e+00, float 1.000000e+00)
  %111 = fmul float %96, 0x3FE6A09E60000000
  %112 = fmul float %98, 0xBFDA20BD80000000
  %113 = fadd float %112, %111
  %114 = fmul float %100, 0x3FE279A740000000
  %115 = fadd float %113, %114
  %116 = call float @llvm.AMDGPU.clamp.(float %115, float 0.000000e+00, float 1.000000e+00)
  %117 = call float @llvm.maxnum.f32(float %104, float 0x3EB0C6F7A0000000)
  %118 = call float @llvm.maxnum.f32(float %110, float 0x3EB0C6F7A0000000)
  %119 = call float @llvm.maxnum.f32(float %116, float 0x3EB0C6F7A0000000)
  %120 = fcmp ugt float %117, 0.000000e+00
  br i1 %120, label %ELSE33, label %ENDIF31

IF29:                                             ; preds = %ELSE
  %121 = call float @llvm.sqrt.f32(float %90)
  %122 = fdiv float 1.000000e+00, %121
  br label %ENDIF

ELSE30:                                           ; preds = %ELSE
  %123 = fcmp ogt float %90, 0.000000e+00
  %124 = select i1 %123, float 1.000000e+00, float %90
  %125 = fcmp oge float %124, 0.000000e+00
  %.op = fmul float %124, 0x47EFFFFFE0000000
  %126 = select i1 %125, float %.op, float 0xC7EFFFFFE0000000
  br label %ENDIF

ELSE33:                                           ; preds = %ENDIF
  %127 = call float @llvm.log2.f32(float %117)
  br label %ENDIF31

ENDIF31:                                          ; preds = %ENDIF, %ELSE33
  %temp8.1 = phi float [ %127, %ELSE33 ], [ 0xC7EFFFFFE0000000, %ENDIF ]
  %128 = fcmp ugt float %118, 0.000000e+00
  br i1 %128, label %ELSE36, label %ENDIF34

ELSE36:                                           ; preds = %ENDIF31
  %129 = call float @llvm.log2.f32(float %118)
  br label %ENDIF34

ENDIF34:                                          ; preds = %ENDIF31, %ELSE36
  %temp12.0 = phi float [ %129, %ELSE36 ], [ 0xC7EFFFFFE0000000, %ENDIF31 ]
  %130 = fcmp ugt float %119, 0.000000e+00
  br i1 %130, label %ELSE39, label %ENDIF37

ELSE39:                                           ; preds = %ENDIF34
  %131 = call float @llvm.log2.f32(float %119)
  br label %ENDIF37

ENDIF37:                                          ; preds = %ENDIF34, %ELSE39
  %temp12.1 = phi float [ %131, %ELSE39 ], [ 0xC7EFFFFFE0000000, %ENDIF34 ]
  %132 = fadd float %44, 1.000000e+00
  %133 = fmul float %temp8.1, %132
  %134 = fmul float %temp12.0, %132
  %135 = fmul float %temp12.1, %132
  %136 = call float @llvm.exp2.f32(float %133)
  %137 = call float @llvm.exp2.f32(float %134)
  %138 = call float @llvm.exp2.f32(float %135)
  %139 = bitcast float %76 to i32
  %140 = bitcast float %77 to i32
  %141 = insertelement <2 x i32> undef, i32 %139, i32 0
  %142 = insertelement <2 x i32> %141, i32 %140, i32 1
  %143 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %142, <8 x i32> %57, <4 x i32> %60, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %144 = extractelement <4 x float> %143, i32 0
  %145 = extractelement <4 x float> %143, i32 1
  %146 = extractelement <4 x float> %143, i32 2
  %147 = fmul float %144, %45
  %148 = fmul float %145, %46
  %149 = fmul float %146, %47
  %150 = fmul float %147, %136
  %151 = fmul float %148, %137
  %152 = fadd float %151, %150
  %153 = fmul float %149, %138
  %154 = fadd float %152, %153
  %155 = fmul float %147, 0x3FD5555560000000
  %156 = fmul float %148, 0x3FD5555560000000
  %157 = fadd float %156, %155
  %158 = fmul float %149, 0x3FD5555560000000
  %159 = fadd float %157, %158
  %160 = fcmp oeq float %82, 0.000000e+00
  %.op46 = fdiv float 1.000000e+00, %82
  %161 = select i1 %160, float 0x47840FD020000000, float %.op46
  %162 = fmul float %80, %161
  %163 = fcmp oeq float %82, 0.000000e+00
  %.op47 = fdiv float 1.000000e+00, %82
  %164 = select i1 %163, float 0x47840FD020000000, float %.op47
  %165 = fmul float %81, %164
  %166 = fmul float %162, %25
  %167 = fadd float %166, %28
  %168 = fmul float %165, %26
  %169 = fadd float %168, %27
  %170 = bitcast float %167 to i32
  %171 = bitcast float %169 to i32
  %172 = insertelement <2 x i32> undef, i32 %170, i32 0
  %173 = insertelement <2 x i32> %172, i32 %171, i32 1
  %174 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %173, <8 x i32> %72, <4 x i32> %75, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %175 = extractelement <4 x float> %174, i32 0
  %176 = extractelement <4 x float> %174, i32 1
  %177 = extractelement <4 x float> %174, i32 2
  %178 = bitcast float %167 to i32
  %179 = bitcast float %169 to i32
  %180 = insertelement <2 x i32> undef, i32 %178, i32 0
  %181 = insertelement <2 x i32> %180, i32 %179, i32 1
  %182 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %181, <8 x i32> %67, <4 x i32> %70, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %183 = extractelement <4 x float> %182, i32 0
  %184 = extractelement <4 x float> %182, i32 1
  %185 = extractelement <4 x float> %182, i32 2
  %186 = bitcast float %76 to i32
  %187 = bitcast float %77 to i32
  %188 = insertelement <2 x i32> undef, i32 %186, i32 0
  %189 = insertelement <2 x i32> %188, i32 %187, i32 1
  %190 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %189, <8 x i32> %52, <4 x i32> %55, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %191 = extractelement <4 x float> %190, i32 0
  %192 = extractelement <4 x float> %190, i32 1
  %193 = extractelement <4 x float> %190, i32 2
  %194 = fmul float %191, %48
  %195 = fmul float %192, %49
  %196 = fmul float %193, %50
  %197 = fmul float %193, %50
  %198 = fmul float %194, %154
  %199 = fadd float %198, %175
  %200 = fmul float %195, %154
  %201 = fadd float %200, %176
  %202 = fmul float %196, %154
  %203 = fadd float %202, %177
  %204 = fmul float %197, %154
  %205 = fadd float %204, %177
  %206 = fmul float %194, %159
  %207 = fadd float %206, %183
  %208 = fmul float %195, %159
  %209 = fadd float %208, %184
  %210 = fmul float %197, %159
  %211 = fadd float %210, %185
  %212 = fmul float %197, %159
  %213 = fadd float %212, %185
  %214 = fmul float %78, %40
  %215 = fmul float %79, %41
  %216 = bitcast float %214 to i32
  %217 = bitcast float %215 to i32
  %218 = insertelement <2 x i32> undef, i32 %216, i32 0
  %219 = insertelement <2 x i32> %218, i32 %217, i32 1
  %220 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> %219, <8 x i32> %62, <4 x i32> %65, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
  %221 = extractelement <4 x float> %220, i32 0
  %222 = extractelement <4 x float> %220, i32 1
  %223 = extractelement <4 x float> %220, i32 2
  %224 = fmul float %221, %221
  %225 = fmul float %222, %222
  %226 = fmul float %223, %223
  %227 = fmul float %223, %223
  %228 = fmul float %221, %224
  %229 = fmul float %222, %225
  %230 = fmul float %223, %226
  %231 = fmul float %223, %227
  %232 = fadd float %229, %228
  %233 = fadd float %232, %231
  %234 = fmul float %224, %221
  %235 = fsub float %233, %234
  %236 = fmul float %225, %222
  %237 = fsub float %233, %236
  %238 = fmul float %227, %223
  %239 = fsub float %233, %238
  %240 = fmul float %227, %223
  %241 = fsub float %233, %240
  %242 = fmul float %221, %34
  %243 = fmul float %222, %35
  %244 = fmul float %223, %36
  %245 = fmul float %223, %36
  %246 = fmul float %42, %235
  %247 = fadd float %246, %228
  %248 = fmul float %42, %237
  %249 = fadd float %248, %229
  %250 = fmul float %42, %239
  %251 = fadd float %250, %230
  %252 = fmul float %42, %241
  %253 = fadd float %252, %231
  %254 = fmul float %247, %37
  %255 = fmul float %249, %38
  %256 = fmul float %251, %39
  %257 = fmul float %253, %39
  %258 = fmul float %254, %43
  %259 = fmul float %255, %43
  %260 = fmul float %256, %43
  %261 = fmul float %257, %43
  %262 = fmul float %199, %258
  %263 = fmul float %201, %259
  %264 = fmul float %203, %260
  %265 = fmul float %205, %261
  %266 = fmul float %242, %207
  %267 = fadd float %266, %262
  %268 = fmul float %243, %209
  %269 = fadd float %268, %263
  %270 = fmul float %244, %211
  %271 = fadd float %270, %264
  %272 = fmul float %245, %213
  %273 = fadd float %272, %265
  %274 = fadd float %267, %31
  %275 = fadd float %269, %32
  %276 = fadd float %271, %33
  %277 = fadd float %273, %33
  %278 = bitcast float %5 to i32
  %279 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> undef, i32 %278, 10
  %280 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %279, float %274, 11
  %281 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %280, float %275, 12
  %282 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %281, float %276, 13
  %283 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %282, float %277, 14
  %284 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %283, float %21, 24
  ret <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float, float, float, float }> %284
}

; Function Attrs: nounwind readnone
declare float @llvm.SI.load.const(<16 x i8>, i32) #1

; Function Attrs: nounwind readnone
declare float @llvm.SI.fs.interp(i32, i32, i32, <2 x i32>) #1

; Function Attrs: nounwind readnone
declare float @llvm.sqrt.f32(float) #1

; Function Attrs: readnone
declare float @llvm.AMDGPU.clamp.(float, float, float) #2

; Function Attrs: nounwind readnone
declare float @llvm.maxnum.f32(float, float) #1

; Function Attrs: nounwind readnone
declare float @llvm.log2.f32(float) #1

; Function Attrs: nounwind readnone
declare float @llvm.exp2.f32(float) #1

; Function Attrs: nounwind readnone
declare <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32>, <8 x i32>, <4 x i32>, i32, i32, i32, i32, i32, i32, i32, i32) #1

attributes #0 = { "InitialPSInputAddr"="36983" }
attributes #1 = { nounwind readnone }
attributes #2 = { readnone }

!0 = !{}
!1 = !{!"const", null, i32 1}

[traceshaders] glDetachShader(program=382, shader=381)
[traceshaders] glDeleteShader(shader=381)
[...]
[traceshaders] glCreateShader(GL_VERTEX_SHADER) → shader=383
[traceshaders] glShaderSource(shader=383)
[traceshaders] ================================================================================
#version 410
#extension GL_ARB_explicit_attrib_location : require
#ifdef GL_ARB_separate_shader_objects
#extension GL_ARB_separate_shader_objects : enable
#endif
subroutine void SubroutineType();
struct vec1 {
	float x;
};
struct uvec1 {
	uint x;
};
struct ivec1 {
	int x;
};
vec4 InstrHelper;
out gl_PerVertex {
vec4 gl_Position;
float gl_PointSize;
float gl_ClipDistance[];};
layout(std140) uniform;
uniform cbuffer_0 { // $Globals 
     vec4 Const0[46];
};
uniform cbuffer_1 { // VSOffsetConstants 
     vec4 Const1[5];
};
layout(location = 0)  in  vec4 dcl_Input0;
vec4 Input0;
layout(location = 1)  in  vec4 dcl_Input1;
vec4 Input1;
layout(location = 2)  in  vec4 dcl_Input2;
vec4 Input2;
layout(location = 4)  in  vec4 dcl_Input4;
vec4 Input4;
layout(location = 5)  in  vec4 dcl_Input5;
vec4 Input5;
layout(location = 6)  in  vec4 dcl_Input6;
vec4 Input6;
layout(location = 7)  in  vec4 dcl_Input7;
vec4 Input7;
layout(location = 0)  out  vec4 VtxGeoOutput0;
#define Output0 VtxGeoOutput0
layout(location = 1)  out  vec4 VtxGeoOutput1;
#define Output1 VtxGeoOutput1
layout(location = 2)  out  vec4 VtxGeoOutput2;
#define Output2 VtxGeoOutput2
layout(location = 3)  out  vec4 VtxGeoOutput3;
#define Output3 VtxGeoOutput3
layout(location = 4)  out  vec4 VtxGeoOutput4;
#define Output4 VtxGeoOutput4
#undef Output5
#define Output5 phase0_Output5
vec4 phase0_Output5;
vec4 Temp[4];
ivec4 Temp_int[4];
uvec4 Temp_uint[4];
void main()
{
    Input0 = dcl_Input0;
    Input1 = dcl_Input1;
    Input2 = dcl_Input2;
    Input4 = dcl_Input4;
    Input5 = dcl_Input5;
    Input6 = dcl_Input6;
    Input7 = dcl_Input7;
    Output0.xy = Input7.xy * Const0[36].xy + Const0[36].wz;
    Output1.xy = Input4.xy;
    Output1.zw = Input5.yx;
    Output2.xy = Input6.xy;
    Output2.zw = vec2(uintBitsToFloat(0u), uintBitsToFloat(0u));
    Temp[0].x = Const0[18].w * uintBitsToFloat(1067114824u) + Const0[30].w;
    Temp[0].y = uintBitsToFloat((Temp[0].x>=(-Temp[0].x)) ? 0xFFFFFFFFu : 0u);
    Temp[0].z = fract(abs(Temp[0].x));
    Temp[0].y = (floatBitsToInt(Temp[0]).y != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].yz = Temp[0].yx * vec2(uintBitsToFloat(1086918619u), uintBitsToFloat(1061481552u));
    Temp[0].w = uintBitsToFloat((Temp[0].z>=(-Temp[0].z)) ? 0xFFFFFFFFu : 0u);
    Temp[0].z = fract(abs(Temp[0].z));
    Temp[0].z = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].z = Temp[0].z * uintBitsToFloat(1086918619u);
    Temp[0].yz = sin(Temp[0].yz);
    Temp[0].y = Temp[0].z + Temp[0].y;
    Temp[0].y = Temp[0].y + uintBitsToFloat(1065353216u);
    Temp[1].xyz = Input0.xyz * Const0[41].xyz + Const0[40].xyz;
    Temp[0].z = Temp[1].x * Const0[33].z;
    Temp[0].y = Temp[0].y * Temp[0].z;
    Temp[0].z = Const0[34].x * Temp[1].x + Temp[0].x;
    Temp[0].z = Temp[0].z * uintBitsToFloat(1056964608u);
    Temp[0].w = uintBitsToFloat((Temp[0].z>=(-Temp[0].z)) ? 0xFFFFFFFFu : 0u);
    Temp[0].z = fract(abs(Temp[0].z));
    Temp[0].z = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].z = Temp[0].z * uintBitsToFloat(1086918619u);
    Temp[0].w = Const0[33].w * Temp[1].x + Temp[0].x;
    Temp[0].x = Temp[0].x + Const0[32].x;
    Temp[2].x = uintBitsToFloat((Temp[0].w>=(-Temp[0].w)) ? 0xFFFFFFFFu : 0u);
    Temp[0].w = fract(abs(Temp[0].w));
    Temp[0].w = (floatBitsToInt(Temp[2]).x != 0) ? Temp[0].w : (-Temp[0].w);
    Temp[0].w = Temp[0].w * uintBitsToFloat(1086918619u);
    Temp[0].zw = sin(Temp[0].zw);
    Temp[0].z = Temp[0].z + Temp[0].w;
    Temp[0].z = Temp[0].z * Const0[34].y;
    Temp[2].z = Temp[0].y * uintBitsToFloat(1045220557u) + Temp[0].z;
    Temp[0].y = Const0[30].z * Temp[1].x + (-Temp[0].x);
    Temp[0].xz = Const0[32].yw * Temp[1].xx + (-Temp[0].xx);
    Temp[0].xy = Temp[0].xy * vec2(uintBitsToFloat(1045494470u), uintBitsToFloat(1055439406u));
    Temp[0].w = uintBitsToFloat((Temp[0].y>=(-Temp[0].y)) ? 0xFFFFFFFFu : 0u);
    Temp[0].y = fract(abs(Temp[0].y));
    Temp[0].y = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].y : (-Temp[0].y);
    Temp[0].w = uintBitsToFloat((Temp[0].x>=(-Temp[0].x)) ? 0xFFFFFFFFu : 0u);
    Temp[0].x = fract(abs(Temp[0].x));
    Temp[0].x = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].x : (-Temp[0].x);
    Temp[0].xy = Temp[0].xy * vec2(uintBitsToFloat(1086918619u), uintBitsToFloat(1086918619u));
    Temp[0].xy = sin(Temp[0].xy);
    Temp[0].x = Temp[0].x + Temp[0].y;
    Temp[0].y = Temp[0].x * Const0[33].y;
    Temp[0].x = abs(Temp[0].x) * Const0[32].z;
    Temp[0].w = uintBitsToFloat((Temp[0].z>=(-Temp[0].z)) ? 0xFFFFFFFFu : 0u);
    Temp[0].z = fract(abs(Temp[0].z));
    Temp[0].z = (floatBitsToInt(Temp[0]).w != 0) ? Temp[0].z : (-Temp[0].z);
    Temp[0].z = Temp[0].z * uintBitsToFloat(1086918619u);
    Temp[0].z = sin(Temp[0].z);
    Temp[2].y = Const0[33].x * Temp[0].z + Temp[0].y;
    Temp[0].y = Const0[30].y * uintBitsToFloat(3256877056u) + uintBitsToFloat(3240099840u);
    Temp[2].x = Temp[0].x * Temp[0].y;
    Temp[0].xyz = Temp[2].xyz * Input6.xxx;
    Temp[1].w = Input0.w;
    Temp[0].w = uintBitsToFloat(0u);
    Temp[0] = Temp[0] + Temp[1];
    Temp[1] = Temp[0].yyyy * Const0[1];
    Temp[1] = Const0[0] * Temp[0].xxxx + Temp[1];
    Temp[1] = Const0[2] * Temp[0].zzzz + Temp[1];
    Temp[0] = Const0[3] * Temp[0].wwww + Temp[1];
    Temp[1] = Temp[0].yyyy * Const1[1];
    Temp[1] = Const1[0] * Temp[0].xxxx + Temp[1];
    Temp[1] = Const1[2] * Temp[0].zzzz + Temp[1];
    Temp[1] = Const1[3] * Temp[0].wwww + Temp[1];
    Temp[0].xyz = Temp[0].xyz * Const1[4].www;
    Output3 = Temp[1];
    Output5 = Temp[1];
    Temp[1].xyz = (-Temp[0].yyy) * Const0[38].xyz;
    Temp[0].xyw = Const0[37].xyz * (-Temp[0].xxx) + Temp[1].xyz;
    Temp[0].xyz = Const0[39].xyz * (-Temp[0].zzz) + Temp[0].xyw;
    Temp[1].xyz = Input1.yzx * vec3(uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u)) + vec3(uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u));
    Temp[2] = Input2 * vec4(uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u), uintBitsToFloat(1073741824u)) + vec4(uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u), uintBitsToFloat(3212836864u));
    Temp[3].xyz = Temp[1].xyz * Temp[2].zxy;
    Temp[1].xyz = Temp[2].yzx * Temp[1].yzx + (-Temp[3].xyz);
    Temp[1].xyz = Temp[2].www * Temp[1].xyz;
    Temp[3].xyz = Temp[2].yzx * Temp[1].zxy;
    Temp[3].xyz = Temp[1].yzx * Temp[2].zxy + (-Temp[3].xyz);
    Output4.y = dot(Temp[1].xyz, Temp[0].xyz);
    Temp[1].xyz = Temp[2].www * Temp[3].xyz;
    Output4.z = dot(Temp[2].xyz, Temp[0].xyz);
    Output4.x = dot(Temp[1].xyz, Temp[0].xyz);
    Output4.w = uintBitsToFloat(1065353216u);
    gl_Position = vec4(phase0_Output5);
    gl_Position.y = -gl_Position.y;
    gl_Position.z = gl_Position.z * 2.0 - gl_Position.w;
    return;
}

[traceshaders] ================================================================================
[traceshaders] glCompileShader(shader=383)
[traceshaders] glCreateProgram() = 384
[traceshaders] glAttachShader(program=384, shader=383)
[traceshaders] glLinkProgram(program=384)
SHADER KEY
  instance_divisors = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
  as_es = 0
  as_ls = 0
  export_prim_id = 0
VERT
PROPERTY NEXT_SHADER FRAG
DCL IN[0]
DCL IN[1]
DCL IN[2]
DCL IN[3]
DCL IN[4]
DCL IN[5]
DCL IN[6]
DCL OUT[0], POSITION
DCL OUT[1], GENERIC[0]
DCL OUT[2], GENERIC[1]
DCL OUT[3], GENERIC[2]
DCL OUT[4], GENERIC[3]
DCL OUT[5], GENERIC[4]
DCL CONST[1][0..45]
DCL CONST[2][0..4]
DCL TEMP[0..9], LOCAL
IMM[0] UINT32 {0, 576, 288, 480}
IMM[1] FLT32 {    0.0000,     1.2100,     6.2832,     0.7692}
IMM[2] UINT32 {4294967295, 656, 640, 528}
IMM[3] INT32 {0, 0, 0, 0}
IMM[4] FLT32 {    1.0000,     0.5000,     0.2000,   -40.0000}
IMM[5] UINT32 {544, 512, 16, 32}
IMM[6] FLT32 {    0.2041,     0.4545,   -10.0000,     2.0000}
IMM[7] UINT32 {48, 1, 64, 608}
IMM[8] UINT32 {592, 624, 0, 0}
IMM[9] FLT32 {   -1.0000,     0.0000,     0.0000,     0.0000}
  0: MAD TEMP[0].xy, IN[6].xyyy, CONST[1][36].xyyy, CONST[1][36].wzzz
  1: MOV TEMP[1].xy, IN[3].xyxx
  2: MOV TEMP[1].zw, IN[4].xxyx
  3: MOV TEMP[2].xy, IN[5].xyxx
  4: MOV TEMP[2].zw, IMM[1].xxxx
  5: MAD TEMP[3].x, CONST[1][18].wwww, IMM[1].yyyy, CONST[1][30].wwww
  6: FSGE TEMP[4].x, TEMP[3].xxxx, -TEMP[3].xxxx
  7: UIF TEMP[4].xxxx :0
  8:   MOV TEMP[4].x, IMM[2].xxxx
  9: ELSE :0
 10:   MOV TEMP[4].x, IMM[0].xxxx
 11: ENDIF
 12: MOV TEMP[3].y, TEMP[4].xxxx
 13: ABS TEMP[4].x, TEMP[3].xxxx
 14: FRC TEMP[4].x, TEMP[4].xxxx
 15: USNE TEMP[5].x, TEMP[3].yyyy, IMM[3].xxxx
 16: UIF TEMP[5].xxxx :0
 17:   MOV TEMP[5].x, TEMP[4].xxxx
 18: ELSE :0
 19:   MOV TEMP[5].x, -TEMP[4].xxxx
 20: ENDIF
 21: MOV TEMP[3].y, TEMP[5].xxxx
 22: MUL TEMP[4].xy, TEMP[3].yxxx, IMM[1].zwww
 23: FSGE TEMP[5].x, TEMP[4].yyyy, -TEMP[4].yyyy
 24: UIF TEMP[5].xxxx :0
 25:   MOV TEMP[5].x, IMM[2].xxxx
 26: ELSE :0
 27:   MOV TEMP[5].x, IMM[0].xxxx
 28: ENDIF
 29: MOV TEMP[3].w, TEMP[5].xxxx
 30: ABS TEMP[5].x, TEMP[4].yyyy
 31: FRC TEMP[5].x, TEMP[5].xxxx
 32: USNE TEMP[6].x, TEMP[3].wwww, IMM[3].xxxx
 33: UIF TEMP[6].xxxx :0
 34:   MOV TEMP[6].x, TEMP[5].xxxx
 35: ELSE :0
 36:   MOV TEMP[6].x, -TEMP[5].xxxx
 37: ENDIF
 38: MUL TEMP[5].x, TEMP[6].xxxx, IMM[1].zzzz
 39: SIN TEMP[4].x, TEMP[4].xxxx
 40: SIN TEMP[4].y, TEMP[5].xxxx
 41: ADD TEMP[4].x, TEMP[4].yyyy, TEMP[4].xxxx
 42: ADD TEMP[4].x, TEMP[4].xxxx, IMM[4].xxxx
 43: MAD TEMP[5].xyz, IN[0].xyzz, CONST[1][41].xyzz, CONST[1][40].xyzz
 44: MUL TEMP[6].x, TEMP[5].xxxx, CONST[1][33].zzzz
 45: MUL TEMP[4].x, TEMP[4].xxxx, TEMP[6].xxxx
 46: MAD TEMP[6].x, CONST[1][34].xxxx, TEMP[5].xxxx, TEMP[3].xxxx
 47: MUL TEMP[6].x, TEMP[6].xxxx, IMM[4].yyyy
 48: FSGE TEMP[7].x, TEMP[6].xxxx, -TEMP[6].xxxx
 49: UIF TEMP[7].xxxx :0
 50:   MOV TEMP[7].x, IMM[2].xxxx
 51: ELSE :0
 52:   MOV TEMP[7].x, IMM[0].xxxx
 53: ENDIF
 54: MOV TEMP[3].w, TEMP[7].xxxx
 55: ABS TEMP[6].x, TEMP[6].xxxx
 56: FRC TEMP[6].x, TEMP[6].xxxx
 57: USNE TEMP[7].x, TEMP[3].wwww, IMM[3].xxxx
 58: UIF TEMP[7].xxxx :0
 59:   MOV TEMP[7].x, TEMP[6].xxxx
 60: ELSE :0
 61:   MOV TEMP[7].x, -TEMP[6].xxxx
 62: ENDIF
 63: MUL TEMP[6].x, TEMP[7].xxxx, IMM[1].zzzz
 64: MAD TEMP[7].x, CONST[1][33].wwww, TEMP[5].xxxx, TEMP[3].xxxx
 65: ADD TEMP[3].x, TEMP[3].xxxx, CONST[1][32].xxxx
 66: FSGE TEMP[8].x, TEMP[7].xxxx, -TEMP[7].xxxx
 67: UIF TEMP[8].xxxx :0
 68:   MOV TEMP[8].x, IMM[2].xxxx
 69: ELSE :0
 70:   MOV TEMP[8].x, IMM[0].xxxx
 71: ENDIF
 72: MOV TEMP[8].x, TEMP[8].xxxx
 73: ABS TEMP[7].x, TEMP[7].xxxx
 74: FRC TEMP[7].x, TEMP[7].xxxx
 75: USNE TEMP[9].x, TEMP[8].xxxx, IMM[3].xxxx
 76: UIF TEMP[9].xxxx :0
 77:   MOV TEMP[9].x, TEMP[7].xxxx
 78: ELSE :0
 79:   MOV TEMP[9].x, -TEMP[7].xxxx
 80: ENDIF
 81: MUL TEMP[7].x, TEMP[9].xxxx, IMM[1].zzzz
 82: SIN TEMP[6].x, TEMP[6].xxxx
 83: SIN TEMP[6].y, TEMP[7].xxxx
 84: ADD TEMP[6].x, TEMP[6].xxxx, TEMP[6].yyyy
 85: MUL TEMP[6].x, TEMP[6].xxxx, CONST[1][34].yyyy
 86: MAD TEMP[4].x, TEMP[4].xxxx, IMM[4].zzzz, TEMP[6].xxxx
 87: MOV TEMP[8].z, TEMP[4].xxxx
 88: MAD TEMP[4].x, CONST[1][30].zzzz, TEMP[5].xxxx, -TEMP[3].xxxx
 89: MOV TEMP[3].y, TEMP[4].xxxx
 90: MAD TEMP[4].xy, CONST[1][32].ywww, TEMP[5].xxxx, -TEMP[3].xxxx
 91: MOV TEMP[3].x, TEMP[4].xxyx
 92: MUL TEMP[3].xy, TEMP[3].xyyy, IMM[6].xyyy
 93: FSGE TEMP[6].x, TEMP[3].yyyy, -TEMP[3].yyyy
 94: UIF TEMP[6].xxxx :0
 95:   MOV TEMP[6].x, IMM[2].xxxx
 96: ELSE :0
 97:   MOV TEMP[6].x, IMM[0].xxxx
 98: ENDIF
 99: MOV TEMP[3].w, TEMP[6].xxxx
100: ABS TEMP[6].x, TEMP[3].yyyy
101: FRC TEMP[6].x, TEMP[6].xxxx
102: USNE TEMP[7].x, TEMP[3].wwww, IMM[3].xxxx
103: UIF TEMP[7].xxxx :0
104:   MOV TEMP[7].x, TEMP[6].xxxx
105: ELSE :0
106:   MOV TEMP[7].x, -TEMP[6].xxxx
107: ENDIF
108: MOV TEMP[3].y, TEMP[7].xxxx
109: FSGE TEMP[6].x, TEMP[3].xxxx, -TEMP[3].xxxx
110: UIF TEMP[6].xxxx :0
111:   MOV TEMP[6].x, IMM[2].xxxx
112: ELSE :0
113:   MOV TEMP[6].x, IMM[0].xxxx
114: ENDIF
115: MOV TEMP[3].w, TEMP[6].xxxx
116: ABS TEMP[6].x, TEMP[3].xxxx
117: FRC TEMP[3].x, TEMP[6].xxxx
118: USNE TEMP[6].x, TEMP[3].wwww, IMM[3].xxxx
119: UIF TEMP[6].xxxx :0
120:   MOV TEMP[6].x, TEMP[3].xxxx
121: ELSE :0
122:   MOV TEMP[6].x, -TEMP[3].xxxx
123: ENDIF
124: MOV TEMP[3].x, TEMP[6].xxxx
125: MUL TEMP[3].xy, TEMP[3].xyyy, IMM[1].zzzz
126: SIN TEMP[6].x, TEMP[3].xxxx
127: SIN TEMP[6].y, TEMP[3].yyyy
128: ADD TEMP[3].x, TEMP[6].xxxx, TEMP[6].yyyy
129: MUL TEMP[6].x, TEMP[3].xxxx, CONST[1][33].yyyy
130: ABS TEMP[7].x, TEMP[3].xxxx
131: MUL TEMP[3].x, TEMP[7].xxxx, CONST[1][32].zzzz
132: FSGE TEMP[7].x, TEMP[4].yyyy, -TEMP[4].yyyy
133: UIF TEMP[7].xxxx :0
134:   MOV TEMP[7].x, IMM[2].xxxx
135: ELSE :0
136:   MOV TEMP[7].x, IMM[0].xxxx
137: ENDIF
138: MOV TEMP[3].w, TEMP[7].xxxx
139: ABS TEMP[4].x, TEMP[4].yyyy
140: FRC TEMP[4].x, TEMP[4].xxxx
141: USNE TEMP[7].x, TEMP[3].wwww, IMM[3].xxxx
142: UIF TEMP[7].xxxx :0
143:   MOV TEMP[7].x, TEMP[4].xxxx
144: ELSE :0
145:   MOV TEMP[7].x, -TEMP[4].xxxx
146: ENDIF
147: MUL TEMP[4].x, TEMP[7].xxxx, IMM[1].zzzz
148: SIN TEMP[4].x, TEMP[4].xxxx
149: MAD TEMP[4].x, CONST[1][33].xxxx, TEMP[4].xxxx, TEMP[6].xxxx
150: MOV TEMP[8].y, TEMP[4].xxxx
151: MAD TEMP[4].x, CONST[1][30].yyyy, IMM[4].wwww, IMM[6].zzzz
152: MUL TEMP[8].x, TEMP[3].xxxx, TEMP[4].xxxx
153: MUL TEMP[3].xyz, TEMP[8].xyzz, IN[5].xxxx
154: MOV TEMP[5].w, IN[0].wwww
155: MOV TEMP[3].w, IMM[1].xxxx
156: ADD TEMP[3], TEMP[3], TEMP[5]
157: MUL TEMP[5], TEMP[3].yyyy, CONST[1][1]
158: MAD TEMP[5], CONST[1][0], TEMP[3].xxxx, TEMP[5]
159: MAD TEMP[5], CONST[1][2], TEMP[3].zzzz, TEMP[5]
160: MAD TEMP[3], CONST[1][3], TEMP[3].wwww, TEMP[5]
161: MUL TEMP[5], TEMP[3].yyyy, CONST[2][1]
162: MAD TEMP[5], CONST[2][0], TEMP[3].xxxx, TEMP[5]
163: MAD TEMP[5], CONST[2][2], TEMP[3].zzzz, TEMP[5]
164: MAD TEMP[5], CONST[2][3], TEMP[3].wwww, TEMP[5]
165: MUL TEMP[3].xyz, TEMP[3].xyzz, CONST[2][4].wwww
166: MOV TEMP[4], TEMP[5]
167: MOV TEMP[6], TEMP[5]
168: MUL TEMP[5].xyz, -TEMP[3].yyyy, CONST[1][38].xyzz
169: MAD TEMP[7].xyz, CONST[1][37].xyzz, -TEMP[3].xxxx, TEMP[5].xyzz
170: MAD TEMP[3].xyz, CONST[1][39].xyzz, -TEMP[3].zzzz, TEMP[7].xyzz
171: MAD TEMP[5].xyz, IN[1].yzxx, IMM[6].wwww, IMM[9].xxxx
172: MAD TEMP[8], IN[2], IMM[6].wwww, IMM[9].xxxx
173: MUL TEMP[7].xyz, TEMP[5].xyzz, TEMP[8].zxyy
174: MAD TEMP[5].xyz, TEMP[8].yzxx, TEMP[5].yzxx, -TEMP[7].xyzz
175: MUL TEMP[5].xyz, TEMP[8].wwww, TEMP[5].xyzz
176: MUL TEMP[7].xyz, TEMP[8].yzxx, TEMP[5].zxyy
177: MAD TEMP[7].xyz, TEMP[5].yzxx, TEMP[8].zxyy, -TEMP[7].xyzz
178: DP3 TEMP[9].x, TEMP[5].xyzz, TEMP[3].xyzz
179: MOV TEMP[9].y, TEMP[9].xxxx
180: MUL TEMP[5].xyz, TEMP[8].wwww, TEMP[7].xyzz
181: DP3 TEMP[7].x, TEMP[8].xyzz, TEMP[3].xyzz
182: MOV TEMP[9].z, TEMP[7].xxxx
183: DP3 TEMP[9].x, TEMP[5].xyzz, TEMP[3].xyzz
184: MOV TEMP[9].w, IMM[4].xxxx
185: MOV TEMP[3].xw, TEMP[6].xxxw
186: MOV TEMP[3].y, -TEMP[6].yyyy
187: MAD TEMP[5].x, TEMP[6].zzzz, IMM[6].wwww, -TEMP[5].wwww
188: MOV TEMP[3].z, TEMP[5].xxxx
189: MOV OUT[1], TEMP[0]
190: MOV OUT[2], TEMP[1]
191: MOV OUT[3], TEMP[2]
192: MOV OUT[4], TEMP[4]
193: MOV OUT[5], TEMP[9]
194: MOV OUT[0], TEMP[3]
195: END
radeonsi: Compiling shader 209
TGSI shader LLVM IR:

; ModuleID = 'tgsi'
source_filename = "tgsi"
target triple = "amdgcn--"

define amdgpu_vs <{ float, float, float }> @main([17 x <16 x i8>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, [32 x <8 x i32>] addrspace(2)* byval, [16 x <8 x i32>] addrspace(2)* byval, [16 x <4 x i32>] addrspace(2)* byval, [16 x <16 x i8>] addrspace(2)* byval, i32 inreg, i32 inreg, i32 inreg, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32) {
main_body:
  %20 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 1, !amdgpu.uniform !0
  %21 = load <16 x i8>, <16 x i8> addrspace(2)* %20, align 16, !tbaa !1
  %22 = call float @llvm.SI.load.const(<16 x i8> %21, i32 0)
  %23 = call float @llvm.SI.load.const(<16 x i8> %21, i32 4)
  %24 = call float @llvm.SI.load.const(<16 x i8> %21, i32 8)
  %25 = call float @llvm.SI.load.const(<16 x i8> %21, i32 12)
  %26 = call float @llvm.SI.load.const(<16 x i8> %21, i32 16)
  %27 = call float @llvm.SI.load.const(<16 x i8> %21, i32 20)
  %28 = call float @llvm.SI.load.const(<16 x i8> %21, i32 24)
  %29 = call float @llvm.SI.load.const(<16 x i8> %21, i32 28)
  %30 = call float @llvm.SI.load.const(<16 x i8> %21, i32 32)
  %31 = call float @llvm.SI.load.const(<16 x i8> %21, i32 36)
  %32 = call float @llvm.SI.load.const(<16 x i8> %21, i32 40)
  %33 = call float @llvm.SI.load.const(<16 x i8> %21, i32 44)
  %34 = call float @llvm.SI.load.const(<16 x i8> %21, i32 48)
  %35 = call float @llvm.SI.load.const(<16 x i8> %21, i32 52)
  %36 = call float @llvm.SI.load.const(<16 x i8> %21, i32 56)
  %37 = call float @llvm.SI.load.const(<16 x i8> %21, i32 60)
  %38 = call float @llvm.SI.load.const(<16 x i8> %21, i32 300)
  %39 = call float @llvm.SI.load.const(<16 x i8> %21, i32 484)
  %40 = call float @llvm.SI.load.const(<16 x i8> %21, i32 488)
  %41 = call float @llvm.SI.load.const(<16 x i8> %21, i32 492)
  %42 = call float @llvm.SI.load.const(<16 x i8> %21, i32 512)
  %43 = call float @llvm.SI.load.const(<16 x i8> %21, i32 516)
  %44 = call float @llvm.SI.load.const(<16 x i8> %21, i32 520)
  %45 = call float @llvm.SI.load.const(<16 x i8> %21, i32 524)
  %46 = call float @llvm.SI.load.const(<16 x i8> %21, i32 528)
  %47 = call float @llvm.SI.load.const(<16 x i8> %21, i32 532)
  %48 = call float @llvm.SI.load.const(<16 x i8> %21, i32 536)
  %49 = call float @llvm.SI.load.const(<16 x i8> %21, i32 540)
  %50 = call float @llvm.SI.load.const(<16 x i8> %21, i32 544)
  %51 = call float @llvm.SI.load.const(<16 x i8> %21, i32 548)
  %52 = call float @llvm.SI.load.const(<16 x i8> %21, i32 576)
  %53 = call float @llvm.SI.load.const(<16 x i8> %21, i32 580)
  %54 = call float @llvm.SI.load.const(<16 x i8> %21, i32 584)
  %55 = call float @llvm.SI.load.const(<16 x i8> %21, i32 588)
  %56 = call float @llvm.SI.load.const(<16 x i8> %21, i32 592)
  %57 = call float @llvm.SI.load.const(<16 x i8> %21, i32 596)
  %58 = call float @llvm.SI.load.const(<16 x i8> %21, i32 600)
  %59 = call float @llvm.SI.load.const(<16 x i8> %21, i32 608)
  %60 = call float @llvm.SI.load.const(<16 x i8> %21, i32 612)
  %61 = call float @llvm.SI.load.const(<16 x i8> %21, i32 616)
  %62 = call float @llvm.SI.load.const(<16 x i8> %21, i32 624)
  %63 = call float @llvm.SI.load.const(<16 x i8> %21, i32 628)
  %64 = call float @llvm.SI.load.const(<16 x i8> %21, i32 632)
  %65 = call float @llvm.SI.load.const(<16 x i8> %21, i32 640)
  %66 = call float @llvm.SI.load.const(<16 x i8> %21, i32 644)
  %67 = call float @llvm.SI.load.const(<16 x i8> %21, i32 648)
  %68 = call float @llvm.SI.load.const(<16 x i8> %21, i32 656)
  %69 = call float @llvm.SI.load.const(<16 x i8> %21, i32 660)
  %70 = call float @llvm.SI.load.const(<16 x i8> %21, i32 664)
  %71 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %1, i64 0, i64 2, !amdgpu.uniform !0
  %72 = load <16 x i8>, <16 x i8> addrspace(2)* %71, align 16, !tbaa !1
  %73 = call float @llvm.SI.load.const(<16 x i8> %72, i32 0)
  %74 = call float @llvm.SI.load.const(<16 x i8> %72, i32 4)
  %75 = call float @llvm.SI.load.const(<16 x i8> %72, i32 8)
  %76 = call float @llvm.SI.load.const(<16 x i8> %72, i32 12)
  %77 = call float @llvm.SI.load.const(<16 x i8> %72, i32 16)
  %78 = call float @llvm.SI.load.const(<16 x i8> %72, i32 20)
  %79 = call float @llvm.SI.load.const(<16 x i8> %72, i32 24)
  %80 = call float @llvm.SI.load.const(<16 x i8> %72, i32 28)
  %81 = call float @llvm.SI.load.const(<16 x i8> %72, i32 32)
  %82 = call float @llvm.SI.load.const(<16 x i8> %72, i32 36)
  %83 = call float @llvm.SI.load.const(<16 x i8> %72, i32 40)
  %84 = call float @llvm.SI.load.const(<16 x i8> %72, i32 44)
  %85 = call float @llvm.SI.load.const(<16 x i8> %72, i32 48)
  %86 = call float @llvm.SI.load.const(<16 x i8> %72, i32 52)
  %87 = call float @llvm.SI.load.const(<16 x i8> %72, i32 56)
  %88 = call float @llvm.SI.load.const(<16 x i8> %72, i32 60)
  %89 = call float @llvm.SI.load.const(<16 x i8> %72, i32 76)
  %90 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 0, !amdgpu.uniform !0
  %91 = load <16 x i8>, <16 x i8> addrspace(2)* %90, align 16, !tbaa !1
  %92 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %91, i32 0, i32 %13)
  %93 = extractelement <4 x float> %92, i32 0
  %94 = extractelement <4 x float> %92, i32 1
  %95 = extractelement <4 x float> %92, i32 2
  %96 = extractelement <4 x float> %92, i32 3
  %97 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 1, !amdgpu.uniform !0
  %98 = load <16 x i8>, <16 x i8> addrspace(2)* %97, align 16, !tbaa !1
  %99 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %98, i32 0, i32 %14)
  %100 = extractelement <4 x float> %99, i32 0
  %101 = extractelement <4 x float> %99, i32 1
  %102 = extractelement <4 x float> %99, i32 2
  %103 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 2, !amdgpu.uniform !0
  %104 = load <16 x i8>, <16 x i8> addrspace(2)* %103, align 16, !tbaa !1
  %105 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %104, i32 0, i32 %15)
  %106 = extractelement <4 x float> %105, i32 0
  %107 = extractelement <4 x float> %105, i32 1
  %108 = extractelement <4 x float> %105, i32 2
  %109 = extractelement <4 x float> %105, i32 3
  %110 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 3, !amdgpu.uniform !0
  %111 = load <16 x i8>, <16 x i8> addrspace(2)* %110, align 16, !tbaa !1
  %112 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %111, i32 0, i32 %16)
  %113 = extractelement <4 x float> %112, i32 0
  %114 = extractelement <4 x float> %112, i32 1
  %115 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 4, !amdgpu.uniform !0
  %116 = load <16 x i8>, <16 x i8> addrspace(2)* %115, align 16, !tbaa !1
  %117 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %116, i32 0, i32 %17)
  %118 = extractelement <4 x float> %117, i32 0
  %119 = extractelement <4 x float> %117, i32 1
  %120 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 5, !amdgpu.uniform !0
  %121 = load <16 x i8>, <16 x i8> addrspace(2)* %120, align 16, !tbaa !1
  %122 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %121, i32 0, i32 %18)
  %123 = extractelement <4 x float> %122, i32 0
  %124 = extractelement <4 x float> %122, i32 1
  %125 = getelementptr [16 x <16 x i8>], [16 x <16 x i8>] addrspace(2)* %5, i64 0, i64 6, !amdgpu.uniform !0
  %126 = load <16 x i8>, <16 x i8> addrspace(2)* %125, align 16, !tbaa !1
  %127 = call <4 x float> @llvm.SI.vs.load.input(<16 x i8> %126, i32 0, i32 %19)
  %128 = extractelement <4 x float> %127, i32 0
  %129 = extractelement <4 x float> %127, i32 1
  %130 = fmul float %128, %52
  %131 = fadd float %130, %55
  %132 = fmul float %129, %53
  %133 = fadd float %132, %54
  %134 = fmul float %38, 0x3FF35C2900000000
  %135 = fadd float %134, %41
  %136 = fsub float -0.000000e+00, %135
  %137 = fcmp oge float %135, %136
  %138 = call float @llvm.fabs.f32(float %135)
  %139 = call float @llvm.floor.f32(float %138)
  %140 = fsub float %138, %139
  %141 = fsub float -0.000000e+00, %140
  %temp20.0 = select i1 %137, float %140, float %141
  %142 = fmul float %temp20.0, 0x401921FB60000000
  %143 = fmul float %135, 0x3FE89D8A00000000
  %144 = fsub float -0.000000e+00, %143
  %145 = fcmp oge float %143, %144
  %146 = call float @llvm.fabs.f32(float %143)
  %147 = call float @llvm.floor.f32(float %146)
  %148 = fsub float %146, %147
  %149 = fsub float -0.000000e+00, %148
  %temp24.0 = select i1 %145, float %148, float %149
  %150 = fmul float %temp24.0, 0x401921FB60000000
  %151 = call float @llvm.sin.f32(float %142)
  %152 = call float @llvm.sin.f32(float %150)
  %153 = fadd float %152, %151
  %154 = fadd float %153, 1.000000e+00
  %155 = fmul float %93, %68
  %156 = fadd float %155, %65
  %157 = fmul float %94, %69
  %158 = fadd float %157, %66
  %159 = fmul float %95, %70
  %160 = fadd float %159, %67
  %161 = fmul float %156, %48
  %162 = fmul float %154, %161
  %163 = fmul float %50, %156
  %164 = fadd float %163, %135
  %165 = fmul float %164, 5.000000e-01
  %166 = fsub float -0.000000e+00, %165
  %167 = fcmp oge float %165, %166
  %168 = call float @llvm.fabs.f32(float %165)
  %169 = call float @llvm.floor.f32(float %168)
  %170 = fsub float %168, %169
  %171 = fsub float -0.000000e+00, %170
  %temp28.1 = select i1 %167, float %170, float %171
  %172 = fmul float %temp28.1, 0x401921FB60000000
  %173 = fmul float %49, %156
  %174 = fadd float %173, %135
  %175 = fadd float %135, %42
  %176 = fsub float -0.000000e+00, %174
  %177 = fcmp oge float %174, %176
  %178 = call float @llvm.fabs.f32(float %174)
  %179 = call float @llvm.floor.f32(float %178)
  %180 = fsub float %178, %179
  %181 = fsub float -0.000000e+00, %180
  %temp36.0 = select i1 %177, float %180, float %181
  %182 = fmul float %temp36.0, 0x401921FB60000000
  %183 = call float @llvm.sin.f32(float %172)
  %184 = call float @llvm.sin.f32(float %182)
  %185 = fadd float %183, %184
  %186 = fmul float %185, %51
  %187 = fmul float %162, 0x3FC99999A0000000
  %188 = fadd float %187, %186
  %189 = fmul float %40, %156
  %190 = fsub float %189, %175
  %191 = fmul float %43, %156
  %192 = fsub float %191, %175
  %193 = fmul float %45, %156
  %194 = fsub float %193, %175
  %195 = fmul float %192, 0x3FCA1F58C0000000
  %196 = fmul float %190, 0x3FDD1745C0000000
  %197 = fsub float -0.000000e+00, %196
  %198 = fcmp oge float %196, %197
  %199 = call float @llvm.fabs.f32(float %196)
  %200 = call float @llvm.floor.f32(float %199)
  %201 = fsub float %199, %200
  %202 = fsub float -0.000000e+00, %201
  %temp28.2 = select i1 %198, float %201, float %202
  %203 = fsub float -0.000000e+00, %195
  %204 = fcmp oge float %195, %203
  %205 = call float @llvm.fabs.f32(float %195)
  %206 = call float @llvm.floor.f32(float %205)
  %207 = fsub float %205, %206
  %208 = fsub float -0.000000e+00, %207
  %temp24.3 = select i1 %204, float %207, float %208
  %209 = fmul float %temp24.3, 0x401921FB60000000
  %210 = fmul float %temp28.2, 0x401921FB60000000
  %211 = call float @llvm.sin.f32(float %209)
  %212 = call float @llvm.sin.f32(float %210)
  %213 = fadd float %211, %212
  %214 = fmul float %213, %47
  %215 = call float @llvm.fabs.f32(float %213)
  %216 = fmul float %215, %44
  %217 = fsub float -0.000000e+00, %194
  %218 = fcmp oge float %194, %217
  %219 = call float @llvm.fabs.f32(float %194)
  %220 = call float @llvm.floor.f32(float %219)
  %221 = fsub float %219, %220
  %222 = fsub float -0.000000e+00, %221
  %temp28.4 = select i1 %218, float %221, float %222
  %223 = fmul float %temp28.4, 0x401921FB60000000
  %224 = call float @llvm.sin.f32(float %223)
  %225 = fmul float %46, %224
  %226 = fadd float %225, %214
  %227 = fmul float %39, -4.000000e+01
  %228 = fadd float %227, -1.000000e+01
  %229 = fmul float %216, %228
  %230 = fmul float %229, %123
  %231 = fmul float %226, %123
  %232 = fmul float %188, %123
  %233 = fadd float %230, %156
  %234 = fadd float %231, %158
  %235 = fadd float %232, %160
  %236 = fadd float %96, 0.000000e+00
  %237 = fmul float %234, %26
  %238 = fmul float %234, %27
  %239 = fmul float %234, %28
  %240 = fmul float %234, %29
  %241 = fmul float %22, %233
  %242 = fadd float %241, %237
  %243 = fmul float %23, %233
  %244 = fadd float %243, %238
  %245 = fmul float %24, %233
  %246 = fadd float %245, %239
  %247 = fmul float %25, %233
  %248 = fadd float %247, %240
  %249 = fmul float %30, %235
  %250 = fadd float %249, %242
  %251 = fmul float %31, %235
  %252 = fadd float %251, %244
  %253 = fmul float %32, %235
  %254 = fadd float %253, %246
  %255 = fmul float %33, %235
  %256 = fadd float %255, %248
  %257 = fmul float %34, %236
  %258 = fadd float %257, %250
  %259 = fmul float %35, %236
  %260 = fadd float %259, %252
  %261 = fmul float %36, %236
  %262 = fadd float %261, %254
  %263 = fmul float %37, %236
  %264 = fadd float %263, %256
  %265 = fmul float %260, %77
  %266 = fmul float %260, %78
  %267 = fmul float %260, %79
  %268 = fmul float %260, %80
  %269 = fmul float %73, %258
  %270 = fadd float %269, %265
  %271 = fmul float %74, %258
  %272 = fadd float %271, %266
  %273 = fmul float %75, %258
  %274 = fadd float %273, %267
  %275 = fmul float %76, %258
  %276 = fadd float %275, %268
  %277 = fmul float %81, %262
  %278 = fadd float %277, %270
  %279 = fmul float %82, %262
  %280 = fadd float %279, %272
  %281 = fmul float %83, %262
  %282 = fadd float %281, %274
  %283 = fmul float %84, %262
  %284 = fadd float %283, %276
  %285 = fmul float %85, %264
  %286 = fadd float %285, %278
  %287 = fmul float %86, %264
  %288 = fadd float %287, %280
  %289 = fmul float %87, %264
  %290 = fadd float %289, %282
  %291 = fmul float %88, %264
  %292 = fadd float %291, %284
  %293 = fmul float %258, %89
  %294 = fmul float %260, %89
  %295 = fmul float %262, %89
  %296 = fmul float %294, %59
  %297 = fsub float -0.000000e+00, %296
  %298 = fmul float %294, %60
  %299 = fsub float -0.000000e+00, %298
  %300 = fmul float %294, %61
  %301 = fsub float -0.000000e+00, %300
  %302 = fmul float %293, %56
  %303 = fsub float %297, %302
  %304 = fmul float %293, %57
  %305 = fsub float %299, %304
  %306 = fmul float %293, %58
  %307 = fsub float %301, %306
  %308 = fmul float %295, %62
  %309 = fsub float %303, %308
  %310 = fmul float %295, %63
  %311 = fsub float %305, %310
  %312 = fmul float %295, %64
  %313 = fsub float %307, %312
  %314 = fmul float %101, 2.000000e+00
  %315 = fadd float %314, -1.000000e+00
  %316 = fmul float %102, 2.000000e+00
  %317 = fadd float %316, -1.000000e+00
  %318 = fmul float %100, 2.000000e+00
  %319 = fadd float %318, -1.000000e+00
  %320 = fmul float %106, 2.000000e+00
  %321 = fadd float %320, -1.000000e+00
  %322 = fmul float %107, 2.000000e+00
  %323 = fadd float %322, -1.000000e+00
  %324 = fmul float %108, 2.000000e+00
  %325 = fadd float %324, -1.000000e+00
  %326 = fmul float %109, 2.000000e+00
  %327 = fadd float %326, -1.000000e+00
  %328 = fmul float %315, %325
  %329 = fmul float %317, %321
  %330 = fmul float %319, %323
  %331 = fmul float %323, %317
  %332 = fsub float %331, %328
  %333 = fmul float %325, %319
  %334 = fsub float %333, %329
  %335 = fmul float %321, %315
  %336 = fsub float %335, %330
  %337 = fmul float %327, %332
  %338 = fmul float %327, %334
  %339 = fmul float %327, %336
  %340 = fmul float %323, %339
  %341 = fmul float %325, %337
  %342 = fmul float %321, %338
  %343 = fmul float %338, %325
  %344 = fsub float %343, %340
  %345 = fmul float %339, %321
  %346 = fsub float %345, %341
  %347 = fmul float %337, %323
  %348 = fsub float %347, %342
  %349 = fmul float %337, %309
  %350 = fmul float %338, %311
  %351 = fadd float %350, %349
  %352 = fmul float %339, %313
  %353 = fadd float %351, %352
  %354 = fmul float %327, %344
  %355 = fmul float %327, %346
  %356 = fmul float %327, %348
  %357 = fmul float %321, %309
  %358 = fmul float %323, %311
  %359 = fadd float %358, %357
  %360 = fmul float %325, %313
  %361 = fadd float %359, %360
  %362 = fmul float %354, %309
  %363 = fmul float %355, %311
  %364 = fadd float %363, %362
  %365 = fmul float %356, %313
  %366 = fadd float %364, %365
  %367 = fsub float -0.000000e+00, %288
  %368 = fmul float %290, 2.000000e+00
  %369 = fsub float %368, %292
  %370 = bitcast i32 %11 to float
  %371 = insertvalue <{ float, float, float }> undef, float %370, 2
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 32, i32 0, float %131, float %133, float undef, float undef)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 33, i32 0, float %113, float %114, float %119, float %118)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 34, i32 0, float %123, float %124, float 0.000000e+00, float 0.000000e+00)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 35, i32 0, float %286, float %288, float %290, float %292)
  call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 36, i32 0, float %366, float %353, float %361, float 1.000000e+00)
  call void @llvm.SI.export(i32 15, i32 0, i32 1, i32 12, i32 0, float %286, float %367, float %369, float %292)
  ret <{ float, float, float }> %371
}

; Function Attrs: nounwind readnone
declare float @llvm.SI.load.const(<16 x i8>, i32) #0

; Function Attrs: nounwind readnone
declare <4 x float> @llvm.SI.vs.load.input(<16 x i8>, i32, i32) #0

; Function Attrs: nounwind readnone
declare float @llvm.fabs.f32(float) #0

; Function Attrs: nounwind readnone
declare float @llvm.floor.f32(float) #0

; Function Attrs: nounwind readnone
declare float @llvm.sin.f32(float) #0

declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)

attributes #0 = { nounwind readnone }

!0 = !{}
!1 = !{!"const", null, i32 1}

[traceshaders] glDetachShader(program=384, shader=383)
[traceshaders] glDeleteShader(shader=383)
[...]
[traceshaders] glGenProgramPipelines():
[traceshaders]  - pipeline=15
[traceshaders] glUseProgramStages(pipeline=15, stages=vert, program=74)
[...]
[traceshaders] glBindProgramPipeline(pipeline=15)
radeonsi: Compiling shader 389
Vertex Shader Prolog LLVM IR:

; ModuleID = 'tgsi'
source_filename = "tgsi"
target triple = "amdgcn--"

define amdgpu_vs <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> @main(i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32 inreg, i32, i32, i32, i32) {
main_body:
  %19 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> undef, i32 %0, 0
  %20 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %19, i32 %1, 1
  %21 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %20, i32 %2, 2
  %22 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %21, i32 %3, 3
  %23 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %22, i32 %4, 4
  %24 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %23, i32 %5, 5
  %25 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %24, i32 %6, 6
  %26 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %25, i32 %7, 7
  %27 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %26, i32 %8, 8
  %28 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %27, i32 %9, 9
  %29 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %28, i32 %10, 10
  %30 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %29, i32 %11, 11
  %31 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %30, i32 %12, 12
  %32 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %31, i32 %13, 13
  %33 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %32, i32 %14, 14
  %34 = bitcast i32 %15 to float
  %35 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %33, float %34, 15
  %36 = bitcast i32 %16 to float
  %37 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %35, float %36, 16
  %38 = bitcast i32 %17 to float
  %39 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %37, float %38, 17
  %40 = bitcast i32 %18 to float
  %41 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %39, float %40, 18
  %42 = add i32 %15, %12
  %43 = bitcast i32 %42 to float
  %44 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %41, float %43, 19
  %45 = add i32 %15, %12
  %46 = bitcast i32 %45 to float
  %47 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %44, float %46, 20
  %48 = add i32 %15, %12
  %49 = bitcast i32 %48 to float
  %50 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %47, float %49, 21
  %51 = add i32 %15, %12
  %52 = bitcast i32 %51 to float
  %53 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %50, float %52, 22
  %54 = add i32 %15, %12
  %55 = bitcast i32 %54 to float
  %56 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %53, float %55, 23
  %57 = add i32 %15, %12
  %58 = bitcast i32 %57 to float
  %59 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %56, float %58, 24
  %60 = add i32 %15, %12
  %61 = bitcast i32 %60 to float
  %62 = insertvalue <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %59, float %61, 25
  ret <{ i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, float, float, float, float, float, float, float, float, float, float }> %62
}


Vertex Shader as VS:
Shader prolog disassembly:
	v_add_i32_e32 v4, vcc, s12, v0 ; 3208000C
	v_mov_b32_e32 v5, v4           ; 7E0A0304
	v_mov_b32_e32 v6, v4           ; 7E0C0304
	v_mov_b32_e32 v7, v4           ; 7E0E0304
	v_mov_b32_e32 v8, v4           ; 7E100304
	v_mov_b32_e32 v9, v4           ; 7E120304
	v_mov_b32_e32 v10, v4          ; 7E140304
Shader main disassembly:
	s_load_dwordx4 s[4:7], s[10:11], 0x0                    ; C00A0105 00000000
	s_load_dwordx4 s[12:15], s[10:11], 0x10                 ; C00A0305 00000010
	s_load_dwordx4 s[16:19], s[10:11], 0x50                 ; C00A0405 00000050
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	buffer_load_format_xyzw v[17:20], v4, s[4:7], 0 idxen   ; E00C2000 80011104
	s_load_dwordx4 s[4:7], s[10:11], 0x20                   ; C00A0105 00000020
	buffer_load_format_xyzw v[25:28], v5, s[12:15], 0 idxen ; E00C2000 80031905
	s_load_dwordx4 s[12:15], s[10:11], 0x30                 ; C00A0305 00000030
	s_waitcnt vmcnt(0) lgkmcnt(0)                           ; BF8C0070
	buffer_load_format_xyzw v[28:31], v6, s[4:7], 0 idxen   ; E00C2000 80011C06
	s_load_dwordx4 s[4:7], s[10:11], 0x40                   ; C00A0105 00000040
	s_load_dwordx4 s[8:11], s[10:11], 0x60                  ; C00A0205 00000060
	buffer_load_format_xyzw v[3:6], v7, s[12:15], 0 idxen   ; E00C2000 80030307
	s_waitcnt vmcnt(0) lgkmcnt(0)                           ; BF8C0070
	buffer_load_format_xyzw v[5:8], v8, s[4:7], 0 idxen     ; E00C2000 80010508
	s_nop 0                                                 ; BF800000
	buffer_load_format_xyzw v[11:14], v9, s[16:19], 0 idxen ; E00C2000 80040B09
	s_nop 0                                                 ; BF800000
	buffer_load_format_xyzw v[21:24], v10, s[8:11], 0 idxen ; E00C2000 8002150A
	s_load_dwordx4 s[16:19], s[2:3], 0x10                   ; C00A0401 00000010
	s_nop 0                                                 ; BF800000
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	s_buffer_load_dword s5, s[16:19], 0x10                  ; C0220148 00000010
	s_buffer_load_dword s4, s[16:19], 0x0                   ; C0220108 00000000
	s_buffer_load_dword s0, s[16:19], 0x20                  ; C0220008 00000020
	s_waitcnt vmcnt(2)                                      ; BF8C0F72
	v_add_f32_e32 v7, v26, v26                              ; 020E351A
	v_add_f32_e32 v0, v25, v25                              ; 02003319
	s_waitcnt vmcnt(1)                                      ; BF8C0F71
	v_mad_f32 v13, 2.0, v27, -1.0                           ; D1C1000D 03CE36F4
	v_mad_f32 v1, 2.0, v25, -1.0                            ; D1C10001 03CE32F4
	v_mad_f32 v15, 2.0, v30, -1.0                           ; D1C1000F 03CE3CF4
	v_mad_f32 v8, 2.0, v29, -1.0                            ; D1C10008 03CE3AF4
	v_mad_f32 v7, v7, v15, -v15                             ; D1C10007 843E1F07
	v_mad_f32 v9, 2.0, v26, -1.0                            ; D1C10009 03CE34F4
	v_add_f32_e32 v10, v27, v27                             ; 0214371B
	v_mad_f32 v7, v8, v13, -v7                              ; D1C10007 841E1B08
	v_add_f32_e32 v25, v28, v28                             ; 0232391C
	v_mad_f32 v14, 2.0, v28, -1.0                           ; D1C1000E 03CE38F4
	v_mad_f32 v0, v0, v8, -v8                               ; D1C10000 84221100
	v_add_f32_e32 v28, v31, v31                             ; 02383F1F
	v_mad_f32 v0, v14, v9, -v0                              ; D1C10000 8402130E
	v_mad_f32 v9, v10, v14, -v14                            ; D1C10009 843A1D0A
	v_add_f32_e32 v26, v30, v30                             ; 02343D1E
	s_waitcnt vmcnt(0)                                      ; BF8C0F70
	v_mad_f32 v24, v28, v7, -v7                             ; D1C10018 841E0F1C
	v_mad_f32 v1, v15, v1, -v9                              ; D1C10001 8426030F
	v_mad_f32 v16, v28, v0, -v0                             ; D1C10010 8402011C
	v_mad_f32 v0, v26, v24, -v24                            ; D1C10000 8462311A
	v_add_f32_e32 v27, v29, v29                             ; 02363B1D
	v_mad_f32 v0, v16, v14, -v0                             ; D1C10000 84021D10
	v_mad_f32 v1, v28, v1, -v1                              ; D1C10001 8406031C
	v_mad_f32 v9, v27, v16, -v16                            ; D1C10009 8442211B
	v_mad_f32 v7, v25, v1, -v1                              ; D1C10007 84060319
	v_mad_f32 v9, v1, v15, -v9                              ; D1C10009 84261F01
	v_mad_f32 v0, v28, v0, -v0                              ; D1C10000 8402011C
	v_mad_f32 v7, v24, v8, -v7                              ; D1C10007 841E1118
	v_mad_f32 v25, v28, v9, -v9                             ; D1C10019 8426131C
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	v_mul_f32_e32 v10, s5, v0                               ; 0A140005
	v_mul_f32_e32 v9, s5, v1                                ; 0A120205
	v_mad_f32 v26, v28, v7, -v7                             ; D1C1001A 841E0F1C
	v_mac_f32_e32 v10, s4, v25                              ; 2C143204
	v_mad_f32 v7, v27, s5, -s5                              ; D1C10007 80140B1B
	v_mac_f32_e32 v9, s4, v24                               ; 2C123004
	v_mac_f32_e32 v10, s0, v26                              ; 2C143400
	v_mac_f32_e32 v7, s4, v14                               ; 2C0E1C04
	v_mac_f32_e32 v9, s0, v16                               ; 2C122000
	v_mul_f32_e32 v27, v10, v10                             ; 0A36150A
	v_mac_f32_e32 v7, s0, v15                               ; 2C0E1E00
	v_mac_f32_e32 v27, v9, v9                               ; 2C361309
	v_mac_f32_e32 v27, v7, v7                               ; 2C360F07
	v_mad_f32 v23, 2.0, v31, -1.0                           ; D1C10017 03CE3EF4
	v_cmp_nlt_f32_e32 vcc, 0, v27                           ; 7C9C3680
	s_and_saveexec_b64 s[6:7], vcc                          ; BE86206A
	s_xor_b64 s[6:7], exec, s[6:7]                          ; 8886067E
	v_cmp_le_f32_e32 vcc, 0, v27                            ; 7C863680
	v_mul_f32_e32 v13, 0x7f7fffff, v27                      ; 0A1A36FF 7F7FFFFF
	v_mov_b32_e32 v28, 0xff7fffff                           ; 7E3802FF FF7FFFFF
	v_cndmask_b32_e32 v13, v28, v13, vcc                    ; 001A1B1C
	s_or_saveexec_b64 s[6:7], s[6:7]                        ; BE862106
	s_buffer_load_dword s11, s[16:19], 0x8                  ; C02202C8 00000008
	s_buffer_load_dword s12, s[16:19], 0x18                 ; C0220308 00000018
	s_buffer_load_dword s1, s[16:19], 0x28                  ; C0220048 00000028
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	s_xor_b64 exec, exec, s[6:7]                            ; 88FE067E
	v_rsq_f32_e32 v13, v27                                  ; 7E1A491B
	s_or_b64 exec, exec, s[6:7]                             ; 87FE067E
	s_load_dwordx4 s[44:47], s[2:3], 0x20                   ; C00A0B01 00000020
	s_buffer_load_dword s41, s[16:19], 0x2ec                ; C0220A48 000002EC
	s_buffer_load_dword s42, s[16:19], 0x348                ; C0220A88 00000348
	s_buffer_load_dword s43, s[16:19], 0x34c                ; C0220AC8 0000034C
	s_buffer_load_dword s48, s[16:19], 0x390                ; C0220C08 00000390
	s_buffer_load_dword s49, s[16:19], 0x394                ; C0220C48 00000394
	s_buffer_load_dword s50, s[16:19], 0x398                ; C0220C88 00000398
	v_mul_f32_e32 v1, s12, v1                               ; 0A02020C
	v_mul_f32_e32 v0, s12, v0                               ; 0A00000C
	v_mac_f32_e32 v1, s11, v24                              ; 2C02300B
	v_mac_f32_e32 v0, s11, v25                              ; 2C00320B
	v_mul_f32_e32 v8, s12, v8                               ; 0A10100C
	v_mac_f32_e32 v1, s1, v16                               ; 2C022001
	v_mac_f32_e32 v8, s11, v14                              ; 2C101C0B
	v_mac_f32_e32 v0, s1, v26                               ; 2C003401
	v_mul_f32_e32 v25, v1, v1                               ; 0A320301
	v_mac_f32_e32 v8, s1, v15                               ; 2C101E01
	v_mac_f32_e32 v25, v0, v0                               ; 2C320100
	v_mac_f32_e32 v25, v8, v8                               ; 2C321108
	v_cmp_nlt_f32_e32 vcc, 0, v25                           ; 7C9C3280
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	s_and_saveexec_b64 s[52:53], vcc                        ; BEB4206A
	s_xor_b64 s[52:53], exec, s[52:53]                      ; 88B4347E
	v_cmp_le_f32_e32 vcc, 0, v25                            ; 7C863280
	v_mul_f32_e32 v14, 0x7f7fffff, v25                      ; 0A1C32FF 7F7FFFFF
	v_mov_b32_e32 v15, 0xff7fffff                           ; 7E1E02FF FF7FFFFF
	v_cndmask_b32_e32 v14, v15, v14, vcc                    ; 001C1D0F
	s_or_saveexec_b64 s[52:53], s[52:53]                    ; BEB42134
	s_buffer_load_dword s29, s[16:19], 0x4                  ; C0220748 00000004
	s_buffer_load_dword s2, s[16:19], 0xc                   ; C0220088 0000000C
	s_buffer_load_dword s31, s[16:19], 0x14                 ; C02207C8 00000014
	s_buffer_load_dword s30, s[16:19], 0x1c                 ; C0220788 0000001C
	s_buffer_load_dword s10, s[16:19], 0x24                 ; C0220288 00000024
	s_buffer_load_dword s3, s[16:19], 0x2c                  ; C02200C8 0000002C
	s_buffer_load_dword s6, s[16:19], 0x30                  ; C0220188 00000030
	s_buffer_load_dword s7, s[16:19], 0x34                  ; C02201C8 00000034
	s_buffer_load_dword s8, s[16:19], 0x38                  ; C0220208 00000038
	s_buffer_load_dword s9, s[16:19], 0x3c                  ; C0220248 0000003C
	s_buffer_load_dword s54, s[16:19], 0x12c                ; C0220D88 0000012C
	s_buffer_load_dword s32, s[16:19], 0x2e4                ; C0220808 000002E4
	s_buffer_load_dword s39, s[16:19], 0x2e8                ; C02209C8 000002E8
	s_buffer_load_dword s40, s[16:19], 0x300                ; C0220A08 00000300
	s_buffer_load_dword s38, s[16:19], 0x304                ; C0220988 00000304
	s_buffer_load_dword s34, s[16:19], 0x308                ; C0220888 00000308
	s_buffer_load_dword s37, s[16:19], 0x30c                ; C0220948 0000030C
	s_buffer_load_dword s33, s[16:19], 0x310                ; C0220848 00000310
	s_buffer_load_dword s35, s[16:19], 0x314                ; C02208C8 00000314
	s_buffer_load_dword s56, s[16:19], 0x318                ; C0220E08 00000318
	s_buffer_load_dword s55, s[16:19], 0x31c                ; C0220DC8 0000031C
	s_buffer_load_dword s51, s[16:19], 0x320                ; C0220CC8 00000320
	s_buffer_load_dword s36, s[16:19], 0x324                ; C0220908 00000324
	s_buffer_load_dword s57, s[16:19], 0x330                ; C0220E48 00000330
	s_buffer_load_dword s58, s[16:19], 0x340                ; C0220E88 00000340
	s_buffer_load_dword s59, s[16:19], 0x344                ; C0220EC8 00000344
	s_buffer_load_dword s60, s[16:19], 0x380                ; C0220F08 00000380
	s_buffer_load_dword s61, s[16:19], 0x384                ; C0220F48 00000384
	s_buffer_load_dword s62, s[16:19], 0x388                ; C0220F88 00000388
	s_buffer_load_dword s13, s[44:47], 0x0                  ; C0220356 00000000
	s_buffer_load_dword s14, s[44:47], 0x4                  ; C0220396 00000004
	s_buffer_load_dword s15, s[44:47], 0x8                  ; C02203D6 00000008
	s_buffer_load_dword s16, s[44:47], 0xc                  ; C0220416 0000000C
	s_buffer_load_dword s17, s[44:47], 0x10                 ; C0220456 00000010
	s_buffer_load_dword s19, s[44:47], 0x14                 ; C02204D6 00000014
	s_buffer_load_dword s21, s[44:47], 0x18                 ; C0220556 00000018
	s_buffer_load_dword s23, s[44:47], 0x1c                 ; C02205D6 0000001C
	s_buffer_load_dword s18, s[44:47], 0x20                 ; C0220496 00000020
	s_buffer_load_dword s20, s[44:47], 0x24                 ; C0220516 00000024
	s_buffer_load_dword s22, s[44:47], 0x28                 ; C0220596 00000028
	s_buffer_load_dword s24, s[44:47], 0x2c                 ; C0220616 0000002C
	s_buffer_load_dword s25, s[44:47], 0x30                 ; C0220656 00000030
	s_buffer_load_dword s26, s[44:47], 0x34                 ; C0220696 00000034
	s_buffer_load_dword s27, s[44:47], 0x38                 ; C02206D6 00000038
	s_buffer_load_dword s28, s[44:47], 0x3c                 ; C0220716 0000003C
	v_mov_b32_e32 v24, s41                                  ; 7E300229
	v_mov_b32_e32 v15, s42                                  ; 7E1E022A
	v_mov_b32_e32 v16, s43                                  ; 7E20022B
	v_mov_b32_e32 v26, s48                                  ; 7E340230
	v_mov_b32_e32 v27, s49                                  ; 7E360231
	v_mov_b32_e32 v28, s50                                  ; 7E380232
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	s_xor_b64 exec, exec, s[52:53]                          ; 88FE347E
	v_rsq_f32_e32 v14, v25                                  ; 7E1C4919
	s_or_b64 exec, exec, s[52:53]                           ; 87FE347E
	v_mac_f32_e32 v15, s59, v22                             ; 2C1E2C3B
	v_mov_b32_e32 v22, 0x3f9ae148                           ; 7E2C02FF 3F9AE148
	v_mad_f32 v17, v26, v17, s60                            ; D1C10011 00F2231A
	v_mac_f32_e32 v24, s54, v22                             ; 2C302C36
	v_mad_f32 v22, s55, v17, v24                            ; D1C10016 04622237
	v_floor_f32_e64 v25, |v22|                              ; D15F0119 00000116
	v_sub_f32_e64 v25, |v22|, v25                           ; D1020119 00023316
	v_bfrev_b32_e32 v26, 1                                  ; 7E345881
	v_mad_f32 v18, v27, v18, s61                            ; D1C10012 00F6251B
	v_xor_b32_e32 v27, v25, v26                             ; 2A363519
	v_cmp_ge_f32_e64 vcc, v22, -v22                         ; D046006A 40022D16
	v_cndmask_b32_e32 v22, v27, v25, vcc                    ; 002C331B
	v_mad_f32 v25, s51, v17, v24                            ; D1C10019 04622233
	v_mul_f32_e32 v25, 0.5, v25                             ; 0A3232F0
	v_floor_f32_e64 v27, |v25|                              ; D15F011B 00000119
	v_sub_f32_e64 v27, |v25|, v27                           ; D102011B 00023719
	v_mad_f32 v19, v28, v19, s62                            ; D1C10013 00FA271C
	v_xor_b32_e32 v28, v27, v26                             ; 2A38351B
	v_cmp_ge_f32_e64 vcc, v25, -v25                         ; D046006A 40023319
	v_cndmask_b32_e32 v25, v28, v27, vcc                    ; 0032371C
	v_floor_f32_e64 v27, |v24|                              ; D15F011B 00000118
	v_sub_f32_e64 v27, |v24|, v27                           ; D102011B 00023718
	v_xor_b32_e32 v28, v27, v26                             ; 2A38351B
	v_cmp_ge_f32_e64 vcc, v24, -v24                         ; D046006A 40023118
	v_cndmask_b32_e32 v27, v28, v27, vcc                    ; 0036371C
	v_mul_f32_e32 v28, 0x3f44ec50, v24                      ; 0A3830FF 3F44EC50
	v_floor_f32_e64 v29, |v28|                              ; D15F011D 0000011C
	v_sub_f32_e64 v29, |v28|, v29                           ; D102011D 00023B1C
	v_xor_b32_e32 v30, v29, v26                             ; 2A3C351D
	v_cmp_ge_f32_e64 vcc, v28, -v28                         ; D046006A 4002391C
	v_add_f32_e32 v24, s40, v24                             ; 02303028
	v_cndmask_b32_e32 v28, v30, v29, vcc                    ; 00383B1E
	v_mad_f32 v29, s39, v17, -v24                           ; D1C1001D 84622227
	v_mul_f32_e32 v29, 0x3ee8ba2e, v29                      ; 0A3A3AFF 3EE8BA2E
	v_floor_f32_e64 v30, |v29|                              ; D15F011E 0000011D
	v_sub_f32_e64 v30, |v29|, v30                           ; D102011E 00023D1D
	v_xor_b32_e32 v31, v30, v26                             ; 2A3E351E
	v_cmp_ge_f32_e64 vcc, v29, -v29                         ; D046006A 40023B1D
	v_cndmask_b32_e32 v29, v31, v30, vcc                    ; 003A3D1F
	v_mad_f32 v30, s38, v17, -v24                           ; D1C1001E 84622226
	v_mul_f32_e32 v30, 0x3e50fac6, v30                      ; 0A3C3CFF 3E50FAC6
	v_floor_f32_e64 v31, |v30|                              ; D15F011F 0000011E
	v_sub_f32_e64 v31, |v30|, v31                           ; D102011F 00023F1E
	v_xor_b32_e32 v32, v31, v26                             ; 2A40351F
	v_cmp_ge_f32_e64 vcc, v30, -v30                         ; D046006A 40023D1E
	v_cndmask_b32_e32 v30, v32, v31, vcc                    ; 003C3F20
	v_mov_b32_e32 v31, 0x40c90fdb                           ; 7E3E02FF 40C90FDB
	v_mul_f32_e32 v22, v31, v22                             ; 0A2C2D1F
	v_mov_b32_e32 v32, 0x3e22f983                           ; 7E4002FF 3E22F983
	v_mul_f32_e32 v25, v31, v25                             ; 0A32331F
	v_mul_f32_e32 v22, v32, v22                             ; 0A2C2D20
	v_mul_f32_e32 v25, v32, v25                             ; 0A323320
	v_fract_f32_e32 v22, v22                                ; 7E2C3716
	v_fract_f32_e32 v25, v25                                ; 7E323719
	v_sin_f32_e32 v22, v22                                  ; 7E2C5316
	v_sin_f32_e32 v25, v25                                  ; 7E325319
	v_mad_f32 v24, s37, v17, -v24                           ; D1C10018 84622225
	v_floor_f32_e64 v33, |v24|                              ; D15F0121 00000118
	v_sub_f32_e64 v33, |v24|, v33                           ; D1020121 00024318
	v_add_f32_e32 v22, v22, v25                             ; 022C3316
	v_mul_f32_e32 v25, v31, v29                             ; 0A323B1F
	v_mul_f32_e32 v29, v31, v30                             ; 0A3A3D1F
	v_xor_b32_e32 v34, v33, v26                             ; 2A443521
	v_cmp_ge_f32_e64 vcc, v24, -v24                         ; D046006A 40023118
	v_cndmask_b32_e32 v24, v34, v33, vcc                    ; 00304322
	v_mul_f32_e32 v29, v32, v29                             ; 0A3A3B20
	v_mul_f32_e32 v25, v32, v25                             ; 0A323320
	v_mul_f32_e32 v24, v31, v24                             ; 0A30311F
	v_fract_f32_e32 v29, v29                                ; 7E3A371D
	v_fract_f32_e32 v25, v25                                ; 7E323719
	v_mul_f32_e32 v10, v10, v13                             ; 0A141B0A
	v_mul_f32_e32 v9, v9, v13                               ; 0A121B09
	v_mul_f32_e32 v7, v7, v13                               ; 0A0E1B07
	exp 15, 32, 0, 0, 0, v10, v9, v7, v0                    ; C400020F 0007090A
	v_mul_f32_e32 v24, v32, v24                             ; 0A303120
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
	v_mul_f32_e32 v7, v31, v27                              ; 0A0E371F
	v_mul_f32_e32 v9, v31, v28                              ; 0A12391F
	v_sin_f32_e32 v29, v29                                  ; 7E3A531D
	v_sin_f32_e32 v25, v25                                  ; 7E325319
	v_fract_f32_e32 v24, v24                                ; 7E303718
	v_mul_f32_e32 v7, v32, v7                               ; 0A0E0F20
	v_mul_f32_e32 v9, v32, v9                               ; 0A121320
	v_fract_f32_e32 v7, v7                                  ; 7E0E3707
	v_fract_f32_e32 v9, v9                                  ; 7E123709
	v_sin_f32_e32 v24, v24                                  ; 7E305318
	v_sin_f32_e32 v7, v7                                    ; 7E0E5307
	v_sin_f32_e32 v9, v9                                    ; 7E125309
	v_add_f32_e32 v25, v25, v29                             ; 02323B19
	v_mul_f32_e32 v29, s35, v25                             ; 0A3A3223
	v_mac_f32_e32 v29, s33, v24                             ; 2C3A3021
	v_mov_b32_e32 v24, 0xc1200000                           ; 7E3002FF C1200000
	v_mov_b32_e32 v30, 0xc2200000                           ; 7E3C02FF C2200000
	v_mac_f32_e32 v16, s58, v21                             ; 2C202A3A
	v_mul_f32_e32 v21, s56, v17                             ; 0A2A2238
	v_add_f32_e32 v7, v7, v9                                ; 020E1307
	v_mul_f32_e64 v25, |v25|, s34                           ; D1050119 00004519
	v_mac_f32_e32 v24, s32, v30                             ; 2C303C20
	v_mac_f32_e32 v18, v11, v29                             ; 2C243B0B
	v_mul_f32_e32 v24, v24, v25                             ; 0A303318
	v_mul_f32_e32 v22, s36, v22                             ; 0A2C2C24
	v_mac_f32_e32 v21, v21, v7                              ; 2C2A0F15
	v_madmk_f32_e32 v7, v21, 0x3e4ccccd, v22                ; 2E0E2D15 3E4CCCCD
	v_mul_f32_e32 v30, s31, v18                             ; 0A3C241F
	v_mac_f32_e32 v17, v11, v24                             ; 2C22310B
	v_mul_f32_e32 v29, s5, v18                              ; 0A3A2405
	v_mac_f32_e32 v30, s29, v17                             ; 2C3C221D
	v_mac_f32_e32 v19, v11, v7                              ; 2C260F0B
	v_mul_f32_e32 v33, s12, v18                             ; 0A42240C
	v_mac_f32_e32 v29, s4, v17                              ; 2C3A2204
	v_add_f32_e32 v20, 0, v20                               ; 02282880
	v_mac_f32_e32 v30, s10, v19                             ; 2C3C260A
	v_mac_f32_e32 v30, s7, v20                              ; 2C3C2807
	v_mul_f32_e32 v18, s30, v18                             ; 0A24241E
	v_mac_f32_e32 v33, s11, v17                             ; 2C42220B
	v_mac_f32_e32 v29, s0, v19                              ; 2C3A2600
	v_mac_f32_e32 v18, s2, v17                              ; 2C242202
	v_mac_f32_e32 v33, s1, v19                              ; 2C422601
	v_mul_f32_e32 v23, s57, v23                             ; 0A2E2E39
	v_mul_f32_e32 v0, v0, v14                               ; 0A001D00
	v_mul_f32_e32 v1, v1, v14                               ; 0A021D01
	v_mul_f32_e32 v8, v8, v14                               ; 0A101D08
	v_mac_f32_e32 v29, s6, v20                              ; 2C3A2806
	v_mul_f32_e32 v9, s19, v30                              ; 0A123C13
	v_mul_f32_e32 v10, s21, v30                             ; 0A143C15
	v_mul_f32_e32 v13, s23, v30                             ; 0A1A3C17
	v_mac_f32_e32 v18, s3, v19                              ; 2C242603
	v_mul_f32_e32 v7, s17, v30                              ; 0A0E3C11
	exp 15, 33, 0, 0, 0, v0, v1, v8, v23                    ; C400021F 17080100
	v_mac_f32_e32 v33, s8, v20                              ; 2C422808
	v_mac_f32_e32 v9, s14, v29                              ; 2C123A0E
	v_mac_f32_e32 v10, s15, v29                             ; 2C143A0F
	v_mac_f32_e32 v13, s16, v29                             ; 2C1A3A10
	v_mac_f32_e32 v7, s13, v29                              ; 2C0E3A0D
	exp 15, 34, 0, 0, 0, v16, v15, v0, v0                   ; C400022F 00000F10
	v_mac_f32_e32 v18, s9, v20                              ; 2C242809
	v_mac_f32_e32 v9, s20, v33                              ; 2C124214
	v_mac_f32_e32 v10, s22, v33                             ; 2C144216
	v_mac_f32_e32 v13, s24, v33                             ; 2C1A4218
	v_mac_f32_e32 v7, s18, v33                              ; 2C0E4212
	v_mac_f32_e32 v9, s26, v18                              ; 2C12241A
	v_mac_f32_e32 v10, s27, v18                             ; 2C14241B
	v_mac_f32_e32 v13, s28, v18                             ; 2C1A241C
	v_mov_b32_e32 v14, 0                                    ; 7E1C0280
	exp 15, 35, 0, 0, 0, v3, v4, v6, v5                     ; C400023F 05060403
	v_mac_f32_e32 v7, s25, v18                              ; 2C0E2419
	v_xor_b32_e32 v9, v9, v26                               ; 2A123509
	exp 15, 36, 0, 0, 0, v11, v12, v14, v14                 ; C400024F 0E0E0C0B
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
	v_mad_f32 v0, 2.0, v10, -v13                            ; D1C10000 843614F4
	exp 15, 12, 0, 1, 0, v7, v9, v0, v13                    ; C40008CF 0D000907
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
Shader epilog disassembly:
	s_endpgm ; BF810000

*** SHADER STATS ***
SGPRS: 80
VGPRS: 36
Spilled VGPRs: 0
Code Size: 1920 bytes
LDS: 0 blocks
Scratch: 0 bytes per wave
Max Waves: 7
********************
[...]
[traceshaders] glGenProgramPipelines():
[traceshaders]  - pipeline=20
[traceshaders] glUseProgramStages(pipeline=20, stages=vert, program=74)
[traceshaders] glUseProgramStages(pipeline=20, stages=frag, program=90)
[traceshaders] glBindProgramPipeline(pipeline=20)

Pixel Shader:
Shader main disassembly:
	s_wqm_b64 exec, exec                                       ; BEFE077E
	s_mov_b32 m0, s11                                          ; BEFC000B
	v_interp_p1_f32 v6, v2, 0, 0, [m0]                         ; D4180002
	v_interp_p2_f32 v6, [v6], v3, 0, 0, [m0]                   ; D4190003
	v_interp_p1_f32 v7, v2, 1, 0, [m0]                         ; D41C0102
	v_interp_p2_f32 v7, [v7], v3, 1, 0, [m0]                   ; D41D0103
	v_interp_p1_f32 v0, v2, 2, 0, [m0]                         ; D4000202
	v_interp_p2_f32 v0, [v0], v3, 2, 0, [m0]                   ; D4010203
	v_interp_p1_f32 v8, v2, 0, 1, [m0]                         ; D4200402
	v_interp_p2_f32 v8, [v8], v3, 0, 1, [m0]                   ; D4210403
	v_interp_p1_f32 v9, v2, 1, 1, [m0]                         ; D4240502
	v_interp_p2_f32 v9, [v9], v3, 1, 1, [m0]                   ; D4250503
	v_interp_p1_f32 v1, v2, 2, 1, [m0]                         ; D4040602
	v_interp_p2_f32 v1, [v1], v3, 2, 1, [m0]                   ; D4050603
	v_interp_p1_f32 v10, v2, 3, 1, [m0]                        ; D4280702
	v_interp_p2_f32 v10, [v10], v3, 3, 1, [m0]                 ; D4290703
	v_interp_p1_f32 v4, v2, 0, 2, [m0]                         ; D4100802
	v_interp_p2_f32 v4, [v4], v3, 0, 2, [m0]                   ; D4110803
	v_interp_p1_f32 v5, v2, 1, 2, [m0]                         ; D4140902
	v_mul_f32_e32 v2, v9, v6                                   ; 0A040D09
	s_load_dwordx4 s[24:27], s[2:3], 0x10                      ; C00A0601 00000010
	v_mad_f32 v2, v8, v7, -v2                                  ; D1C10002 840A0F08
	v_interp_p2_f32 v5, [v5], v3, 1, 2, [m0]                   ; D4150903
	v_mul_f32_e32 v2, v10, v2                                  ; 0A04050A
	v_mul_f32_e32 v3, v0, v0                                   ; 0A060100
	v_mac_f32_e32 v3, v2, v2                                   ; 2C060502
	v_mac_f32_e32 v3, v1, v1                                   ; 2C060301
	v_cmp_nlt_f32_e32 vcc, 0, v3                               ; 7C9C0680
	s_waitcnt lgkmcnt(0)                                       ; BF8C007F
	s_and_saveexec_b64 s[28:29], vcc                           ; BE9C206A
	s_xor_b64 s[28:29], exec, s[28:29]                         ; 889C1C7E
	v_cmp_le_f32_e32 vcc, 0, v3                                ; 7C860680
	v_mul_f32_e32 v6, 0x7f7fffff, v3                           ; 0A0C06FF 7F7FFFFF
	v_mov_b32_e32 v7, 0xff7fffff                               ; 7E0E02FF FF7FFFFF
	v_cndmask_b32_e32 v6, v7, v6, vcc                          ; 000C0D07
	s_or_saveexec_b64 s[28:29], s[28:29]                       ; BE9C211C
	s_buffer_load_dword s23, s[24:27], 0x2d0                   ; C02205CC 000002D0
	s_buffer_load_dword s17, s[24:27], 0x3a0                   ; C022044C 000003A0
	s_buffer_load_dword s15, s[24:27], 0x3a4                   ; C02203CC 000003A4
	s_buffer_load_dword s13, s[24:27], 0x3a8                   ; C022034C 000003A8
	s_buffer_load_dword s11, s[24:27], 0x3ac                   ; C02202CC 000003AC
	s_buffer_load_dword s19, s[24:27], 0x3b0                   ; C02204CC 000003B0
	s_buffer_load_dword s20, s[24:27], 0x3b4                   ; C022050C 000003B4
	s_buffer_load_dword s21, s[24:27], 0x3b8                   ; C022054C 000003B8
	s_buffer_load_dword s22, s[24:27], 0x3bc                   ; C022058C 000003BC
	s_buffer_load_dword s18, s[24:27], 0x3c0                   ; C022048C 000003C0
	s_buffer_load_dword s16, s[24:27], 0x3c4                   ; C022040C 000003C4
	s_buffer_load_dword s14, s[24:27], 0x3c8                   ; C022038C 000003C8
	s_buffer_load_dword s12, s[24:27], 0x3cc                   ; C022030C 000003CC
	s_buffer_load_dword s9, s[24:27], 0x3d0                    ; C022024C 000003D0
	s_buffer_load_dword s8, s[24:27], 0x3d4                    ; C022020C 000003D4
	s_buffer_load_dword s7, s[24:27], 0x3d8                    ; C02201CC 000003D8
	s_buffer_load_dword s0, s[24:27], 0x3dc                    ; C022000C 000003DC
	s_buffer_load_dword s1, s[24:27], 0x3e0                    ; C022004C 000003E0
	s_buffer_load_dword s2, s[24:27], 0x3e4                    ; C022008C 000003E4
	s_buffer_load_dword s3, s[24:27], 0x3e8                    ; C02200CC 000003E8
	s_buffer_load_dword s6, s[24:27], 0x3ec                    ; C022018C 000003EC
	s_load_dwordx8 s[32:39], s[4:5], 0x0                       ; C00E0802 00000000
	s_nop 0                                                    ; BF800000
	s_load_dwordx4 s[24:27], s[4:5], 0x30                      ; C00A0602 00000030
	s_waitcnt lgkmcnt(0)                                       ; BF8C007F
	s_xor_b64 exec, exec, s[28:29]                             ; 88FE1C7E
	v_rsq_f32_e32 v6, v3                                       ; 7E0C4903
	s_or_b64 exec, exec, s[28:29]                              ; 87FE1C7E
	v_mul_f32_e32 v2, v2, v6                                   ; 0A040D02
	v_mul_f32_e32 v1, v1, v6                                   ; 0A020D01
	v_mad_f32 v2, 0.5, v2, 0.5                                 ; D1C10002 03C204F0
	v_mad_f32 v11, 0.5, v1, 0.5                                ; D1C1000B 03C202F0
	v_add_f32_e64 v1, 0, v2 clamp                              ; D1018001 00020480
	v_mov_b32_e32 v2, 0xc0400000                               ; 7E0402FF C0400000
	v_add_f32_e32 v2, s23, v2                                  ; 02040417
	v_mul_f32_e32 v2, 0x3b000000, v2                           ; 0A0404FF 3B000000
	v_add_f32_e64 v2, 0, v2 clamp                              ; D1018002 00020480
	image_sample v[7:10], v[4:5], s[32:39], s[24:27] dmask:0xf ; F0800F00 00C80704
	v_cmp_gt_f32_e32 vcc, 0, v2                                ; 7C880480
	v_sqrt_f32_e32 v2, v2                                      ; 7E044F02
	v_mul_f32_e32 v0, v0, v6                                   ; 0A000D00
	s_waitcnt vmcnt(0)                                         ; BF8C0F70
	v_mul_f32_e32 v5, s20, v8                                  ; 0A0A1014
	v_mul_f32_e32 v6, s21, v8                                  ; 0A0C1015
	v_cndmask_b32_e64 v3, v2, 0, vcc                           ; D1000003 01A90102
	v_mul_f32_e32 v2, s19, v8                                  ; 0A041013
	v_mul_f32_e32 v8, s22, v8                                  ; 0A101016
	v_mac_f32_e32 v2, s17, v7                                  ; 2C040E11
	v_mac_f32_e32 v5, s15, v7                                  ; 2C0A0E0F
	v_mac_f32_e32 v6, s13, v7                                  ; 2C0C0E0D
	v_mac_f32_e32 v8, s11, v7                                  ; 2C100E0B
	v_mac_f32_e32 v2, s18, v9                                  ; 2C041212
	v_mac_f32_e32 v5, s16, v9                                  ; 2C0A1210
	v_mac_f32_e32 v6, s14, v9                                  ; 2C0C120E
	v_mac_f32_e32 v8, s12, v9                                  ; 2C10120C
	v_mad_f32 v0, 0.5, v0, 0.5                                 ; D1C10000 03C200F0
	v_mac_f32_e32 v5, s8, v10                                  ; 2C0A1408
	v_mac_f32_e32 v6, s7, v10                                  ; 2C0C1407
	v_mac_f32_e32 v2, s9, v10                                  ; 2C041409
	v_mac_f32_e32 v8, s0, v10                                  ; 2C101400
	v_add_f32_e32 v4, s1, v2                                   ; 02080401
	v_add_f32_e64 v0, 0, v0 clamp                              ; D1018000 00020080
	v_add_f32_e32 v5, s2, v5                                   ; 020A0A02
	v_add_f32_e32 v6, s3, v6                                   ; 020C0C03
	v_add_f32_e32 v7, s6, v8                                   ; 020E1006
	v_add_f32_e64 v2, 0, v11 clamp                             ; D1018002 00021680
Shader epilog disassembly:
	v_cvt_pkrtz_f16_f32_e64 v0, v0, v1 ; D2960000 00020300
	v_cvt_pkrtz_f16_f32_e64 v1, v2, v3 ; D2960001 00020702
	exp 15, 0, 1, 0, 0, v0, v1, v0, v0 ; C400040F 00000100
	s_waitcnt expcnt(0)                ; BF8C0F0F
	v_cvt_pkrtz_f16_f32_e64 v0, v4, v5 ; D2960000 00020B04
	v_cvt_pkrtz_f16_f32_e64 v1, v6, v7 ; D2960001 00020F06
	exp 15, 1, 1, 1, 1, v0, v1, v0, v0 ; C4001C1F 00000100
	s_endpgm                           ; BF810000

*** SHADER CONFIG ***
SPI_PS_INPUT_ADDR = 0xd077
SPI_PS_INPUT_ENA  = 0x0002
*** SHADER STATS ***
SGPRS: 80
VGPRS: 16
Spilled VGPRs: 0
Code Size: 616 bytes
LDS: 0 blocks
Scratch: 0 bytes per wave
Max Waves: 10
********************
[...]
[traceshaders] glGenProgramPipelines():
[traceshaders]  - pipeline=105
[traceshaders] glUseProgramStages(pipeline=105, stages=vert, program=384)
[traceshaders] glUseProgramStages(pipeline=105, stages=frag, program=382)
[traceshaders] glBindProgramPipeline(pipeline=105)

Vertex Shader as VS:
Shader prolog disassembly:
	v_add_i32_e32 v4, vcc, s12, v0 ; 3208000C
	v_mov_b32_e32 v5, v4           ; 7E0A0304
	v_mov_b32_e32 v6, v4           ; 7E0C0304
	v_mov_b32_e32 v7, v4           ; 7E0E0304
	v_mov_b32_e32 v8, v4           ; 7E100304
	v_mov_b32_e32 v9, v4           ; 7E120304
	v_mov_b32_e32 v10, v4          ; 7E140304
Shader main disassembly:
	s_load_dwordx4 s[4:7], s[10:11], 0x0                    ; C00A0105 00000000
	s_load_dwordx4 s[12:15], s[10:11], 0x10                 ; C00A0305 00000010
	s_load_dwordx4 s[16:19], s[10:11], 0x20                 ; C00A0405 00000020
	v_mov_b32_e32 v1, 0x3f9ae148                            ; 7E0202FF 3F9AE148
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	buffer_load_format_xyzw v[11:14], v4, s[4:7], 0 idxen   ; E00C2000 80010B04
	s_nop 0                                                 ; BF800000
	buffer_load_format_xyzw v[15:18], v5, s[12:15], 0 idxen ; E00C2000 80030F05
	s_load_dwordx4 s[12:15], s[10:11], 0x30                 ; C00A0305 00000030
	buffer_load_format_xyzw v[3:6], v6, s[16:19], 0 idxen   ; E00C2000 80040306
	s_load_dwordx4 s[16:19], s[10:11], 0x40                 ; C00A0405 00000040
	s_load_dwordx4 s[4:7], s[2:3], 0x10                     ; C00A0101 00000010
	s_load_dwordx4 s[0:3], s[2:3], 0x20                     ; C00A0001 00000020
	s_waitcnt vmcnt(1) lgkmcnt(0)                           ; BF8C0071
	buffer_load_format_xyzw v[18:21], v7, s[12:15], 0 idxen ; E00C2000 80031207
	s_load_dwordx4 s[12:15], s[10:11], 0x50                 ; C00A0305 00000050
	s_load_dwordx4 s[8:11], s[10:11], 0x60                  ; C00A0205 00000060
	s_waitcnt vmcnt(0)                                      ; BF8C0F70
	buffer_load_format_xyzw v[20:23], v8, s[16:19], 0 idxen ; E00C2000 80041408
	s_buffer_load_dword s27, s[4:7], 0x1ec                  ; C02206C2 000001EC
	s_buffer_load_dword s51, s[4:7], 0x280                  ; C0220CC2 00000280
	s_waitcnt vmcnt(0) lgkmcnt(0)                           ; BF8C0070
	buffer_load_format_xyzw v[22:25], v9, s[12:15], 0 idxen ; E00C2000 80031609
	s_nop 0                                                 ; BF800000
	buffer_load_format_xyzw v[7:10], v10, s[8:11], 0 idxen  ; E00C2000 8002070A
	s_buffer_load_dword s11, s[4:7], 0x12c                  ; C02202C2 0000012C
	v_mov_b32_e32 v0, s27                                   ; 7E00021B
	s_waitcnt vmcnt(0)                                      ; BF8C0F70
	v_bfrev_b32_e32 v9, 1                                   ; 7E125881
	s_buffer_load_dword s54, s[4:7], 0x290                  ; C0220D82 00000290
	s_buffer_load_dword s36, s[4:7], 0x220                  ; C0220902 00000220
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	v_mac_f32_e32 v0, s11, v1                               ; 2C00020B
	v_floor_f32_e64 v1, |v0|                                ; D15F0101 00000100
	v_sub_f32_e64 v1, |v0|, v1                              ; D1020101 00020300
	v_xor_b32_e32 v10, v1, v9                               ; 2A141301
	v_cmp_ge_f32_e64 vcc, v0, -v0                           ; D046006A 40020100
	v_cndmask_b32_e32 v1, v10, v1, vcc                      ; 0002030A
	v_mul_f32_e32 v10, 0x3f44ec50, v0                       ; 0A1400FF 3F44EC50
	v_floor_f32_e64 v24, |v10|                              ; D15F0118 0000010A
	v_sub_f32_e64 v24, |v10|, v24                           ; D1020118 0002310A
	v_xor_b32_e32 v25, v24, v9                              ; 2A321318
	v_cmp_ge_f32_e64 vcc, v10, -v10                         ; D046006A 4002150A
	v_cndmask_b32_e32 v10, v25, v24, vcc                    ; 00143119
	v_mov_b32_e32 v24, s51                                  ; 7E300233
	s_buffer_load_dword s28, s[4:7], 0x200                  ; C0220702 00000200
	s_buffer_load_dword s35, s[4:7], 0x21c                  ; C02208C2 0000021C
	s_buffer_load_dword s26, s[4:7], 0x1e8                  ; C0220682 000001E8
	s_buffer_load_dword s29, s[4:7], 0x204                  ; C0220742 00000204
	s_buffer_load_dword s31, s[4:7], 0x20c                  ; C02207C2 0000020C
	s_buffer_load_dword s41, s[4:7], 0x24c                  ; C0220A42 0000024C
	s_buffer_load_dword s38, s[4:7], 0x240                  ; C0220982 00000240
	s_buffer_load_dword s40, s[4:7], 0x248                  ; C0220A02 00000248
	s_buffer_load_dword s39, s[4:7], 0x244                  ; C02209C2 00000244
	s_buffer_load_dword s52, s[4:7], 0x284                  ; C0220D02 00000284
	s_buffer_load_dword s55, s[4:7], 0x294                  ; C0220DC2 00000294
	s_buffer_load_dword s33, s[4:7], 0x214                  ; C0220842 00000214
	s_buffer_load_dword s53, s[4:7], 0x288                  ; C0220D42 00000288
	s_buffer_load_dword s32, s[4:7], 0x210                  ; C0220802 00000210
	s_buffer_load_dword s20, s[4:7], 0x0                    ; C0220502 00000000
	s_buffer_load_dword s21, s[4:7], 0x4                    ; C0220542 00000004
	s_buffer_load_dword s22, s[4:7], 0x8                    ; C0220582 00000008
	s_buffer_load_dword s23, s[4:7], 0xc                    ; C02205C2 0000000C
	s_buffer_load_dword s24, s[4:7], 0x10                   ; C0220602 00000010
	s_buffer_load_dword s16, s[4:7], 0x14                   ; C0220402 00000014
	s_buffer_load_dword s17, s[4:7], 0x18                   ; C0220442 00000018
	s_buffer_load_dword s18, s[4:7], 0x1c                   ; C0220482 0000001C
	s_buffer_load_dword s19, s[4:7], 0x20                   ; C02204C2 00000020
	s_buffer_load_dword s12, s[4:7], 0x24                   ; C0220302 00000024
	s_buffer_load_dword s13, s[4:7], 0x28                   ; C0220342 00000028
	s_buffer_load_dword s14, s[4:7], 0x2c                   ; C0220382 0000002C
	s_buffer_load_dword s15, s[4:7], 0x30                   ; C02203C2 00000030
	s_buffer_load_dword s8, s[4:7], 0x34                    ; C0220202 00000034
	s_buffer_load_dword s9, s[4:7], 0x38                    ; C0220242 00000038
	s_buffer_load_dword s10, s[4:7], 0x3c                   ; C0220282 0000003C
	s_buffer_load_dword s25, s[4:7], 0x1e4                  ; C0220642 000001E4
	s_buffer_load_dword s30, s[4:7], 0x208                  ; C0220782 00000208
	s_buffer_load_dword s34, s[4:7], 0x218                  ; C0220882 00000218
	s_buffer_load_dword s37, s[4:7], 0x224                  ; C0220942 00000224
	s_buffer_load_dword s42, s[4:7], 0x250                  ; C0220A82 00000250
	s_buffer_load_dword s43, s[4:7], 0x254                  ; C0220AC2 00000254
	s_buffer_load_dword s44, s[4:7], 0x258                  ; C0220B02 00000258
	s_buffer_load_dword s45, s[4:7], 0x260                  ; C0220B42 00000260
	s_buffer_load_dword s46, s[4:7], 0x264                  ; C0220B82 00000264
	s_buffer_load_dword s47, s[4:7], 0x268                  ; C0220BC2 00000268
	s_buffer_load_dword s48, s[4:7], 0x270                  ; C0220C02 00000270
	s_buffer_load_dword s49, s[4:7], 0x274                  ; C0220C42 00000274
	s_buffer_load_dword s50, s[4:7], 0x278                  ; C0220C82 00000278
	s_buffer_load_dword s4, s[4:7], 0x298                   ; C0220102 00000298
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	v_mov_b32_e32 v30, s53                                  ; 7E3C0235
	s_buffer_load_dword s57, s[0:3], 0x10                   ; C0220E40 00000010
	s_buffer_load_dword s58, s[0:3], 0x14                   ; C0220E80 00000014
	v_mac_f32_e32 v24, s54, v11                             ; 2C301636
	v_mad_f32 v11, s36, v24, v0                             ; D1C1000B 04023024
	v_mul_f32_e32 v11, 0.5, v11                             ; 0A1616F0
	v_floor_f32_e64 v25, |v11|                              ; D15F0119 0000010B
	v_sub_f32_e64 v25, |v11|, v25                           ; D1020119 0002330B
	v_xor_b32_e32 v26, v25, v9                              ; 2A341319
	v_cmp_ge_f32_e64 vcc, v11, -v11                         ; D046006A 4002170B
	v_cndmask_b32_e32 v11, v26, v25, vcc                    ; 0016331A
	v_add_f32_e32 v25, s28, v0                              ; 0232001C
	v_mac_f32_e32 v0, s35, v24                              ; 2C003023
	v_floor_f32_e64 v26, |v0|                               ; D15F011A 00000100
	v_sub_f32_e64 v26, |v0|, v26                            ; D102011A 00023500
	v_xor_b32_e32 v27, v26, v9                              ; 2A36131A
	v_cmp_ge_f32_e64 vcc, v0, -v0                           ; D046006A 40020100
	v_cndmask_b32_e32 v0, v27, v26, vcc                     ; 0000351B
	v_mad_f32 v26, s26, v24, -v25                           ; D1C1001A 8466301A
	v_mul_f32_e32 v26, 0x3ee8ba2e, v26                      ; 0A3434FF 3EE8BA2E
	v_floor_f32_e64 v27, |v26|                              ; D15F011B 0000011A
	v_sub_f32_e64 v27, |v26|, v27                           ; D102011B 0002371A
	v_xor_b32_e32 v28, v27, v9                              ; 2A38131B
	v_cmp_ge_f32_e64 vcc, v26, -v26                         ; D046006A 4002351A
	v_cndmask_b32_e32 v26, v28, v27, vcc                    ; 0034371C
	v_mad_f32 v27, s29, v24, -v25                           ; D1C1001B 8466301D
	v_mul_f32_e32 v27, 0x3e50fac6, v27                      ; 0A3636FF 3E50FAC6
	v_floor_f32_e64 v28, |v27|                              ; D15F011C 0000011B
	v_sub_f32_e64 v28, |v27|, v28                           ; D102011C 0002391B
	v_xor_b32_e32 v29, v28, v9                              ; 2A3A131C
	v_cmp_ge_f32_e64 vcc, v27, -v27                         ; D046006A 4002371B
	v_mad_f32 v25, s31, v24, -v25                           ; D1C10019 8466301F
	v_cndmask_b32_e32 v27, v29, v28, vcc                    ; 0036391D
	v_floor_f32_e64 v28, |v25|                              ; D15F011C 00000119
	v_sub_f32_e64 v28, |v25|, v28                           ; D102011C 00023919
	v_xor_b32_e32 v29, v28, v9                              ; 2A3A131C
	v_cmp_ge_f32_e64 vcc, v25, -v25                         ; D046006A 40023319
	v_cndmask_b32_e32 v25, v29, v28, vcc                    ; 0032391D
	v_mov_b32_e32 v28, s41                                  ; 7E380229
	v_mac_f32_e32 v28, s38, v7                              ; 2C380E26
	v_mov_b32_e32 v7, s40                                   ; 7E0E0228
	v_mac_f32_e32 v7, s39, v8                               ; 2C0E1027
	v_mov_b32_e32 v8, s52                                   ; 7E100234
	v_mac_f32_e32 v8, s55, v12                              ; 2C101837
	v_mov_b32_e32 v12, 0x40c90fdb                           ; 7E1802FF 40C90FDB
	v_mul_f32_e32 v11, v12, v11                             ; 0A16170C
	v_mov_b32_e32 v29, 0x3e22f983                           ; 7E3A02FF 3E22F983
	v_mul_f32_e32 v0, v12, v0                               ; 0A00010C
	v_mul_f32_e32 v11, v29, v11                             ; 0A16171D
	v_mul_f32_e32 v0, v29, v0                               ; 0A00011D
	v_fract_f32_e32 v11, v11                                ; 7E16370B
	v_fract_f32_e32 v0, v0                                  ; 7E003700
	v_sin_f32_e32 v11, v11                                  ; 7E16530B
	v_sin_f32_e32 v0, v0                                    ; 7E005300
	v_mul_f32_e32 v26, v12, v26                             ; 0A34350C
	v_mul_f32_e32 v26, v29, v26                             ; 0A34351D
	v_fract_f32_e32 v26, v26                                ; 7E34371A
	v_add_f32_e32 v0, v0, v11                               ; 02001700
	v_mul_f32_e32 v11, v12, v27                             ; 0A16370C
	v_mul_f32_e32 v11, v29, v11                             ; 0A16171D
	v_fract_f32_e32 v11, v11                                ; 7E16370B
	v_mul_f32_e32 v25, v12, v25                             ; 0A32330C
	v_mul_f32_e32 v25, v29, v25                             ; 0A32331D
	v_mul_f32_e32 v1, v12, v1                               ; 0A02030C
	v_mul_f32_e32 v10, v12, v10                             ; 0A14150C
	v_sin_f32_e32 v11, v11                                  ; 7E16530B
	v_sin_f32_e32 v26, v26                                  ; 7E34531A
	v_fract_f32_e32 v25, v25                                ; 7E323719
	v_mul_f32_e32 v1, v29, v1                               ; 0A02031D
	v_mul_f32_e32 v10, v29, v10                             ; 0A14151D
	v_fract_f32_e32 v1, v1                                  ; 7E023701
	v_fract_f32_e32 v10, v10                                ; 7E14370A
	v_sin_f32_e32 v25, v25                                  ; 7E325319
	v_sin_f32_e32 v1, v1                                    ; 7E025301
	v_sin_f32_e32 v10, v10                                  ; 7E14530A
	v_add_f32_e32 v11, v26, v11                             ; 0216171A
	v_mul_f32_e32 v26, s33, v11                             ; 0A341621
	v_mac_f32_e32 v26, s32, v25                             ; 2C343220
	v_mov_b32_e32 v25, 0xc1200000                           ; 7E3202FF C1200000
	v_mov_b32_e32 v27, 0xc2200000                           ; 7E3602FF C2200000
	v_mac_f32_e32 v30, s4, v13                              ; 2C3C1A04
	v_add_f32_e32 v1, v1, v10                               ; 02021501
	v_mul_f32_e32 v13, s34, v24                             ; 0A1A3022
	v_mul_f32_e64 v11, |v11|, s30                           ; D105010B 00003D0B
	v_mac_f32_e32 v25, s25, v27                             ; 2C323619
	v_mac_f32_e32 v8, v22, v26                              ; 2C103516
	v_mul_f32_e32 v11, v25, v11                             ; 0A161719
	v_mac_f32_e32 v13, v13, v1                              ; 2C1A030D
	v_mul_f32_e32 v0, s37, v0                               ; 0A000025
	v_madmk_f32_e32 v0, v13, 0x3e4ccccd, v0                 ; 2E00010D 3E4CCCCD
	v_mac_f32_e32 v24, v22, v11                             ; 2C301716
	v_mul_f32_e32 v27, s16, v8                              ; 0A361010
	v_mul_f32_e32 v26, s24, v8                              ; 0A341018
	v_mul_f32_e32 v31, s17, v8                              ; 0A3E1011
	v_mul_f32_e32 v8, s18, v8                               ; 0A101012
	v_mac_f32_e32 v30, v22, v0                              ; 2C3C0116
	v_mac_f32_e32 v27, s21, v24                             ; 2C363015
	s_buffer_load_dword s59, s[0:3], 0x18                   ; C0220EC0 00000018
	s_buffer_load_dword s60, s[0:3], 0x1c                   ; C0220F00 0000001C
	v_mac_f32_e32 v26, s20, v24                             ; 2C343014
	v_mac_f32_e32 v31, s22, v24                             ; 2C3E3016
	v_mac_f32_e32 v8, s23, v24                              ; 2C103017
	s_buffer_load_dword s5, s[0:3], 0x0                     ; C0220140 00000000
	s_buffer_load_dword s6, s[0:3], 0x4                     ; C0220180 00000004
	s_buffer_load_dword s7, s[0:3], 0x8                     ; C02201C0 00000008
	s_buffer_load_dword s56, s[0:3], 0xc                    ; C0220E00 0000000C
	v_mac_f32_e32 v27, s12, v30                             ; 2C363C0C
	v_add_f32_e32 v0, 0, v14                                ; 02001C80
	v_mac_f32_e32 v26, s19, v30                             ; 2C343C13
	v_mac_f32_e32 v31, s13, v30                             ; 2C3E3C0D
	v_mac_f32_e32 v8, s14, v30                              ; 2C103C0E
	v_mac_f32_e32 v27, s8, v0                               ; 2C360008
	s_buffer_load_dword s61, s[0:3], 0x20                   ; C0220F40 00000020
	s_buffer_load_dword s62, s[0:3], 0x24                   ; C0220F80 00000024
	s_buffer_load_dword s63, s[0:3], 0x28                   ; C0220FC0 00000028
	s_buffer_load_dword s64, s[0:3], 0x2c                   ; C0221000 0000002C
	s_buffer_load_dword s65, s[0:3], 0x30                   ; C0221040 00000030
	s_buffer_load_dword s66, s[0:3], 0x34                   ; C0221080 00000034
	s_buffer_load_dword s67, s[0:3], 0x38                   ; C02210C0 00000038
	s_buffer_load_dword s72, s[0:3], 0x3c                   ; C0221200 0000003C
	s_buffer_load_dword s0, s[0:3], 0x4c                    ; C0220000 0000004C
	v_mac_f32_e32 v26, s15, v0                              ; 2C34000F
	v_mac_f32_e32 v31, s9, v0                               ; 2C3E0009
	v_mac_f32_e32 v8, s10, v0                               ; 2C10000A
	s_waitcnt lgkmcnt(0)                                    ; BF8C007F
	v_mul_f32_e32 v0, s57, v27                              ; 0A003639
	exp 15, 32, 0, 0, 0, v28, v7, v0, v0                    ; C400020F 0000071C
	exp 15, 33, 0, 0, 0, v18, v19, v21, v20                 ; C400021F 14151312
	v_mul_f32_e32 v14, s58, v27                             ; 0A1C363A
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
	v_mul_f32_e32 v18, s59, v27                             ; 0A24363B
	v_mul_f32_e32 v19, s60, v27                             ; 0A26363C
	v_add_f32_e32 v1, v16, v16                              ; 02022110
	v_add_f32_e32 v10, v17, v17                             ; 02142311
	v_mad_f32 v11, 2.0, v17, -1.0                           ; D1C1000B 03CE22F4
	v_add_f32_e32 v17, v5, v5                               ; 02220B05
	v_mad_f32 v5, 2.0, v5, -1.0                             ; D1C10005 03CE0AF4
	v_add_f32_e32 v12, v15, v15                             ; 02181F0F
	v_mad_f32 v13, 2.0, v15, -1.0                           ; D1C1000D 03CE1EF4
	v_mov_b32_e32 v15, 0                                    ; 7E1E0280
	v_mac_f32_e32 v0, s5, v26                               ; 2C003405
	v_mac_f32_e32 v14, s6, v26                              ; 2C1C3406
	v_mac_f32_e32 v18, s7, v26                              ; 2C243407
	v_mac_f32_e32 v19, s56, v26                             ; 2C263438
	v_mad_f32 v7, 2.0, v16, -1.0                            ; D1C10007 03CE20F4
	v_add_f32_e32 v16, v3, v3                               ; 02200703
	exp 15, 34, 0, 0, 0, v22, v23, v15, v15                 ; C400022F 0F0F1716
	v_mad_f32 v3, 2.0, v3, -1.0                             ; D1C10003 03CE06F4
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
	v_add_f32_e32 v15, v4, v4                               ; 021E0904
	v_mad_f32 v4, 2.0, v4, -1.0                             ; D1C10004 03CE08F4
	v_mad_f32 v1, v1, v5, -v5                               ; D1C10001 84160B01
	v_mac_f32_e32 v0, s61, v31                              ; 2C003E3D
	v_mac_f32_e32 v14, s62, v31                             ; 2C1C3E3E
	v_mac_f32_e32 v18, s63, v31                             ; 2C243E3F
	v_mac_f32_e32 v19, s64, v31                             ; 2C263E40
	v_mul_f32_e32 v20, s0, v27                              ; 0A283600
	v_mad_f32 v10, v10, v3, -v3                             ; D1C1000A 840E070A
	v_mad_f32 v1, v4, v11, -v1                              ; D1C10001 84061704
	v_add_f32_e32 v6, v6, v6                                ; 020C0D06
	v_mad_f32 v11, v12, v4, -v4                             ; D1C1000B 8412090C
	v_mad_f32 v10, v5, v13, -v10                            ; D1C1000A 842A1B05
	v_mad_f32 v7, v3, v7, -v11                              ; D1C10007 842E0F03
	v_mad_f32 v1, v6, v1, -v1                               ; D1C10001 84060306
	v_mac_f32_e32 v0, s65, v8                               ; 2C001041
	v_mac_f32_e32 v14, s66, v8                              ; 2C1C1042
	v_mac_f32_e32 v18, s67, v8                              ; 2C241043
	v_mac_f32_e32 v19, s72, v8                              ; 2C261048
	v_mul_f32_e32 v8, s0, v26                               ; 0A103400
	v_mul_f32_e32 v22, s45, v20                             ; 0A2C282D
	v_mul_f32_e32 v23, s46, v20                             ; 0A2E282E
	v_mad_f32 v7, v6, v7, -v7                               ; D1C10007 841E0F06
	v_mad_f32 v12, v17, v1, -v1                             ; D1C1000C 84060311
	v_mul_f32_e32 v21, s0, v31                              ; 0A2A3E00
	v_mul_f32_e32 v20, s47, v20                             ; 0A28282F
	v_mad_f32 v22, -v8, s42, -v22                           ; D1C10016 A4585508
	v_mad_f32 v10, v6, v10, -v10                            ; D1C1000A 842A1506
	v_mad_f32 v3, v7, v3, -v12                              ; D1C10003 84320707
	v_mad_f32 v12, v16, v10, -v10                           ; D1C1000C 842A1510
	v_mad_f32 v23, -v8, s43, -v23                           ; D1C10017 A45C5708
	v_mad_f32 v8, -v8, s44, -v20                            ; D1C10008 A4505908
	v_mad_f32 v20, -v21, s48, v22                           ; D1C10014 24586115
	v_mad_f32 v11, v15, v7, -v7                             ; D1C1000B 841E0F0F
	v_mad_f32 v12, v1, v4, -v12                             ; D1C1000C 84320901
	v_mad_f32 v22, -v21, s49, v23                           ; D1C10016 245C6315
	v_mul_f32_e32 v1, v20, v1                               ; 0A020314
	v_mad_f32 v11, v10, v5, -v11                            ; D1C1000B 842E0B0A
	v_mac_f32_e32 v1, v22, v10                              ; 2C021516
	v_mad_f32 v10, v16, v20, -v20                           ; D1C1000A 84522910
	v_mac_f32_e32 v10, v22, v4                              ; 2C140916
	v_mad_f32 v4, v6, v11, -v11                             ; D1C10004 842E1706
	v_mul_f32_e32 v4, v20, v4                               ; 0A080914
	v_mad_f32 v3, v6, v3, -v3                               ; D1C10003 840E0706
	v_mad_f32 v8, -v21, s50, v8                             ; D1C10008 24206515
	v_mac_f32_e32 v4, v22, v3                               ; 2C080716
	v_mad_f32 v3, v6, v12, -v12                             ; D1C10003 84321906
	v_mac_f32_e32 v4, v8, v3                                ; 2C080708
	v_mac_f32_e32 v1, v8, v7                                ; 2C020F08
	v_mac_f32_e32 v10, v8, v5                               ; 2C140B08
	exp 15, 35, 0, 0, 0, v0, v14, v18, v19                  ; C400023F 13120E00
	v_mov_b32_e32 v3, 1.0                                   ; 7E0602F2
	exp 15, 36, 0, 0, 0, v4, v1, v10, v3                    ; C400024F 030A0104
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
	v_xor_b32_e32 v1, v14, v9                               ; 2A02130E
	v_mad_f32 v3, 2.0, v18, -v19                            ; D1C10003 844E24F4
	exp 15, 12, 0, 1, 0, v0, v1, v3, v19                    ; C40008CF 13030100
	s_waitcnt expcnt(0)                                     ; BF8C0F0F
Shader epilog disassembly:
	s_endpgm ; BF810000

*** SHADER STATS ***
SGPRS: 80
VGPRS: 32
Spilled VGPRs: 0
Code Size: 1840 bytes
LDS: 0 blocks
Scratch: 0 bytes per wave
Max Waves: 8
********************

Pixel Shader:
Shader main disassembly:
	s_wqm_b64 exec, exec                                          ; BEFE077E
	s_mov_b32 m0, s11                                             ; BEFC000B
	v_interp_p1_f32 v0, v2, 0, 0, [m0]                            ; D4000002
	v_interp_p2_f32 v0, [v0], v3, 0, 0, [m0]                      ; D4010003
	v_interp_p1_f32 v1, v2, 1, 0, [m0]                            ; D4040102
	v_interp_p2_f32 v1, [v1], v3, 1, 0, [m0]                      ; D4050103
	v_interp_p1_f32 v7, v2, 0, 1, [m0]                            ; D41C0402
	v_interp_p2_f32 v7, [v7], v3, 0, 1, [m0]                      ; D41D0403
	v_interp_p1_f32 v8, v2, 1, 1, [m0]                            ; D4200502
	v_interp_p2_f32 v8, [v8], v3, 1, 1, [m0]                      ; D4210503
	v_interp_p1_f32 v5, v2, 0, 2, [m0]                            ; D4140802
	v_interp_p2_f32 v5, [v5], v3, 0, 2, [m0]                      ; D4150803
	v_interp_p1_f32 v4, v2, 1, 2, [m0]                            ; D4100902
	v_interp_p2_f32 v4, [v4], v3, 1, 2, [m0]                      ; D4110903
	v_interp_p1_f32 v6, v2, 3, 2, [m0]                            ; D4180B02
	v_interp_p2_f32 v6, [v6], v3, 3, 2, [m0]                      ; D4190B03
	v_interp_p1_f32 v9, v2, 0, 3, [m0]                            ; D4240C02
	v_interp_p2_f32 v9, [v9], v3, 0, 3, [m0]                      ; D4250C03
	v_interp_p1_f32 v10, v2, 1, 3, [m0]                           ; D4280D02
	v_interp_p2_f32 v10, [v10], v3, 1, 3, [m0]                    ; D4290D03
	v_interp_p1_f32 v2, v2, 2, 3, [m0]                            ; D4080E02
	v_mul_f32_e32 v11, v9, v9                                     ; 0A161309
	v_interp_p2_f32 v2, [v2], v3, 2, 3, [m0]                      ; D4090E03
	v_mac_f32_e32 v11, v10, v10                                   ; 2C16150A
	v_mac_f32_e32 v11, v2, v2                                     ; 2C160502
	v_writelane_b32 v22, s10, 0                                   ; D28A0016 0001000A
	v_cmp_nlt_f32_e32 vcc, 0, v11                                 ; 7C9C1680
	s_and_saveexec_b64 s[0:1], vcc                                ; BE80206A
	s_xor_b64 s[0:1], exec, s[0:1]                                ; 8880007E
	v_cmp_le_f32_e32 vcc, 0, v11                                  ; 7C861680
	v_mul_f32_e32 v3, 0x7f7fffff, v11                             ; 0A0616FF 7F7FFFFF
	v_mov_b32_e32 v12, 0xff7fffff                                 ; 7E1802FF FF7FFFFF
	v_cndmask_b32_e32 v3, v12, v3, vcc                            ; 0006070C
	s_or_saveexec_b64 s[0:1], s[0:1]                              ; BE802100
	s_xor_b64 exec, exec, s[0:1]                                  ; 88FE007E
	v_rsq_f32_e32 v3, v11                                         ; 7E06490B
	s_or_b64 exec, exec, s[0:1]                                   ; 87FE007E
	s_load_dwordx4 s[16:19], s[2:3], 0x10                         ; C00A0401 00000010
	v_mul_f32_e32 v9, v9, v3                                      ; 0A120709
	v_mul_f32_e32 v11, v2, v3                                     ; 0A160702
	v_mul_f32_e32 v10, v10, v3                                    ; 0A14070A
	v_mad_f32 v9, 0, v11, -v9                                     ; D1C10009 84261680
	v_mad_f32 v10, 0, v11, -v10                                   ; D1C1000A 842A1680
	v_mac_f32_e32 v11, v2, v3                                     ; 2C160702
	v_mad_f32 v2, -v3, v2, v11                                    ; D1C10002 242E0503
	v_mov_b32_e32 v11, 0x3f13cd3a                                 ; 7E1602FF 3F13CD3A
	v_mul_f32_e32 v3, v11, v2                                     ; 0A06050B
	v_madmk_f32_e32 v12, v10, 0x3f5105ec, v3                      ; 2E18070A 3F5105EC
	v_madmk_f32_e32 v3, v9, 0xbf3504f3, v3                        ; 2E060709 BF3504F3
	v_mov_b32_e32 v14, 0xbed105ec                                 ; 7E1C02FF BED105EC
	v_mul_f32_e32 v9, 0x3f3504f3, v9                              ; 0A1212FF 3F3504F3
	v_mac_f32_e32 v9, v14, v10                                    ; 2C12150E
	v_mac_f32_e32 v9, v11, v2                                     ; 2C12050B
	v_mac_f32_e32 v3, v14, v10                                    ; 2C06150E
	v_add_f32_e64 v12, 0, v12 clamp                               ; D101800C 00021880
	v_max_f32_e32 v10, 0x358637bd, v12                            ; 161418FF 358637BD
	v_add_f32_e64 v3, 0, v3 clamp                                 ; D1018003 00020680
	v_add_f32_e64 v9, 0, v9 clamp                                 ; D1018009 00021280
	v_mov_b32_e32 v2, 0xff7fffff                                  ; 7E0402FF FF7FFFFF
	v_cmp_nge_f32_e32 vcc, 0, v10                                 ; 7C921480
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	s_and_saveexec_b64 s[6:7], vcc                                ; BE86206A
	s_xor_b64 s[6:7], exec, s[6:7]                                ; 8886067E
	v_log_f32_e32 v2, v10                                         ; 7E04430A
	s_or_b64 exec, exec, s[6:7]                                   ; 87FE067E
	s_load_dwordx4 s[0:3], s[2:3], 0x20                           ; C00A0001 00000020
	s_buffer_load_dword s23, s[16:19], 0x8                        ; C02205C8 00000008
	s_buffer_load_dword s52, s[16:19], 0xc                        ; C0220D08 0000000C
	v_max_f32_e32 v10, 0x358637bd, v3                             ; 161406FF 358637BD
	v_cmp_nge_f32_e32 vcc, 0, v10                                 ; 7C921480
	v_mov_b32_e32 v3, 0xff7fffff                                  ; 7E0602FF FF7FFFFF
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	s_and_saveexec_b64 vcc, vcc                                   ; BEEA206A
	s_xor_b64 vcc, exec, vcc                                      ; 88EA6A7E
	v_log_f32_e32 v3, v10                                         ; 7E06430A
	s_or_b64 exec, exec, vcc                                      ; 87FE6A7E
	s_buffer_load_dword s6, s[0:3], 0x150                         ; C0220180 00000150
	s_buffer_load_dword s9, s[0:3], 0x160                         ; C0220240 00000160
	s_buffer_load_dword s11, s[0:3], 0x164                        ; C02202C0 00000164
	s_buffer_load_dword s12, s[0:3], 0x168                        ; C0220300 00000168
	s_buffer_load_dword s64, s[0:3], 0x180                        ; C0221000 00000180
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	v_writelane_b32 v22, s6, 3                                    ; D28A0016 00010606
	s_buffer_load_dword s6, s[0:3], 0x154                         ; C0220180 00000154
	s_buffer_load_dword s65, s[0:3], 0x184                        ; C0221040 00000184
	s_buffer_load_dword s66, s[0:3], 0x188                        ; C0221080 00000188
	s_buffer_load_dword s67, s[0:3], 0x1c0                        ; C02210C0 000001C0
	s_buffer_load_dword s53, s[0:3], 0x1c4                        ; C0220D40 000001C4
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	v_writelane_b32 v22, s6, 2                                    ; D28A0016 00010406
	s_buffer_load_dword s6, s[0:3], 0x158                         ; C0220180 00000158
	s_buffer_load_dword s13, s[0:3], 0x1c8                        ; C0220340 000001C8
	s_buffer_load_dword s8, s[0:3], 0x1cc                         ; C0220200 000001CC
	s_buffer_load_dword s72, s[0:3], 0x1d0                        ; C0221200 000001D0
	s_buffer_load_dword s10, s[0:3], 0x2c0                        ; C0220280 000002C0
	s_buffer_load_dword s54, s[0:3], 0x2c4                        ; C0220D80 000002C4
	s_buffer_load_dword s55, s[0:3], 0x2c8                        ; C0220DC0 000002C8
	s_buffer_load_dword s21, s[0:3], 0x2d0                        ; C0220540 000002D0
	s_buffer_load_dword s22, s[0:3], 0x2d4                        ; C0220580 000002D4
	s_buffer_load_dword s20, s[0:3], 0x2d8                        ; C0220500 000002D8
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	v_writelane_b32 v22, s6, 1                                    ; D28A0016 00010206
	s_load_dwordx4 s[0:3], s[4:5], 0xb0                           ; C00A0002 000000B0
	s_load_dwordx8 s[56:63], s[4:5], 0x80                         ; C00E0E02 00000080
	s_buffer_load_dword s15, s[16:19], 0x0                        ; C02203C8 00000000
	s_buffer_load_dword s14, s[16:19], 0x4                        ; C0220388 00000004
	s_load_dwordx4 s[32:35], s[4:5], 0xf0                         ; C00A0802 000000F0
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	v_writelane_b32 v22, s0, 4                                    ; D28A0016 00010800
	v_writelane_b32 v22, s1, 5                                    ; D28A0016 00010A01
	v_writelane_b32 v22, s2, 6                                    ; D28A0016 00010C02
	v_writelane_b32 v22, s3, 7                                    ; D28A0016 00010E03
	v_writelane_b32 v22, s56, 8                                   ; D28A0016 00011038
	v_writelane_b32 v22, s57, 9                                   ; D28A0016 00011239
	v_writelane_b32 v22, s58, 10                                  ; D28A0016 0001143A
	v_writelane_b32 v22, s59, 11                                  ; D28A0016 0001163B
	v_writelane_b32 v22, s60, 12                                  ; D28A0016 0001183C
	v_writelane_b32 v22, s61, 13                                  ; D28A0016 00011A3D
	v_writelane_b32 v22, s62, 14                                  ; D28A0016 00011C3E
	v_writelane_b32 v22, s63, 15                                  ; D28A0016 00011E3F
	s_load_dwordx4 s[48:51], s[4:5], 0x130                        ; C00A0C02 00000130
	s_load_dwordx4 s[36:39], s[4:5], 0x30                         ; C00A0902 00000030
	s_load_dwordx8 s[40:47], s[4:5], 0x0                          ; C00E0A02 00000000
	s_load_dwordx8 s[24:31], s[4:5], 0x40                         ; C00E0602 00000040
	s_load_dwordx4 s[16:19], s[4:5], 0x70                         ; C00A0402 00000070
	s_load_dwordx8 s[56:63], s[4:5], 0xc0                         ; C00E0E02 000000C0
	s_load_dwordx8 s[0:7], s[4:5], 0x100                          ; C00E0002 00000100
	v_mov_b32_e32 v11, s23                                        ; 7E160217
	v_mov_b32_e32 v10, s52                                        ; 7E140234
	v_max_f32_e32 v12, 0x358637bd, v9                             ; 161812FF 358637BD
	v_cmp_nge_f32_e32 vcc, 0, v12                                 ; 7C921880
	v_mov_b32_e32 v9, 0xff7fffff                                  ; 7E1202FF FF7FFFFF
	s_waitcnt lgkmcnt(0)                                          ; BF8C007F
	s_and_saveexec_b64 vcc, vcc                                   ; BEEA206A
	s_xor_b64 vcc, exec, vcc                                      ; 88EA6A7E
	v_log_f32_e32 v9, v12                                         ; 7E12430C
	s_or_b64 exec, exec, vcc                                      ; 87FE6A7E
	v_mul_f32_e32 v17, s67, v7                                    ; 0A220E43
	v_rcp_f32_e32 v7, v6                                          ; 7E0E4506
	image_sample v[14:16], v[0:1], s[24:31], s[16:19] dmask:0x7   ; F0800700 00860E00
	v_mac_f32_e32 v2, s72, v2                                     ; 2C040448
	v_mac_f32_e32 v3, s72, v3                                     ; 2C060648
	s_waitcnt vmcnt(0)                                            ; BF8C0F70
	v_mul_f32_e32 v19, s10, v14                                   ; 0A261C0A
	v_exp_f32_e32 v2, v2                                          ; 7E044102
	v_mul_f32_e32 v18, s53, v8                                    ; 0A241035
	v_mov_b32_e32 v8, 0x7c207e81                                  ; 7E1002FF 7C207E81
	v_cmp_eq_f32_e32 vcc, 0, v6                                   ; 7C840C80
	v_cndmask_b32_e32 v6, v7, v8, vcc                             ; 000C1107
	v_mac_f32_e32 v9, s72, v9                                     ; 2C121248
	v_mul_f32_e32 v20, s54, v15                                   ; 0A281E36
	v_exp_f32_e32 v3, v3                                          ; 7E064103
	v_mul_f32_e32 v2, v2, v19                                     ; 0A042702
	v_mul_f32_e32 v5, v6, v5                                      ; 0A0A0B06
	v_mul_f32_e32 v4, v6, v4                                      ; 0A080906
	v_mac_f32_e32 v2, v3, v20                                     ; 2C042903
	v_mul_f32_e32 v21, s55, v16                                   ; 0A2A2037
	v_exp_f32_e32 v7, v9                                          ; 7E0E4109
	v_mov_b32_e32 v3, 0x3eaaaaab                                  ; 7E0602FF 3EAAAAAB
	v_mac_f32_e32 v2, v7, v21                                     ; 2C042B07
	v_mul_f32_e32 v7, v3, v19                                     ; 0A0E2703
	v_mac_f32_e32 v10, s15, v5                                    ; 2C140A0F
	v_mac_f32_e32 v11, s14, v4                                    ; 2C16080E
	image_sample v[4:6], v[10:11], s[0:7], s[48:51] dmask:0x7     ; F0800700 0180040A
	v_mac_f32_e32 v7, v3, v20                                     ; 2C0E2903
	image_sample v[10:12], v[10:11], s[56:63], s[32:35] dmask:0x7 ; F0800700 010E0A0A
	s_nop 0                                                       ; BF800000
	image_sample v[14:16], v[0:1], s[40:47], s[36:39] dmask:0x7   ; F0800700 012A0E00
	s_waitcnt vmcnt(0)                                            ; BF8C0F70
	v_mul_f32_e32 v0, s21, v14                                    ; 0A001C15
	v_mac_f32_e32 v7, v3, v21                                     ; 2C0E2B03
	v_mul_f32_e32 v1, s22, v15                                    ; 0A021E16
	v_mul_f32_e32 v3, s20, v16                                    ; 0A062014
	v_readlane_b32 s16, v22, 8                                    ; D2890010 00011116
	v_readlane_b32 s17, v22, 9                                    ; D2890011 00011316
	v_readlane_b32 s18, v22, 10                                   ; D2890012 00011516
	v_readlane_b32 s19, v22, 11                                   ; D2890013 00011716
	v_readlane_b32 s0, v22, 4                                     ; D2890000 00010916
	v_readlane_b32 s20, v22, 12                                   ; D2890014 00011916
	v_readlane_b32 s1, v22, 5                                     ; D2890001 00010B16
	v_readlane_b32 s21, v22, 13                                   ; D2890015 00011B16
	v_readlane_b32 s2, v22, 6                                     ; D2890002 00010D16
	v_readlane_b32 s22, v22, 14                                   ; D2890016 00011D16
	v_readlane_b32 s3, v22, 7                                     ; D2890003 00010F16
	v_readlane_b32 s23, v22, 15                                   ; D2890017 00011F16
	v_mad_f32 v4, v2, v0, v4                                      ; D1C10004 04120102
	v_mad_f32 v5, v2, v1, v5                                      ; D1C10005 04160302
	v_mac_f32_e32 v6, v2, v3                                      ; 2C0C0702
	v_mad_f32 v8, v7, v0, v10                                     ; D1C10008 042A0107
	v_mad_f32 v9, v7, v1, v11                                     ; D1C10009 042E0307
	image_sample v[0:2], v[17:18], s[16:23], s[0:3] dmask:0x7     ; F0800700 00040011
	v_mac_f32_e32 v12, v7, v3                                     ; 2C180707
	s_waitcnt vmcnt(0)                                            ; BF8C0F70
	v_mul_f32_e32 v3, v0, v0                                      ; 0A060100
	v_mul_f32_e32 v7, v1, v1                                      ; 0A0E0301
	v_mul_f32_e32 v11, v3, v0                                     ; 0A160103
	v_mul_f32_e32 v10, v2, v2                                     ; 0A140502
	v_mad_f32 v16, v7, v1, v11                                    ; D1C10010 042E0307
	v_mac_f32_e32 v16, v10, v2                                    ; 2C20050A
	v_mad_f32 v3, -v0, v3, v16                                    ; D1C10003 24420700
	v_mul_f32_e32 v14, v7, v1                                     ; 0A1C0307
	v_mad_f32 v7, -v1, v7, v16                                    ; D1C10007 24420F01
	v_mac_f32_e32 v11, s13, v3                                    ; 2C16060D
	v_mul_f32_e32 v15, v10, v2                                    ; 0A1E050A
	v_mad_f32 v10, -v2, v10, v16                                  ; D1C1000A 24421502
	v_mul_f32_e32 v16, s9, v0                                     ; 0A200009
	v_mul_f32_e32 v0, s64, v11                                    ; 0A001640
	v_mac_f32_e32 v14, s13, v7                                    ; 2C1C0E0D
	v_mul_f32_e32 v17, s11, v1                                    ; 0A22020B
	v_mul_f32_e32 v1, s65, v14                                    ; 0A021C41
	v_mac_f32_e32 v15, s13, v10                                   ; 2C1E140D
	v_mul_f32_e32 v0, s8, v0                                      ; 0A000008
	v_readlane_b32 s0, v22, 3                                     ; D2890000 00010716
	v_mad_f32 v0, v4, v0, s0                                      ; D1C10000 00020104
	v_mul_f32_e32 v18, s12, v2                                    ; 0A24040C
	v_mul_f32_e32 v2, s66, v15                                    ; 0A041E42
	v_mul_f32_e32 v1, s8, v1                                      ; 0A020208
	v_readlane_b32 s0, v22, 2                                     ; D2890000 00010516
	v_mad_f32 v1, v5, v1, s0                                      ; D1C10001 00020305
	v_mul_f32_e32 v2, s8, v2                                      ; 0A040408
	v_readlane_b32 s0, v22, 1                                     ; D2890000 00010316
	v_mad_f32 v2, v6, v2, s0                                      ; D1C10002 00020506
	v_mac_f32_e32 v2, v12, v18                                    ; 2C04250C
	v_mac_f32_e32 v0, v8, v16                                     ; 2C002108
	v_mac_f32_e32 v1, v9, v17                                     ; 2C022309
	v_readlane_b32 s10, v22, 0                                    ; D289000A 00010116
	v_mov_b32_e32 v3, v2                                          ; 7E060302
Shader epilog disassembly:
	v_cvt_pkrtz_f16_f32_e64 v0, v0, v1 ; D2960000 00020300
	v_cvt_pkrtz_f16_f32_e64 v1, v2, v3 ; D2960001 00020702
	exp 15, 0, 1, 1, 1, v0, v1, v0, v0 ; C4001C0F 00000100
	s_endpgm                           ; BF810000

*** SHADER CONFIG ***
SPI_PS_INPUT_ADDR = 0xd077
SPI_PS_INPUT_ENA  = 0x0002
*** SHADER STATS ***
SGPRS: 80
VGPRS: 24
Spilled VGPRs: 0
Code Size: 1356 bytes
LDS: 0 blocks
Scratch: 0 bytes per wave
Max Waves: 10
********************
[...]